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 * Switches for gnatxref::
394 * Switches for gnatfind::
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 * Example of gnatcheck Usage::
497 Sample Bodies Using gnatstub
500 * Switches for gnatstub::
502 Other Utility Programs
504 * Using Other Utility Programs with GNAT::
505 * The External Symbol Naming Scheme of GNAT::
506 * Converting Ada Files to html with gnathtml::
509 Code Coverage and Profiling
511 * Code Coverage of Ada Programs using gcov::
512 * Profiling an Ada Program using gprof::
515 Running and Debugging Ada Programs
517 * The GNAT Debugger GDB::
519 * Introduction to GDB Commands::
520 * Using Ada Expressions::
521 * Calling User-Defined Subprograms::
522 * Using the Next Command in a Function::
525 * Debugging Generic Units::
526 * GNAT Abnormal Termination or Failure to Terminate::
527 * Naming Conventions for GNAT Source Files::
528 * Getting Internal Debugging Information::
536 Compatibility with HP Ada
538 * Ada Language Compatibility::
539 * Differences in the Definition of Package System::
540 * Language-Related Features::
541 * The Package STANDARD::
542 * The Package SYSTEM::
543 * Tasking and Task-Related Features::
544 * Pragmas and Pragma-Related Features::
545 * Library of Predefined Units::
547 * Main Program Definition::
548 * Implementation-Defined Attributes::
549 * Compiler and Run-Time Interfacing::
550 * Program Compilation and Library Management::
552 * Implementation Limits::
553 * Tools and Utilities::
555 Language-Related Features
557 * Integer Types and Representations::
558 * Floating-Point Types and Representations::
559 * Pragmas Float_Representation and Long_Float::
560 * Fixed-Point Types and Representations::
561 * Record and Array Component Alignment::
563 * Other Representation Clauses::
565 Tasking and Task-Related Features
567 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
568 * Assigning Task IDs::
569 * Task IDs and Delays::
570 * Task-Related Pragmas::
571 * Scheduling and Task Priority::
573 * External Interrupts::
575 Pragmas and Pragma-Related Features
577 * Restrictions on the Pragma INLINE::
578 * Restrictions on the Pragma INTERFACE::
579 * Restrictions on the Pragma SYSTEM_NAME::
581 Library of Predefined Units
583 * Changes to DECLIB::
587 * Shared Libraries and Options Files::
591 Platform-Specific Information for the Run-Time Libraries
593 * Summary of Run-Time Configurations::
594 * Specifying a Run-Time Library::
595 * Choosing the Scheduling Policy::
596 * Solaris-Specific Considerations::
597 * Linux-Specific Considerations::
598 * AIX-Specific Considerations::
599 * Irix-Specific Considerations::
601 Example of Binder Output File
603 Elaboration Order Handling in GNAT
606 * Checking the Elaboration Order::
607 * Controlling the Elaboration Order::
608 * Controlling Elaboration in GNAT - Internal Calls::
609 * Controlling Elaboration in GNAT - External Calls::
610 * Default Behavior in GNAT - Ensuring Safety::
611 * Treatment of Pragma Elaborate::
612 * Elaboration Issues for Library Tasks::
613 * Mixing Elaboration Models::
614 * What to Do If the Default Elaboration Behavior Fails::
615 * Elaboration for Access-to-Subprogram Values::
616 * Summary of Procedures for Elaboration Control::
617 * Other Elaboration Order Considerations::
619 Conditional Compilation
620 * Use of Boolean Constants::
621 * Debugging - A Special Case::
622 * Conditionalizing Declarations::
623 * Use of Alternative Implementations::
628 * Basic Assembler Syntax::
629 * A Simple Example of Inline Assembler::
630 * Output Variables in Inline Assembler::
631 * Input Variables in Inline Assembler::
632 * Inlining Inline Assembler Code::
633 * Other Asm Functionality::
635 Compatibility and Porting Guide
637 * Compatibility with Ada 83::
638 * Compatibility between Ada 95 and Ada 2005::
639 * Implementation-dependent characteristics::
641 @c This brief section is only in the non-VMS version
642 @c The complete chapter on HP Ada issues is in the VMS version
643 * Compatibility with HP Ada 83::
645 * Compatibility with Other Ada Systems::
646 * Representation Clauses::
648 * Transitioning to 64-Bit GNAT for OpenVMS::
652 Microsoft Windows Topics
654 * Using GNAT on Windows::
655 * CONSOLE and WINDOWS subsystems::
657 * Mixed-Language Programming on Windows::
658 * Windows Calling Conventions::
659 * Introduction to Dynamic Link Libraries (DLLs)::
660 * Using DLLs with GNAT::
661 * Building DLLs with GNAT::
662 * GNAT and Windows Resources::
664 * Setting Stack Size from gnatlink::
665 * Setting Heap Size from gnatlink::
672 @node About This Guide
673 @unnumbered About This Guide
677 This guide describes the use of @value{EDITION},
678 a compiler and software development toolset for the full Ada
679 programming language, implemented on OpenVMS for HP's Alpha and
680 Integrity server (I64) platforms.
683 This guide describes the use of @value{EDITION},
684 a compiler and software development
685 toolset for the full Ada programming language.
687 It documents the features of the compiler and tools, and explains
688 how to use them to build Ada applications.
690 @value{EDITION} implements Ada 95 and Ada 2005, and it may also be invoked in
691 Ada 83 compatibility mode.
692 By default, @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
693 but you can override with a compiler switch
694 (@pxref{Compiling Different Versions of Ada})
695 to explicitly specify the language version.
696 Throughout this manual, references to ``Ada'' without a year suffix
697 apply to both the Ada 95 and Ada 2005 versions of the language.
701 For ease of exposition, ``@value{EDITION}'' will be referred to simply as
702 ``GNAT'' in the remainder of this document.
709 * What This Guide Contains::
710 * What You Should Know before Reading This Guide::
711 * Related Information::
715 @node What This Guide Contains
716 @unnumberedsec What This Guide Contains
719 This guide contains the following chapters:
723 @ref{Getting Started with GNAT}, describes how to get started compiling
724 and running Ada programs with the GNAT Ada programming environment.
726 @ref{The GNAT Compilation Model}, describes the compilation model used
730 @ref{Compiling Using gcc}, describes how to compile
731 Ada programs with @command{gcc}, the Ada compiler.
734 @ref{Binding Using gnatbind}, describes how to
735 perform binding of Ada programs with @code{gnatbind}, the GNAT binding
739 @ref{Linking Using gnatlink},
740 describes @command{gnatlink}, a
741 program that provides for linking using the GNAT run-time library to
742 construct a program. @command{gnatlink} can also incorporate foreign language
743 object units into the executable.
746 @ref{The GNAT Make Program gnatmake}, describes @command{gnatmake}, a
747 utility that automatically determines the set of sources
748 needed by an Ada compilation unit, and executes the necessary compilations
752 @ref{Improving Performance}, shows various techniques for making your
753 Ada program run faster or take less space.
754 It discusses the effect of the compiler's optimization switch and
755 also describes the @command{gnatelim} tool and unused subprogram/data
759 @ref{Renaming Files Using gnatchop}, describes
760 @code{gnatchop}, a utility that allows you to preprocess a file that
761 contains Ada source code, and split it into one or more new files, one
762 for each compilation unit.
765 @ref{Configuration Pragmas}, describes the configuration pragmas
769 @ref{Handling Arbitrary File Naming Conventions Using gnatname},
770 shows how to override the default GNAT file naming conventions,
771 either for an individual unit or globally.
774 @ref{GNAT Project Manager}, describes how to use project files
775 to organize large projects.
778 @ref{The Cross-Referencing Tools gnatxref and gnatfind}, discusses
779 @code{gnatxref} and @code{gnatfind}, two tools that provide an easy
780 way to navigate through sources.
783 @ref{The GNAT Pretty-Printer gnatpp}, shows how to produce a reformatted
784 version of an Ada source file with control over casing, indentation,
785 comment placement, and other elements of program presentation style.
788 @ref{The GNAT Metric Tool gnatmetric}, shows how to compute various
789 metrics for an Ada source file, such as the number of types and subprograms,
790 and assorted complexity measures.
793 @ref{File Name Krunching Using gnatkr}, describes the @code{gnatkr}
794 file name krunching utility, used to handle shortened
795 file names on operating systems with a limit on the length of names.
798 @ref{Preprocessing Using gnatprep}, describes @code{gnatprep}, a
799 preprocessor utility that allows a single source file to be used to
800 generate multiple or parameterized source files by means of macro
805 @ref{The GNAT Run-Time Library Builder gnatlbr}, describes @command{gnatlbr},
806 a tool for rebuilding the GNAT run time with user-supplied
807 configuration pragmas.
811 @ref{The GNAT Library Browser gnatls}, describes @code{gnatls}, a
812 utility that displays information about compiled units, including dependences
813 on the corresponding sources files, and consistency of compilations.
816 @ref{Cleaning Up Using gnatclean}, describes @code{gnatclean}, a utility
817 to delete files that are produced by the compiler, binder and linker.
821 @ref{GNAT and Libraries}, describes the process of creating and using
822 Libraries with GNAT. It also describes how to recompile the GNAT run-time
826 @ref{Using the GNU make Utility}, describes some techniques for using
827 the GNAT toolset in Makefiles.
831 @ref{Memory Management Issues}, describes some useful predefined storage pools
832 and in particular the GNAT Debug Pool facility, which helps detect incorrect
835 It also describes @command{gnatmem}, a utility that monitors dynamic
836 allocation and deallocation and helps detect ``memory leaks''.
840 @ref{Stack Related Facilities}, describes some useful tools associated with
841 stack checking and analysis.
844 @ref{Verifying Properties Using gnatcheck}, discusses @code{gnatcheck},
845 a utility that checks Ada code against a set of rules.
848 @ref{Creating Sample Bodies Using gnatstub}, discusses @code{gnatstub},
849 a utility that generates empty but compilable bodies for library units.
852 @ref{Generating Ada Bindings for C and C++ headers}, describes how to
853 generate automatically Ada bindings from C and C++ headers.
856 @ref{Other Utility Programs}, discusses several other GNAT utilities,
857 including @code{gnathtml}.
861 @ref{Code Coverage and Profiling}, describes how to perform a structural
862 coverage and profile the execution of Ada programs.
866 @ref{Running and Debugging Ada Programs}, describes how to run and debug
871 @ref{Compatibility with HP Ada}, details the compatibility of GNAT with
872 HP Ada 83 @footnote{``HP Ada'' refers to the legacy product originally
873 developed by Digital Equipment Corporation and currently supported by HP.}
874 for OpenVMS Alpha. This product was formerly known as DEC Ada,
877 historical compatibility reasons, the relevant libraries still use the
882 @ref{Platform-Specific Information for the Run-Time Libraries},
883 describes the various run-time
884 libraries supported by GNAT on various platforms and explains how to
885 choose a particular library.
888 @ref{Example of Binder Output File}, shows the source code for the binder
889 output file for a sample program.
892 @ref{Elaboration Order Handling in GNAT}, describes how GNAT helps
893 you deal with elaboration order issues.
896 @ref{Conditional Compilation}, describes how to model conditional compilation,
897 both with Ada in general and with GNAT facilities in particular.
900 @ref{Inline Assembler}, shows how to use the inline assembly facility
904 @ref{Compatibility and Porting Guide}, contains sections on compatibility
905 of GNAT with other Ada development environments (including Ada 83 systems),
906 to assist in porting code from those environments.
910 @ref{Microsoft Windows Topics}, presents information relevant to the
911 Microsoft Windows platform.
915 @c *************************************************
916 @node What You Should Know before Reading This Guide
917 @c *************************************************
918 @unnumberedsec What You Should Know before Reading This Guide
920 @cindex Ada 95 Language Reference Manual
921 @cindex Ada 2005 Language Reference Manual
923 This guide assumes a basic familiarity with the Ada 95 language, as
924 described in the International Standard ANSI/ISO/IEC-8652:1995, January
926 It does not require knowledge of the new features introduced by Ada 2005,
927 (officially known as ISO/IEC 8652:1995 with Technical Corrigendum 1
929 Both reference manuals are included in the GNAT documentation
932 @node Related Information
933 @unnumberedsec Related Information
936 For further information about related tools, refer to the following
941 @xref{Top, GNAT Reference Manual, About This Guide, gnat_rm, GNAT
942 Reference Manual}, which contains all reference material for the GNAT
943 implementation of Ada.
947 @cite{Using the GNAT Programming Studio}, which describes the GPS
948 Integrated Development Environment.
951 @cite{GNAT Programming Studio Tutorial}, which introduces the
952 main GPS features through examples.
956 @cite{Ada 95 Reference Manual}, which contains reference
957 material for the Ada 95 programming language.
960 @cite{Ada 2005 Reference Manual}, which contains reference
961 material for the Ada 2005 programming language.
964 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
966 in the GNU:[DOCS] directory,
968 for all details on the use of the GNU source-level debugger.
971 @xref{Top,, The extensible self-documenting text editor, emacs,
974 located in the GNU:[DOCS] directory if the EMACS kit is installed,
976 for full information on the extensible editor and programming
983 @unnumberedsec Conventions
985 @cindex Typographical conventions
988 Following are examples of the typographical and graphic conventions used
993 @code{Functions}, @command{utility program names}, @code{standard names},
997 @option{Option flags}
1000 @file{File names}, @samp{button names}, and @samp{field names}.
1003 @code{Variables}, @env{environment variables}, and @var{metasyntactic
1010 @r{[}optional information or parameters@r{]}
1013 Examples are described by text
1015 and then shown this way.
1020 Commands that are entered by the user are preceded in this manual by the
1021 characters @w{``@code{$ }''} (dollar sign followed by space). If your system
1022 uses this sequence as a prompt, then the commands will appear exactly as
1023 you see them in the manual. If your system uses some other prompt, then
1024 the command will appear with the @code{$} replaced by whatever prompt
1025 character you are using.
1028 Full file names are shown with the ``@code{/}'' character
1029 as the directory separator; e.g., @file{parent-dir/subdir/myfile.adb}.
1030 If you are using GNAT on a Windows platform, please note that
1031 the ``@code{\}'' character should be used instead.
1034 @c ****************************
1035 @node Getting Started with GNAT
1036 @chapter Getting Started with GNAT
1039 This chapter describes some simple ways of using GNAT to build
1040 executable Ada programs.
1042 @ref{Running GNAT}, through @ref{Using the gnatmake Utility},
1043 show how to use the command line environment.
1044 @ref{Introduction to GPS}, provides a brief
1045 introduction to the GNAT Programming Studio, a visually-oriented
1046 Integrated Development Environment for GNAT.
1047 GPS offers a graphical ``look and feel'', support for development in
1048 other programming languages, comprehensive browsing features, and
1049 many other capabilities.
1050 For information on GPS please refer to
1051 @cite{Using the GNAT Programming Studio}.
1056 * Running a Simple Ada Program::
1057 * Running a Program with Multiple Units::
1058 * Using the gnatmake Utility::
1060 * Editing with Emacs::
1063 * Introduction to GPS::
1068 @section Running GNAT
1071 Three steps are needed to create an executable file from an Ada source
1076 The source file(s) must be compiled.
1078 The file(s) must be bound using the GNAT binder.
1080 All appropriate object files must be linked to produce an executable.
1084 All three steps are most commonly handled by using the @command{gnatmake}
1085 utility program that, given the name of the main program, automatically
1086 performs the necessary compilation, binding and linking steps.
1088 @node Running a Simple Ada Program
1089 @section Running a Simple Ada Program
1092 Any text editor may be used to prepare an Ada program.
1094 used, the optional Ada mode may be helpful in laying out the program.)
1096 program text is a normal text file. We will assume in our initial
1097 example that you have used your editor to prepare the following
1098 standard format text file:
1100 @smallexample @c ada
1102 with Ada.Text_IO; use Ada.Text_IO;
1105 Put_Line ("Hello WORLD!");
1111 This file should be named @file{hello.adb}.
1112 With the normal default file naming conventions, GNAT requires
1114 contain a single compilation unit whose file name is the
1116 with periods replaced by hyphens; the
1117 extension is @file{ads} for a
1118 spec and @file{adb} for a body.
1119 You can override this default file naming convention by use of the
1120 special pragma @code{Source_File_Name} (@pxref{Using Other File Names}).
1121 Alternatively, if you want to rename your files according to this default
1122 convention, which is probably more convenient if you will be using GNAT
1123 for all your compilations, then the @code{gnatchop} utility
1124 can be used to generate correctly-named source files
1125 (@pxref{Renaming Files Using gnatchop}).
1127 You can compile the program using the following command (@code{$} is used
1128 as the command prompt in the examples in this document):
1135 @command{gcc} is the command used to run the compiler. This compiler is
1136 capable of compiling programs in several languages, including Ada and
1137 C. It assumes that you have given it an Ada program if the file extension is
1138 either @file{.ads} or @file{.adb}, and it will then call
1139 the GNAT compiler to compile the specified file.
1142 The @option{-c} switch is required. It tells @command{gcc} to only do a
1143 compilation. (For C programs, @command{gcc} can also do linking, but this
1144 capability is not used directly for Ada programs, so the @option{-c}
1145 switch must always be present.)
1148 This compile command generates a file
1149 @file{hello.o}, which is the object
1150 file corresponding to your Ada program. It also generates
1151 an ``Ada Library Information'' file @file{hello.ali},
1152 which contains additional information used to check
1153 that an Ada program is consistent.
1154 To build an executable file,
1155 use @code{gnatbind} to bind the program
1156 and @command{gnatlink} to link it. The
1157 argument to both @code{gnatbind} and @command{gnatlink} is the name of the
1158 @file{ALI} file, but the default extension of @file{.ali} can
1159 be omitted. This means that in the most common case, the argument
1160 is simply the name of the main program:
1168 A simpler method of carrying out these steps is to use
1170 a master program that invokes all the required
1171 compilation, binding and linking tools in the correct order. In particular,
1172 @command{gnatmake} automatically recompiles any sources that have been
1173 modified since they were last compiled, or sources that depend
1174 on such modified sources, so that ``version skew'' is avoided.
1175 @cindex Version skew (avoided by @command{gnatmake})
1178 $ gnatmake hello.adb
1182 The result is an executable program called @file{hello}, which can be
1190 assuming that the current directory is on the search path
1191 for executable programs.
1194 and, if all has gone well, you will see
1201 appear in response to this command.
1203 @c ****************************************
1204 @node Running a Program with Multiple Units
1205 @section Running a Program with Multiple Units
1208 Consider a slightly more complicated example that has three files: a
1209 main program, and the spec and body of a package:
1211 @smallexample @c ada
1214 package Greetings is
1219 with Ada.Text_IO; use Ada.Text_IO;
1220 package body Greetings is
1223 Put_Line ("Hello WORLD!");
1226 procedure Goodbye is
1228 Put_Line ("Goodbye WORLD!");
1245 Following the one-unit-per-file rule, place this program in the
1246 following three separate files:
1250 spec of package @code{Greetings}
1253 body of package @code{Greetings}
1256 body of main program
1260 To build an executable version of
1261 this program, we could use four separate steps to compile, bind, and link
1262 the program, as follows:
1266 $ gcc -c greetings.adb
1272 Note that there is no required order of compilation when using GNAT.
1273 In particular it is perfectly fine to compile the main program first.
1274 Also, it is not necessary to compile package specs in the case where
1275 there is an accompanying body; you only need to compile the body. If you want
1276 to submit these files to the compiler for semantic checking and not code
1277 generation, then use the
1278 @option{-gnatc} switch:
1281 $ gcc -c greetings.ads -gnatc
1285 Although the compilation can be done in separate steps as in the
1286 above example, in practice it is almost always more convenient
1287 to use the @command{gnatmake} tool. All you need to know in this case
1288 is the name of the main program's source file. The effect of the above four
1289 commands can be achieved with a single one:
1292 $ gnatmake gmain.adb
1296 In the next section we discuss the advantages of using @command{gnatmake} in
1299 @c *****************************
1300 @node Using the gnatmake Utility
1301 @section Using the @command{gnatmake} Utility
1304 If you work on a program by compiling single components at a time using
1305 @command{gcc}, you typically keep track of the units you modify. In order to
1306 build a consistent system, you compile not only these units, but also any
1307 units that depend on the units you have modified.
1308 For example, in the preceding case,
1309 if you edit @file{gmain.adb}, you only need to recompile that file. But if
1310 you edit @file{greetings.ads}, you must recompile both
1311 @file{greetings.adb} and @file{gmain.adb}, because both files contain
1312 units that depend on @file{greetings.ads}.
1314 @code{gnatbind} will warn you if you forget one of these compilation
1315 steps, so that it is impossible to generate an inconsistent program as a
1316 result of forgetting to do a compilation. Nevertheless it is tedious and
1317 error-prone to keep track of dependencies among units.
1318 One approach to handle the dependency-bookkeeping is to use a
1319 makefile. However, makefiles present maintenance problems of their own:
1320 if the dependencies change as you change the program, you must make
1321 sure that the makefile is kept up-to-date manually, which is also an
1322 error-prone process.
1324 The @command{gnatmake} utility takes care of these details automatically.
1325 Invoke it using either one of the following forms:
1328 $ gnatmake gmain.adb
1329 $ gnatmake ^gmain^GMAIN^
1333 The argument is the name of the file containing the main program;
1334 you may omit the extension. @command{gnatmake}
1335 examines the environment, automatically recompiles any files that need
1336 recompiling, and binds and links the resulting set of object files,
1337 generating the executable file, @file{^gmain^GMAIN.EXE^}.
1338 In a large program, it
1339 can be extremely helpful to use @command{gnatmake}, because working out by hand
1340 what needs to be recompiled can be difficult.
1342 Note that @command{gnatmake}
1343 takes into account all the Ada rules that
1344 establish dependencies among units. These include dependencies that result
1345 from inlining subprogram bodies, and from
1346 generic instantiation. Unlike some other
1347 Ada make tools, @command{gnatmake} does not rely on the dependencies that were
1348 found by the compiler on a previous compilation, which may possibly
1349 be wrong when sources change. @command{gnatmake} determines the exact set of
1350 dependencies from scratch each time it is run.
1353 @node Editing with Emacs
1354 @section Editing with Emacs
1358 Emacs is an extensible self-documenting text editor that is available in a
1359 separate VMSINSTAL kit.
1361 Invoke Emacs by typing @kbd{Emacs} at the command prompt. To get started,
1362 click on the Emacs Help menu and run the Emacs Tutorial.
1363 In a character cell terminal, Emacs help is invoked with @kbd{Ctrl-h} (also
1364 written as @kbd{C-h}), and the tutorial by @kbd{C-h t}.
1366 Documentation on Emacs and other tools is available in Emacs under the
1367 pull-down menu button: @code{Help - Info}. After selecting @code{Info},
1368 use the middle mouse button to select a topic (e.g.@: Emacs).
1370 In a character cell terminal, do @kbd{C-h i} to invoke info, and then @kbd{m}
1371 (stands for menu) followed by the menu item desired, as in @kbd{m Emacs}, to
1372 get to the Emacs manual.
1373 Help on Emacs is also available by typing @kbd{HELP EMACS} at the DCL command
1376 The tutorial is highly recommended in order to learn the intricacies of Emacs,
1377 which is sufficiently extensible to provide for a complete programming
1378 environment and shell for the sophisticated user.
1382 @node Introduction to GPS
1383 @section Introduction to GPS
1384 @cindex GPS (GNAT Programming Studio)
1385 @cindex GNAT Programming Studio (GPS)
1387 Although the command line interface (@command{gnatmake}, etc.) alone
1388 is sufficient, a graphical Interactive Development
1389 Environment can make it easier for you to compose, navigate, and debug
1390 programs. This section describes the main features of GPS
1391 (``GNAT Programming Studio''), the GNAT graphical IDE.
1392 You will see how to use GPS to build and debug an executable, and
1393 you will also learn some of the basics of the GNAT ``project'' facility.
1395 GPS enables you to do much more than is presented here;
1396 e.g., you can produce a call graph, interface to a third-party
1397 Version Control System, and inspect the generated assembly language
1399 Indeed, GPS also supports languages other than Ada.
1400 Such additional information, and an explanation of all of the GPS menu
1401 items. may be found in the on-line help, which includes
1402 a user's guide and a tutorial (these are also accessible from the GNAT
1406 * Building a New Program with GPS::
1407 * Simple Debugging with GPS::
1410 @node Building a New Program with GPS
1411 @subsection Building a New Program with GPS
1413 GPS invokes the GNAT compilation tools using information
1414 contained in a @emph{project} (also known as a @emph{project file}):
1415 a collection of properties such
1416 as source directories, identities of main subprograms, tool switches, etc.,
1417 and their associated values.
1418 See @ref{GNAT Project Manager} for details.
1419 In order to run GPS, you will need to either create a new project
1420 or else open an existing one.
1422 This section will explain how you can use GPS to create a project,
1423 to associate Ada source files with a project, and to build and run
1427 @item @emph{Creating a project}
1429 Invoke GPS, either from the command line or the platform's IDE.
1430 After it starts, GPS will display a ``Welcome'' screen with three
1435 @code{Start with default project in directory}
1438 @code{Create new project with wizard}
1441 @code{Open existing project}
1445 Select @code{Create new project with wizard} and press @code{OK}.
1446 A new window will appear. In the text box labeled with
1447 @code{Enter the name of the project to create}, type @file{sample}
1448 as the project name.
1449 In the next box, browse to choose the directory in which you
1450 would like to create the project file.
1451 After selecting an appropriate directory, press @code{Forward}.
1453 A window will appear with the title
1454 @code{Version Control System Configuration}.
1455 Simply press @code{Forward}.
1457 A window will appear with the title
1458 @code{Please select the source directories for this project}.
1459 The directory that you specified for the project file will be selected
1460 by default as the one to use for sources; simply press @code{Forward}.
1462 A window will appear with the title
1463 @code{Please select the build directory for this project}.
1464 The directory that you specified for the project file will be selected
1465 by default for object files and executables;
1466 simply press @code{Forward}.
1468 A window will appear with the title
1469 @code{Please select the main units for this project}.
1470 You will supply this information later, after creating the source file.
1471 Simply press @code{Forward} for now.
1473 A window will appear with the title
1474 @code{Please select the switches to build the project}.
1475 Press @code{Apply}. This will create a project file named
1476 @file{sample.prj} in the directory that you had specified.
1478 @item @emph{Creating and saving the source file}
1480 After you create the new project, a GPS window will appear, which is
1481 partitioned into two main sections:
1485 A @emph{Workspace area}, initially greyed out, which you will use for
1486 creating and editing source files
1489 Directly below, a @emph{Messages area}, which initially displays a
1490 ``Welcome'' message.
1491 (If the Messages area is not visible, drag its border upward to expand it.)
1495 Select @code{File} on the menu bar, and then the @code{New} command.
1496 The Workspace area will become white, and you can now
1497 enter the source program explicitly.
1498 Type the following text
1500 @smallexample @c ada
1502 with Ada.Text_IO; use Ada.Text_IO;
1505 Put_Line("Hello from GPS!");
1511 Select @code{File}, then @code{Save As}, and enter the source file name
1513 The file will be saved in the same directory you specified as the
1514 location of the default project file.
1516 @item @emph{Updating the project file}
1518 You need to add the new source file to the project.
1520 the @code{Project} menu and then @code{Edit project properties}.
1521 Click the @code{Main files} tab on the left, and then the
1523 Choose @file{hello.adb} from the list, and press @code{Open}.
1524 The project settings window will reflect this action.
1527 @item @emph{Building and running the program}
1529 In the main GPS window, now choose the @code{Build} menu, then @code{Make},
1530 and select @file{hello.adb}.
1531 The Messages window will display the resulting invocations of @command{gcc},
1532 @command{gnatbind}, and @command{gnatlink}
1533 (reflecting the default switch settings from the
1534 project file that you created) and then a ``successful compilation/build''
1537 To run the program, choose the @code{Build} menu, then @code{Run}, and
1538 select @command{hello}.
1539 An @emph{Arguments Selection} window will appear.
1540 There are no command line arguments, so just click @code{OK}.
1542 The Messages window will now display the program's output (the string
1543 @code{Hello from GPS}), and at the bottom of the GPS window a status
1544 update is displayed (@code{Run: hello}).
1545 Close the GPS window (or select @code{File}, then @code{Exit}) to
1546 terminate this GPS session.
1549 @node Simple Debugging with GPS
1550 @subsection Simple Debugging with GPS
1552 This section illustrates basic debugging techniques (setting breakpoints,
1553 examining/modifying variables, single stepping).
1556 @item @emph{Opening a project}
1558 Start GPS and select @code{Open existing project}; browse to
1559 specify the project file @file{sample.prj} that you had created in the
1562 @item @emph{Creating a source file}
1564 Select @code{File}, then @code{New}, and type in the following program:
1566 @smallexample @c ada
1568 with Ada.Text_IO; use Ada.Text_IO;
1569 procedure Example is
1570 Line : String (1..80);
1573 Put_Line("Type a line of text at each prompt; an empty line to exit");
1577 Put_Line (Line (1..N) );
1585 Select @code{File}, then @code{Save as}, and enter the file name
1588 @item @emph{Updating the project file}
1590 Add @code{Example} as a new main unit for the project:
1593 Select @code{Project}, then @code{Edit Project Properties}.
1596 Select the @code{Main files} tab, click @code{Add}, then
1597 select the file @file{example.adb} from the list, and
1599 You will see the file name appear in the list of main units
1605 @item @emph{Building/running the executable}
1607 To build the executable
1608 select @code{Build}, then @code{Make}, and then choose @file{example.adb}.
1610 Run the program to see its effect (in the Messages area).
1611 Each line that you enter is displayed; an empty line will
1612 cause the loop to exit and the program to terminate.
1614 @item @emph{Debugging the program}
1616 Note that the @option{-g} switches to @command{gcc} and @command{gnatlink},
1617 which are required for debugging, are on by default when you create
1619 Thus unless you intentionally remove these settings, you will be able
1620 to debug any program that you develop using GPS.
1623 @item @emph{Initializing}
1625 Select @code{Debug}, then @code{Initialize}, then @file{example}
1627 @item @emph{Setting a breakpoint}
1629 After performing the initialization step, you will observe a small
1630 icon to the right of each line number.
1631 This serves as a toggle for breakpoints; clicking the icon will
1632 set a breakpoint at the corresponding line (the icon will change to
1633 a red circle with an ``x''), and clicking it again
1634 will remove the breakpoint / reset the icon.
1636 For purposes of this example, set a breakpoint at line 10 (the
1637 statement @code{Put_Line@ (Line@ (1..N));}
1639 @item @emph{Starting program execution}
1641 Select @code{Debug}, then @code{Run}. When the
1642 @code{Program Arguments} window appears, click @code{OK}.
1643 A console window will appear; enter some line of text,
1644 e.g.@: @code{abcde}, at the prompt.
1645 The program will pause execution when it gets to the
1646 breakpoint, and the corresponding line is highlighted.
1648 @item @emph{Examining a variable}
1650 Move the mouse over one of the occurrences of the variable @code{N}.
1651 You will see the value (5) displayed, in ``tool tip'' fashion.
1652 Right click on @code{N}, select @code{Debug}, then select @code{Display N}.
1653 You will see information about @code{N} appear in the @code{Debugger Data}
1654 pane, showing the value as 5.
1656 @item @emph{Assigning a new value to a variable}
1658 Right click on the @code{N} in the @code{Debugger Data} pane, and
1659 select @code{Set value of N}.
1660 When the input window appears, enter the value @code{4} and click
1662 This value does not automatically appear in the @code{Debugger Data}
1663 pane; to see it, right click again on the @code{N} in the
1664 @code{Debugger Data} pane and select @code{Update value}.
1665 The new value, 4, will appear in red.
1667 @item @emph{Single stepping}
1669 Select @code{Debug}, then @code{Next}.
1670 This will cause the next statement to be executed, in this case the
1671 call of @code{Put_Line} with the string slice.
1672 Notice in the console window that the displayed string is simply
1673 @code{abcd} and not @code{abcde} which you had entered.
1674 This is because the upper bound of the slice is now 4 rather than 5.
1676 @item @emph{Removing a breakpoint}
1678 Toggle the breakpoint icon at line 10.
1680 @item @emph{Resuming execution from a breakpoint}
1682 Select @code{Debug}, then @code{Continue}.
1683 The program will reach the next iteration of the loop, and
1684 wait for input after displaying the prompt.
1685 This time, just hit the @kbd{Enter} key.
1686 The value of @code{N} will be 0, and the program will terminate.
1687 The console window will disappear.
1692 @node The GNAT Compilation Model
1693 @chapter The GNAT Compilation Model
1694 @cindex GNAT compilation model
1695 @cindex Compilation model
1698 * Source Representation::
1699 * Foreign Language Representation::
1700 * File Naming Rules::
1701 * Using Other File Names::
1702 * Alternative File Naming Schemes::
1703 * Generating Object Files::
1704 * Source Dependencies::
1705 * The Ada Library Information Files::
1706 * Binding an Ada Program::
1707 * Mixed Language Programming::
1709 * Building Mixed Ada & C++ Programs::
1710 * Comparison between GNAT and C/C++ Compilation Models::
1712 * Comparison between GNAT and Conventional Ada Library Models::
1714 * Placement of temporary files::
1719 This chapter describes the compilation model used by GNAT. Although
1720 similar to that used by other languages, such as C and C++, this model
1721 is substantially different from the traditional Ada compilation models,
1722 which are based on a library. The model is initially described without
1723 reference to the library-based model. If you have not previously used an
1724 Ada compiler, you need only read the first part of this chapter. The
1725 last section describes and discusses the differences between the GNAT
1726 model and the traditional Ada compiler models. If you have used other
1727 Ada compilers, this section will help you to understand those
1728 differences, and the advantages of the GNAT model.
1730 @node Source Representation
1731 @section Source Representation
1735 Ada source programs are represented in standard text files, using
1736 Latin-1 coding. Latin-1 is an 8-bit code that includes the familiar
1737 7-bit ASCII set, plus additional characters used for
1738 representing foreign languages (@pxref{Foreign Language Representation}
1739 for support of non-USA character sets). The format effector characters
1740 are represented using their standard ASCII encodings, as follows:
1745 Vertical tab, @code{16#0B#}
1749 Horizontal tab, @code{16#09#}
1753 Carriage return, @code{16#0D#}
1757 Line feed, @code{16#0A#}
1761 Form feed, @code{16#0C#}
1765 Source files are in standard text file format. In addition, GNAT will
1766 recognize a wide variety of stream formats, in which the end of
1767 physical lines is marked by any of the following sequences:
1768 @code{LF}, @code{CR}, @code{CR-LF}, or @code{LF-CR}. This is useful
1769 in accommodating files that are imported from other operating systems.
1771 @cindex End of source file
1772 @cindex Source file, end
1774 The end of a source file is normally represented by the physical end of
1775 file. However, the control character @code{16#1A#} (@code{SUB}) is also
1776 recognized as signalling the end of the source file. Again, this is
1777 provided for compatibility with other operating systems where this
1778 code is used to represent the end of file.
1780 Each file contains a single Ada compilation unit, including any pragmas
1781 associated with the unit. For example, this means you must place a
1782 package declaration (a package @dfn{spec}) and the corresponding body in
1783 separate files. An Ada @dfn{compilation} (which is a sequence of
1784 compilation units) is represented using a sequence of files. Similarly,
1785 you will place each subunit or child unit in a separate file.
1787 @node Foreign Language Representation
1788 @section Foreign Language Representation
1791 GNAT supports the standard character sets defined in Ada as well as
1792 several other non-standard character sets for use in localized versions
1793 of the compiler (@pxref{Character Set Control}).
1796 * Other 8-Bit Codes::
1797 * Wide Character Encodings::
1805 The basic character set is Latin-1. This character set is defined by ISO
1806 standard 8859, part 1. The lower half (character codes @code{16#00#}
1807 @dots{} @code{16#7F#)} is identical to standard ASCII coding, but the upper half
1808 is used to represent additional characters. These include extended letters
1809 used by European languages, such as French accents, the vowels with umlauts
1810 used in German, and the extra letter A-ring used in Swedish.
1812 @findex Ada.Characters.Latin_1
1813 For a complete list of Latin-1 codes and their encodings, see the source
1814 file of library unit @code{Ada.Characters.Latin_1} in file
1815 @file{a-chlat1.ads}.
1816 You may use any of these extended characters freely in character or
1817 string literals. In addition, the extended characters that represent
1818 letters can be used in identifiers.
1820 @node Other 8-Bit Codes
1821 @subsection Other 8-Bit Codes
1824 GNAT also supports several other 8-bit coding schemes:
1827 @item ISO 8859-2 (Latin-2)
1830 Latin-2 letters allowed in identifiers, with uppercase and lowercase
1833 @item ISO 8859-3 (Latin-3)
1836 Latin-3 letters allowed in identifiers, with uppercase and lowercase
1839 @item ISO 8859-4 (Latin-4)
1842 Latin-4 letters allowed in identifiers, with uppercase and lowercase
1845 @item ISO 8859-5 (Cyrillic)
1848 ISO 8859-5 letters (Cyrillic) allowed in identifiers, with uppercase and
1849 lowercase equivalence.
1851 @item ISO 8859-15 (Latin-9)
1854 ISO 8859-15 (Latin-9) letters allowed in identifiers, with uppercase and
1855 lowercase equivalence
1857 @item IBM PC (code page 437)
1858 @cindex code page 437
1859 This code page is the normal default for PCs in the U.S. It corresponds
1860 to the original IBM PC character set. This set has some, but not all, of
1861 the extended Latin-1 letters, but these letters do not have the same
1862 encoding as Latin-1. In this mode, these letters are allowed in
1863 identifiers with uppercase and lowercase equivalence.
1865 @item IBM PC (code page 850)
1866 @cindex code page 850
1867 This code page is a modification of 437 extended to include all the
1868 Latin-1 letters, but still not with the usual Latin-1 encoding. In this
1869 mode, all these letters are allowed in identifiers with uppercase and
1870 lowercase equivalence.
1872 @item Full Upper 8-bit
1873 Any character in the range 80-FF allowed in identifiers, and all are
1874 considered distinct. In other words, there are no uppercase and lowercase
1875 equivalences in this range. This is useful in conjunction with
1876 certain encoding schemes used for some foreign character sets (e.g.,
1877 the typical method of representing Chinese characters on the PC).
1880 No upper-half characters in the range 80-FF are allowed in identifiers.
1881 This gives Ada 83 compatibility for identifier names.
1885 For precise data on the encodings permitted, and the uppercase and lowercase
1886 equivalences that are recognized, see the file @file{csets.adb} in
1887 the GNAT compiler sources. You will need to obtain a full source release
1888 of GNAT to obtain this file.
1890 @node Wide Character Encodings
1891 @subsection Wide Character Encodings
1894 GNAT allows wide character codes to appear in character and string
1895 literals, and also optionally in identifiers, by means of the following
1896 possible encoding schemes:
1901 In this encoding, a wide character is represented by the following five
1909 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1910 characters (using uppercase letters) of the wide character code. For
1911 example, ESC A345 is used to represent the wide character with code
1913 This scheme is compatible with use of the full Wide_Character set.
1915 @item Upper-Half Coding
1916 @cindex Upper-Half Coding
1917 The wide character with encoding @code{16#abcd#} where the upper bit is on
1918 (in other words, ``a'' is in the range 8-F) is represented as two bytes,
1919 @code{16#ab#} and @code{16#cd#}. The second byte cannot be a format control
1920 character, but is not required to be in the upper half. This method can
1921 be also used for shift-JIS or EUC, where the internal coding matches the
1924 @item Shift JIS Coding
1925 @cindex Shift JIS Coding
1926 A wide character is represented by a two-character sequence,
1928 @code{16#cd#}, with the restrictions described for upper-half encoding as
1929 described above. The internal character code is the corresponding JIS
1930 character according to the standard algorithm for Shift-JIS
1931 conversion. Only characters defined in the JIS code set table can be
1932 used with this encoding method.
1936 A wide character is represented by a two-character sequence
1938 @code{16#cd#}, with both characters being in the upper half. The internal
1939 character code is the corresponding JIS character according to the EUC
1940 encoding algorithm. Only characters defined in the JIS code set table
1941 can be used with this encoding method.
1944 A wide character is represented using
1945 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
1946 10646-1/Am.2. Depending on the character value, the representation
1947 is a one, two, or three byte sequence:
1952 16#0000#-16#007f#: 2#0@var{xxxxxxx}#
1953 16#0080#-16#07ff#: 2#110@var{xxxxx}# 2#10@var{xxxxxx}#
1954 16#0800#-16#ffff#: 2#1110@var{xxxx}# 2#10@var{xxxxxx}# 2#10@var{xxxxxx}#
1959 where the @var{xxx} bits correspond to the left-padded bits of the
1960 16-bit character value. Note that all lower half ASCII characters
1961 are represented as ASCII bytes and all upper half characters and
1962 other wide characters are represented as sequences of upper-half
1963 (The full UTF-8 scheme allows for encoding 31-bit characters as
1964 6-byte sequences, but in this implementation, all UTF-8 sequences
1965 of four or more bytes length will be treated as illegal).
1966 @item Brackets Coding
1967 In this encoding, a wide character is represented by the following eight
1975 Where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1976 characters (using uppercase letters) of the wide character code. For
1977 example, [``A345''] is used to represent the wide character with code
1978 @code{16#A345#}. It is also possible (though not required) to use the
1979 Brackets coding for upper half characters. For example, the code
1980 @code{16#A3#} can be represented as @code{[``A3'']}.
1982 This scheme is compatible with use of the full Wide_Character set,
1983 and is also the method used for wide character encoding in the standard
1984 ACVC (Ada Compiler Validation Capability) test suite distributions.
1989 Note: Some of these coding schemes do not permit the full use of the
1990 Ada character set. For example, neither Shift JIS, nor EUC allow the
1991 use of the upper half of the Latin-1 set.
1993 @node File Naming Rules
1994 @section File Naming Rules
1997 The default file name is determined by the name of the unit that the
1998 file contains. The name is formed by taking the full expanded name of
1999 the unit and replacing the separating dots with hyphens and using
2000 ^lowercase^uppercase^ for all letters.
2002 An exception arises if the file name generated by the above rules starts
2003 with one of the characters
2005 @samp{A}, @samp{G}, @samp{I}, or @samp{S},
2008 @samp{a}, @samp{g}, @samp{i}, or @samp{s},
2010 and the second character is a
2011 minus. In this case, the character ^tilde^dollar sign^ is used in place
2012 of the minus. The reason for this special rule is to avoid clashes with
2013 the standard names for child units of the packages System, Ada,
2014 Interfaces, and GNAT, which use the prefixes
2016 @samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},
2019 @samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},
2023 The file extension is @file{.ads} for a spec and
2024 @file{.adb} for a body. The following list shows some
2025 examples of these rules.
2032 @item arith_functions.ads
2033 Arith_Functions (package spec)
2034 @item arith_functions.adb
2035 Arith_Functions (package body)
2037 Func.Spec (child package spec)
2039 Func.Spec (child package body)
2041 Sub (subunit of Main)
2042 @item ^a~bad.adb^A$BAD.ADB^
2043 A.Bad (child package body)
2047 Following these rules can result in excessively long
2048 file names if corresponding
2049 unit names are long (for example, if child units or subunits are
2050 heavily nested). An option is available to shorten such long file names
2051 (called file name ``krunching''). This may be particularly useful when
2052 programs being developed with GNAT are to be used on operating systems
2053 with limited file name lengths. @xref{Using gnatkr}.
2055 Of course, no file shortening algorithm can guarantee uniqueness over
2056 all possible unit names; if file name krunching is used, it is your
2057 responsibility to ensure no name clashes occur. Alternatively you
2058 can specify the exact file names that you want used, as described
2059 in the next section. Finally, if your Ada programs are migrating from a
2060 compiler with a different naming convention, you can use the gnatchop
2061 utility to produce source files that follow the GNAT naming conventions.
2062 (For details @pxref{Renaming Files Using gnatchop}.)
2064 Note: in the case of @code{Windows NT/XP} or @code{OpenVMS} operating
2065 systems, case is not significant. So for example on @code{Windows XP}
2066 if the canonical name is @code{main-sub.adb}, you can use the file name
2067 @code{Main-Sub.adb} instead. However, case is significant for other
2068 operating systems, so for example, if you want to use other than
2069 canonically cased file names on a Unix system, you need to follow
2070 the procedures described in the next section.
2072 @node Using Other File Names
2073 @section Using Other File Names
2077 In the previous section, we have described the default rules used by
2078 GNAT to determine the file name in which a given unit resides. It is
2079 often convenient to follow these default rules, and if you follow them,
2080 the compiler knows without being explicitly told where to find all
2083 However, in some cases, particularly when a program is imported from
2084 another Ada compiler environment, it may be more convenient for the
2085 programmer to specify which file names contain which units. GNAT allows
2086 arbitrary file names to be used by means of the Source_File_Name pragma.
2087 The form of this pragma is as shown in the following examples:
2088 @cindex Source_File_Name pragma
2090 @smallexample @c ada
2092 pragma Source_File_Name (My_Utilities.Stacks,
2093 Spec_File_Name => "myutilst_a.ada");
2094 pragma Source_File_name (My_Utilities.Stacks,
2095 Body_File_Name => "myutilst.ada");
2100 As shown in this example, the first argument for the pragma is the unit
2101 name (in this example a child unit). The second argument has the form
2102 of a named association. The identifier
2103 indicates whether the file name is for a spec or a body;
2104 the file name itself is given by a string literal.
2106 The source file name pragma is a configuration pragma, which means that
2107 normally it will be placed in the @file{gnat.adc}
2108 file used to hold configuration
2109 pragmas that apply to a complete compilation environment.
2110 For more details on how the @file{gnat.adc} file is created and used
2111 see @ref{Handling of Configuration Pragmas}.
2112 @cindex @file{gnat.adc}
2115 GNAT allows completely arbitrary file names to be specified using the
2116 source file name pragma. However, if the file name specified has an
2117 extension other than @file{.ads} or @file{.adb} it is necessary to use
2118 a special syntax when compiling the file. The name in this case must be
2119 preceded by the special sequence @option{-x} followed by a space and the name
2120 of the language, here @code{ada}, as in:
2123 $ gcc -c -x ada peculiar_file_name.sim
2128 @command{gnatmake} handles non-standard file names in the usual manner (the
2129 non-standard file name for the main program is simply used as the
2130 argument to gnatmake). Note that if the extension is also non-standard,
2131 then it must be included in the @command{gnatmake} command, it may not
2134 @node Alternative File Naming Schemes
2135 @section Alternative File Naming Schemes
2136 @cindex File naming schemes, alternative
2139 In the previous section, we described the use of the @code{Source_File_Name}
2140 pragma to allow arbitrary names to be assigned to individual source files.
2141 However, this approach requires one pragma for each file, and especially in
2142 large systems can result in very long @file{gnat.adc} files, and also create
2143 a maintenance problem.
2145 GNAT also provides a facility for specifying systematic file naming schemes
2146 other than the standard default naming scheme previously described. An
2147 alternative scheme for naming is specified by the use of
2148 @code{Source_File_Name} pragmas having the following format:
2149 @cindex Source_File_Name pragma
2151 @smallexample @c ada
2152 pragma Source_File_Name (
2153 Spec_File_Name => FILE_NAME_PATTERN
2154 @r{[},Casing => CASING_SPEC@r{]}
2155 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2157 pragma Source_File_Name (
2158 Body_File_Name => FILE_NAME_PATTERN
2159 @r{[},Casing => CASING_SPEC@r{]}
2160 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2162 pragma Source_File_Name (
2163 Subunit_File_Name => FILE_NAME_PATTERN
2164 @r{[},Casing => CASING_SPEC@r{]}
2165 @r{[},Dot_Replacement => STRING_LITERAL@r{]});
2167 FILE_NAME_PATTERN ::= STRING_LITERAL
2168 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
2172 The @code{FILE_NAME_PATTERN} string shows how the file name is constructed.
2173 It contains a single asterisk character, and the unit name is substituted
2174 systematically for this asterisk. The optional parameter
2175 @code{Casing} indicates
2176 whether the unit name is to be all upper-case letters, all lower-case letters,
2177 or mixed-case. If no
2178 @code{Casing} parameter is used, then the default is all
2179 ^lower-case^upper-case^.
2181 The optional @code{Dot_Replacement} string is used to replace any periods
2182 that occur in subunit or child unit names. If no @code{Dot_Replacement}
2183 argument is used then separating dots appear unchanged in the resulting
2185 Although the above syntax indicates that the
2186 @code{Casing} argument must appear
2187 before the @code{Dot_Replacement} argument, but it
2188 is also permissible to write these arguments in the opposite order.
2190 As indicated, it is possible to specify different naming schemes for
2191 bodies, specs, and subunits. Quite often the rule for subunits is the
2192 same as the rule for bodies, in which case, there is no need to give
2193 a separate @code{Subunit_File_Name} rule, and in this case the
2194 @code{Body_File_name} rule is used for subunits as well.
2196 The separate rule for subunits can also be used to implement the rather
2197 unusual case of a compilation environment (e.g.@: a single directory) which
2198 contains a subunit and a child unit with the same unit name. Although
2199 both units cannot appear in the same partition, the Ada Reference Manual
2200 allows (but does not require) the possibility of the two units coexisting
2201 in the same environment.
2203 The file name translation works in the following steps:
2208 If there is a specific @code{Source_File_Name} pragma for the given unit,
2209 then this is always used, and any general pattern rules are ignored.
2212 If there is a pattern type @code{Source_File_Name} pragma that applies to
2213 the unit, then the resulting file name will be used if the file exists. If
2214 more than one pattern matches, the latest one will be tried first, and the
2215 first attempt resulting in a reference to a file that exists will be used.
2218 If no pattern type @code{Source_File_Name} pragma that applies to the unit
2219 for which the corresponding file exists, then the standard GNAT default
2220 naming rules are used.
2225 As an example of the use of this mechanism, consider a commonly used scheme
2226 in which file names are all lower case, with separating periods copied
2227 unchanged to the resulting file name, and specs end with @file{.1.ada}, and
2228 bodies end with @file{.2.ada}. GNAT will follow this scheme if the following
2231 @smallexample @c ada
2232 pragma Source_File_Name
2233 (Spec_File_Name => "*.1.ada");
2234 pragma Source_File_Name
2235 (Body_File_Name => "*.2.ada");
2239 The default GNAT scheme is actually implemented by providing the following
2240 default pragmas internally:
2242 @smallexample @c ada
2243 pragma Source_File_Name
2244 (Spec_File_Name => "*.ads", Dot_Replacement => "-");
2245 pragma Source_File_Name
2246 (Body_File_Name => "*.adb", Dot_Replacement => "-");
2250 Our final example implements a scheme typically used with one of the
2251 Ada 83 compilers, where the separator character for subunits was ``__''
2252 (two underscores), specs were identified by adding @file{_.ADA}, bodies
2253 by adding @file{.ADA}, and subunits by
2254 adding @file{.SEP}. All file names were
2255 upper case. Child units were not present of course since this was an
2256 Ada 83 compiler, but it seems reasonable to extend this scheme to use
2257 the same double underscore separator for child units.
2259 @smallexample @c ada
2260 pragma Source_File_Name
2261 (Spec_File_Name => "*_.ADA",
2262 Dot_Replacement => "__",
2263 Casing = Uppercase);
2264 pragma Source_File_Name
2265 (Body_File_Name => "*.ADA",
2266 Dot_Replacement => "__",
2267 Casing = Uppercase);
2268 pragma Source_File_Name
2269 (Subunit_File_Name => "*.SEP",
2270 Dot_Replacement => "__",
2271 Casing = Uppercase);
2274 @node Generating Object Files
2275 @section Generating Object Files
2278 An Ada program consists of a set of source files, and the first step in
2279 compiling the program is to generate the corresponding object files.
2280 These are generated by compiling a subset of these source files.
2281 The files you need to compile are the following:
2285 If a package spec has no body, compile the package spec to produce the
2286 object file for the package.
2289 If a package has both a spec and a body, compile the body to produce the
2290 object file for the package. The source file for the package spec need
2291 not be compiled in this case because there is only one object file, which
2292 contains the code for both the spec and body of the package.
2295 For a subprogram, compile the subprogram body to produce the object file
2296 for the subprogram. The spec, if one is present, is as usual in a
2297 separate file, and need not be compiled.
2301 In the case of subunits, only compile the parent unit. A single object
2302 file is generated for the entire subunit tree, which includes all the
2306 Compile child units independently of their parent units
2307 (though, of course, the spec of all the ancestor unit must be present in order
2308 to compile a child unit).
2312 Compile generic units in the same manner as any other units. The object
2313 files in this case are small dummy files that contain at most the
2314 flag used for elaboration checking. This is because GNAT always handles generic
2315 instantiation by means of macro expansion. However, it is still necessary to
2316 compile generic units, for dependency checking and elaboration purposes.
2320 The preceding rules describe the set of files that must be compiled to
2321 generate the object files for a program. Each object file has the same
2322 name as the corresponding source file, except that the extension is
2325 You may wish to compile other files for the purpose of checking their
2326 syntactic and semantic correctness. For example, in the case where a
2327 package has a separate spec and body, you would not normally compile the
2328 spec. However, it is convenient in practice to compile the spec to make
2329 sure it is error-free before compiling clients of this spec, because such
2330 compilations will fail if there is an error in the spec.
2332 GNAT provides an option for compiling such files purely for the
2333 purposes of checking correctness; such compilations are not required as
2334 part of the process of building a program. To compile a file in this
2335 checking mode, use the @option{-gnatc} switch.
2337 @node Source Dependencies
2338 @section Source Dependencies
2341 A given object file clearly depends on the source file which is compiled
2342 to produce it. Here we are using @dfn{depends} in the sense of a typical
2343 @code{make} utility; in other words, an object file depends on a source
2344 file if changes to the source file require the object file to be
2346 In addition to this basic dependency, a given object may depend on
2347 additional source files as follows:
2351 If a file being compiled @code{with}'s a unit @var{X}, the object file
2352 depends on the file containing the spec of unit @var{X}. This includes
2353 files that are @code{with}'ed implicitly either because they are parents
2354 of @code{with}'ed child units or they are run-time units required by the
2355 language constructs used in a particular unit.
2358 If a file being compiled instantiates a library level generic unit, the
2359 object file depends on both the spec and body files for this generic
2363 If a file being compiled instantiates a generic unit defined within a
2364 package, the object file depends on the body file for the package as
2365 well as the spec file.
2369 @cindex @option{-gnatn} switch
2370 If a file being compiled contains a call to a subprogram for which
2371 pragma @code{Inline} applies and inlining is activated with the
2372 @option{-gnatn} switch, the object file depends on the file containing the
2373 body of this subprogram as well as on the file containing the spec. Note
2374 that for inlining to actually occur as a result of the use of this switch,
2375 it is necessary to compile in optimizing mode.
2377 @cindex @option{-gnatN} switch
2378 The use of @option{-gnatN} activates inlining optimization
2379 that is performed by the front end of the compiler. This inlining does
2380 not require that the code generation be optimized. Like @option{-gnatn},
2381 the use of this switch generates additional dependencies.
2383 When using a gcc-based back end (in practice this means using any version
2384 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
2385 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
2386 Historically front end inlining was more extensive than the gcc back end
2387 inlining, but that is no longer the case.
2390 If an object file @file{O} depends on the proper body of a subunit through
2391 inlining or instantiation, it depends on the parent unit of the subunit.
2392 This means that any modification of the parent unit or one of its subunits
2393 affects the compilation of @file{O}.
2396 The object file for a parent unit depends on all its subunit body files.
2399 The previous two rules meant that for purposes of computing dependencies and
2400 recompilation, a body and all its subunits are treated as an indivisible whole.
2403 These rules are applied transitively: if unit @code{A} @code{with}'s
2404 unit @code{B}, whose elaboration calls an inlined procedure in package
2405 @code{C}, the object file for unit @code{A} will depend on the body of
2406 @code{C}, in file @file{c.adb}.
2408 The set of dependent files described by these rules includes all the
2409 files on which the unit is semantically dependent, as dictated by the
2410 Ada language standard. However, it is a superset of what the
2411 standard describes, because it includes generic, inline, and subunit
2414 An object file must be recreated by recompiling the corresponding source
2415 file if any of the source files on which it depends are modified. For
2416 example, if the @code{make} utility is used to control compilation,
2417 the rule for an Ada object file must mention all the source files on
2418 which the object file depends, according to the above definition.
2419 The determination of the necessary
2420 recompilations is done automatically when one uses @command{gnatmake}.
2423 @node The Ada Library Information Files
2424 @section The Ada Library Information Files
2425 @cindex Ada Library Information files
2426 @cindex @file{ALI} files
2429 Each compilation actually generates two output files. The first of these
2430 is the normal object file that has a @file{.o} extension. The second is a
2431 text file containing full dependency information. It has the same
2432 name as the source file, but an @file{.ali} extension.
2433 This file is known as the Ada Library Information (@file{ALI}) file.
2434 The following information is contained in the @file{ALI} file.
2438 Version information (indicates which version of GNAT was used to compile
2439 the unit(s) in question)
2442 Main program information (including priority and time slice settings,
2443 as well as the wide character encoding used during compilation).
2446 List of arguments used in the @command{gcc} command for the compilation
2449 Attributes of the unit, including configuration pragmas used, an indication
2450 of whether the compilation was successful, exception model used etc.
2453 A list of relevant restrictions applying to the unit (used for consistency)
2457 Categorization information (e.g.@: use of pragma @code{Pure}).
2460 Information on all @code{with}'ed units, including presence of
2461 @code{Elaborate} or @code{Elaborate_All} pragmas.
2464 Information from any @code{Linker_Options} pragmas used in the unit
2467 Information on the use of @code{Body_Version} or @code{Version}
2468 attributes in the unit.
2471 Dependency information. This is a list of files, together with
2472 time stamp and checksum information. These are files on which
2473 the unit depends in the sense that recompilation is required
2474 if any of these units are modified.
2477 Cross-reference data. Contains information on all entities referenced
2478 in the unit. Used by tools like @code{gnatxref} and @code{gnatfind} to
2479 provide cross-reference information.
2484 For a full detailed description of the format of the @file{ALI} file,
2485 see the source of the body of unit @code{Lib.Writ}, contained in file
2486 @file{lib-writ.adb} in the GNAT compiler sources.
2488 @node Binding an Ada Program
2489 @section Binding an Ada Program
2492 When using languages such as C and C++, once the source files have been
2493 compiled the only remaining step in building an executable program
2494 is linking the object modules together. This means that it is possible to
2495 link an inconsistent version of a program, in which two units have
2496 included different versions of the same header.
2498 The rules of Ada do not permit such an inconsistent program to be built.
2499 For example, if two clients have different versions of the same package,
2500 it is illegal to build a program containing these two clients.
2501 These rules are enforced by the GNAT binder, which also determines an
2502 elaboration order consistent with the Ada rules.
2504 The GNAT binder is run after all the object files for a program have
2505 been created. It is given the name of the main program unit, and from
2506 this it determines the set of units required by the program, by reading the
2507 corresponding ALI files. It generates error messages if the program is
2508 inconsistent or if no valid order of elaboration exists.
2510 If no errors are detected, the binder produces a main program, in Ada by
2511 default, that contains calls to the elaboration procedures of those
2512 compilation unit that require them, followed by
2513 a call to the main program. This Ada program is compiled to generate the
2514 object file for the main program. The name of
2515 the Ada file is @file{b~@var{xxx}.adb} (with the corresponding spec
2516 @file{b~@var{xxx}.ads}) where @var{xxx} is the name of the
2519 Finally, the linker is used to build the resulting executable program,
2520 using the object from the main program from the bind step as well as the
2521 object files for the Ada units of the program.
2523 @node Mixed Language Programming
2524 @section Mixed Language Programming
2525 @cindex Mixed Language Programming
2528 This section describes how to develop a mixed-language program,
2529 specifically one that comprises units in both Ada and C.
2532 * Interfacing to C::
2533 * Calling Conventions::
2536 @node Interfacing to C
2537 @subsection Interfacing to C
2539 Interfacing Ada with a foreign language such as C involves using
2540 compiler directives to import and/or export entity definitions in each
2541 language---using @code{extern} statements in C, for instance, and the
2542 @code{Import}, @code{Export}, and @code{Convention} pragmas in Ada.
2543 A full treatment of these topics is provided in Appendix B, section 1
2544 of the Ada Reference Manual.
2546 There are two ways to build a program using GNAT that contains some Ada
2547 sources and some foreign language sources, depending on whether or not
2548 the main subprogram is written in Ada. Here is a source example with
2549 the main subprogram in Ada:
2555 void print_num (int num)
2557 printf ("num is %d.\n", num);
2563 /* num_from_Ada is declared in my_main.adb */
2564 extern int num_from_Ada;
2568 return num_from_Ada;
2572 @smallexample @c ada
2574 procedure My_Main is
2576 -- Declare then export an Integer entity called num_from_Ada
2577 My_Num : Integer := 10;
2578 pragma Export (C, My_Num, "num_from_Ada");
2580 -- Declare an Ada function spec for Get_Num, then use
2581 -- C function get_num for the implementation.
2582 function Get_Num return Integer;
2583 pragma Import (C, Get_Num, "get_num");
2585 -- Declare an Ada procedure spec for Print_Num, then use
2586 -- C function print_num for the implementation.
2587 procedure Print_Num (Num : Integer);
2588 pragma Import (C, Print_Num, "print_num");
2591 Print_Num (Get_Num);
2597 To build this example, first compile the foreign language files to
2598 generate object files:
2600 ^gcc -c file1.c^gcc -c FILE1.C^
2601 ^gcc -c file2.c^gcc -c FILE2.C^
2605 Then, compile the Ada units to produce a set of object files and ALI
2608 gnatmake ^-c^/ACTIONS=COMPILE^ my_main.adb
2612 Run the Ada binder on the Ada main program:
2614 gnatbind my_main.ali
2618 Link the Ada main program, the Ada objects and the other language
2621 gnatlink my_main.ali file1.o file2.o
2625 The last three steps can be grouped in a single command:
2627 gnatmake my_main.adb -largs file1.o file2.o
2630 @cindex Binder output file
2632 If the main program is in a language other than Ada, then you may have
2633 more than one entry point into the Ada subsystem. You must use a special
2634 binder option to generate callable routines that initialize and
2635 finalize the Ada units (@pxref{Binding with Non-Ada Main Programs}).
2636 Calls to the initialization and finalization routines must be inserted
2637 in the main program, or some other appropriate point in the code. The
2638 call to initialize the Ada units must occur before the first Ada
2639 subprogram is called, and the call to finalize the Ada units must occur
2640 after the last Ada subprogram returns. The binder will place the
2641 initialization and finalization subprograms into the
2642 @file{b~@var{xxx}.adb} file where they can be accessed by your C
2643 sources. To illustrate, we have the following example:
2647 extern void adainit (void);
2648 extern void adafinal (void);
2649 extern int add (int, int);
2650 extern int sub (int, int);
2652 int main (int argc, char *argv[])
2658 /* Should print "21 + 7 = 28" */
2659 printf ("%d + %d = %d\n", a, b, add (a, b));
2660 /* Should print "21 - 7 = 14" */
2661 printf ("%d - %d = %d\n", a, b, sub (a, b));
2667 @smallexample @c ada
2670 function Add (A, B : Integer) return Integer;
2671 pragma Export (C, Add, "add");
2675 package body Unit1 is
2676 function Add (A, B : Integer) return Integer is
2684 function Sub (A, B : Integer) return Integer;
2685 pragma Export (C, Sub, "sub");
2689 package body Unit2 is
2690 function Sub (A, B : Integer) return Integer is
2699 The build procedure for this application is similar to the last
2700 example's. First, compile the foreign language files to generate object
2703 ^gcc -c main.c^gcc -c main.c^
2707 Next, compile the Ada units to produce a set of object files and ALI
2710 gnatmake ^-c^/ACTIONS=COMPILE^ unit1.adb
2711 gnatmake ^-c^/ACTIONS=COMPILE^ unit2.adb
2715 Run the Ada binder on every generated ALI file. Make sure to use the
2716 @option{-n} option to specify a foreign main program:
2718 gnatbind ^-n^/NOMAIN^ unit1.ali unit2.ali
2722 Link the Ada main program, the Ada objects and the foreign language
2723 objects. You need only list the last ALI file here:
2725 gnatlink unit2.ali main.o -o exec_file
2728 This procedure yields a binary executable called @file{exec_file}.
2732 Depending on the circumstances (for example when your non-Ada main object
2733 does not provide symbol @code{main}), you may also need to instruct the
2734 GNAT linker not to include the standard startup objects by passing the
2735 @option{^-nostartfiles^/NOSTART_FILES^} switch to @command{gnatlink}.
2737 @node Calling Conventions
2738 @subsection Calling Conventions
2739 @cindex Foreign Languages
2740 @cindex Calling Conventions
2741 GNAT follows standard calling sequence conventions and will thus interface
2742 to any other language that also follows these conventions. The following
2743 Convention identifiers are recognized by GNAT:
2746 @cindex Interfacing to Ada
2747 @cindex Other Ada compilers
2748 @cindex Convention Ada
2750 This indicates that the standard Ada calling sequence will be
2751 used and all Ada data items may be passed without any limitations in the
2752 case where GNAT is used to generate both the caller and callee. It is also
2753 possible to mix GNAT generated code and code generated by another Ada
2754 compiler. In this case, the data types should be restricted to simple
2755 cases, including primitive types. Whether complex data types can be passed
2756 depends on the situation. Probably it is safe to pass simple arrays, such
2757 as arrays of integers or floats. Records may or may not work, depending
2758 on whether both compilers lay them out identically. Complex structures
2759 involving variant records, access parameters, tasks, or protected types,
2760 are unlikely to be able to be passed.
2762 Note that in the case of GNAT running
2763 on a platform that supports HP Ada 83, a higher degree of compatibility
2764 can be guaranteed, and in particular records are layed out in an identical
2765 manner in the two compilers. Note also that if output from two different
2766 compilers is mixed, the program is responsible for dealing with elaboration
2767 issues. Probably the safest approach is to write the main program in the
2768 version of Ada other than GNAT, so that it takes care of its own elaboration
2769 requirements, and then call the GNAT-generated adainit procedure to ensure
2770 elaboration of the GNAT components. Consult the documentation of the other
2771 Ada compiler for further details on elaboration.
2773 However, it is not possible to mix the tasking run time of GNAT and
2774 HP Ada 83, All the tasking operations must either be entirely within
2775 GNAT compiled sections of the program, or entirely within HP Ada 83
2776 compiled sections of the program.
2778 @cindex Interfacing to Assembly
2779 @cindex Convention Assembler
2781 Specifies assembler as the convention. In practice this has the
2782 same effect as convention Ada (but is not equivalent in the sense of being
2783 considered the same convention).
2785 @cindex Convention Asm
2788 Equivalent to Assembler.
2790 @cindex Interfacing to COBOL
2791 @cindex Convention COBOL
2794 Data will be passed according to the conventions described
2795 in section B.4 of the Ada Reference Manual.
2798 @cindex Interfacing to C
2799 @cindex Convention C
2801 Data will be passed according to the conventions described
2802 in section B.3 of the Ada Reference Manual.
2804 A note on interfacing to a C ``varargs'' function:
2805 @findex C varargs function
2806 @cindex Interfacing to C varargs function
2807 @cindex varargs function interfaces
2811 In C, @code{varargs} allows a function to take a variable number of
2812 arguments. There is no direct equivalent in this to Ada. One
2813 approach that can be used is to create a C wrapper for each
2814 different profile and then interface to this C wrapper. For
2815 example, to print an @code{int} value using @code{printf},
2816 create a C function @code{printfi} that takes two arguments, a
2817 pointer to a string and an int, and calls @code{printf}.
2818 Then in the Ada program, use pragma @code{Import} to
2819 interface to @code{printfi}.
2822 It may work on some platforms to directly interface to
2823 a @code{varargs} function by providing a specific Ada profile
2824 for a particular call. However, this does not work on
2825 all platforms, since there is no guarantee that the
2826 calling sequence for a two argument normal C function
2827 is the same as for calling a @code{varargs} C function with
2828 the same two arguments.
2831 @cindex Convention Default
2836 @cindex Convention External
2843 @cindex Interfacing to C++
2844 @cindex Convention C++
2845 @item C_Plus_Plus (or CPP)
2846 This stands for C++. For most purposes this is identical to C.
2847 See the separate description of the specialized GNAT pragmas relating to
2848 C++ interfacing for further details.
2852 @cindex Interfacing to Fortran
2853 @cindex Convention Fortran
2855 Data will be passed according to the conventions described
2856 in section B.5 of the Ada Reference Manual.
2859 This applies to an intrinsic operation, as defined in the Ada
2860 Reference Manual. If a pragma Import (Intrinsic) applies to a subprogram,
2861 this means that the body of the subprogram is provided by the compiler itself,
2862 usually by means of an efficient code sequence, and that the user does not
2863 supply an explicit body for it. In an application program, the pragma may
2864 be applied to the following sets of names:
2868 Rotate_Left, Rotate_Right, Shift_Left, Shift_Right,
2869 Shift_Right_Arithmetic. The corresponding subprogram declaration must have
2870 two formal parameters. The
2871 first one must be a signed integer type or a modular type with a binary
2872 modulus, and the second parameter must be of type Natural.
2873 The return type must be the same as the type of the first argument. The size
2874 of this type can only be 8, 16, 32, or 64.
2877 Binary arithmetic operators: ``+'', ``-'', ``*'', ``/''
2878 The corresponding operator declaration must have parameters and result type
2879 that have the same root numeric type (for example, all three are long_float
2880 types). This simplifies the definition of operations that use type checking
2881 to perform dimensional checks:
2883 @smallexample @c ada
2884 type Distance is new Long_Float;
2885 type Time is new Long_Float;
2886 type Velocity is new Long_Float;
2887 function "/" (D : Distance; T : Time)
2889 pragma Import (Intrinsic, "/");
2893 This common idiom is often programmed with a generic definition and an
2894 explicit body. The pragma makes it simpler to introduce such declarations.
2895 It incurs no overhead in compilation time or code size, because it is
2896 implemented as a single machine instruction.
2899 General subprogram entities, to bind an Ada subprogram declaration to
2900 a compiler builtin by name with back-ends where such interfaces are
2901 available. A typical example is the set of ``__builtin'' functions
2902 exposed by the GCC back-end, as in the following example:
2904 @smallexample @c ada
2905 function builtin_sqrt (F : Float) return Float;
2906 pragma Import (Intrinsic, builtin_sqrt, "__builtin_sqrtf");
2909 Most of the GCC builtins are accessible this way, and as for other
2910 import conventions (e.g. C), it is the user's responsibility to ensure
2911 that the Ada subprogram profile matches the underlying builtin
2919 @cindex Convention Stdcall
2921 This is relevant only to Windows XP/2000/NT implementations of GNAT,
2922 and specifies that the @code{Stdcall} calling sequence will be used,
2923 as defined by the NT API. Nevertheless, to ease building
2924 cross-platform bindings this convention will be handled as a @code{C} calling
2925 convention on non-Windows platforms.
2928 @cindex Convention DLL
2930 This is equivalent to @code{Stdcall}.
2933 @cindex Convention Win32
2935 This is equivalent to @code{Stdcall}.
2939 @cindex Convention Stubbed
2941 This is a special convention that indicates that the compiler
2942 should provide a stub body that raises @code{Program_Error}.
2946 GNAT additionally provides a useful pragma @code{Convention_Identifier}
2947 that can be used to parametrize conventions and allow additional synonyms
2948 to be specified. For example if you have legacy code in which the convention
2949 identifier Fortran77 was used for Fortran, you can use the configuration
2952 @smallexample @c ada
2953 pragma Convention_Identifier (Fortran77, Fortran);
2957 And from now on the identifier Fortran77 may be used as a convention
2958 identifier (for example in an @code{Import} pragma) with the same
2962 @node Building Mixed Ada & C++ Programs
2963 @section Building Mixed Ada and C++ Programs
2966 A programmer inexperienced with mixed-language development may find that
2967 building an application containing both Ada and C++ code can be a
2968 challenge. This section gives a few
2969 hints that should make this task easier. The first section addresses
2970 the differences between interfacing with C and interfacing with C++.
2972 looks into the delicate problem of linking the complete application from
2973 its Ada and C++ parts. The last section gives some hints on how the GNAT
2974 run-time library can be adapted in order to allow inter-language dispatching
2975 with a new C++ compiler.
2978 * Interfacing to C++::
2979 * Linking a Mixed C++ & Ada Program::
2980 * A Simple Example::
2981 * Interfacing with C++ constructors::
2982 * Interfacing with C++ at the Class Level::
2985 @node Interfacing to C++
2986 @subsection Interfacing to C++
2989 GNAT supports interfacing with the G++ compiler (or any C++ compiler
2990 generating code that is compatible with the G++ Application Binary
2991 Interface ---see http://www.codesourcery.com/archives/cxx-abi).
2994 Interfacing can be done at 3 levels: simple data, subprograms, and
2995 classes. In the first two cases, GNAT offers a specific @code{Convention
2996 C_Plus_Plus} (or @code{CPP}) that behaves exactly like @code{Convention C}.
2997 Usually, C++ mangles the names of subprograms. To generate proper mangled
2998 names automatically, see @ref{Generating Ada Bindings for C and C++ headers}).
2999 This problem can also be addressed manually in two ways:
3003 by modifying the C++ code in order to force a C convention using
3004 the @code{extern "C"} syntax.
3007 by figuring out the mangled name (using e.g. @command{nm}) and using it as the
3008 Link_Name argument of the pragma import.
3012 Interfacing at the class level can be achieved by using the GNAT specific
3013 pragmas such as @code{CPP_Constructor}. @xref{Interfacing to C++,,,
3014 gnat_rm, GNAT Reference Manual}, for additional information.
3016 @node Linking a Mixed C++ & Ada Program
3017 @subsection Linking a Mixed C++ & Ada Program
3020 Usually the linker of the C++ development system must be used to link
3021 mixed applications because most C++ systems will resolve elaboration
3022 issues (such as calling constructors on global class instances)
3023 transparently during the link phase. GNAT has been adapted to ease the
3024 use of a foreign linker for the last phase. Three cases can be
3029 Using GNAT and G++ (GNU C++ compiler) from the same GCC installation:
3030 The C++ linker can simply be called by using the C++ specific driver
3033 Note that if the C++ code uses inline functions, you will need to
3034 compile your C++ code with the @code{-fkeep-inline-functions} switch in
3035 order to provide an existing function implementation that the Ada code can
3039 $ g++ -c -fkeep-inline-functions file1.C
3040 $ g++ -c -fkeep-inline-functions file2.C
3041 $ gnatmake ada_unit -largs file1.o file2.o --LINK=g++
3045 Using GNAT and G++ from two different GCC installations: If both
3046 compilers are on the @env{PATH}, the previous method may be used. It is
3047 important to note that environment variables such as
3048 @env{C_INCLUDE_PATH}, @env{GCC_EXEC_PREFIX}, @env{BINUTILS_ROOT}, and
3049 @env{GCC_ROOT} will affect both compilers
3050 at the same time and may make one of the two compilers operate
3051 improperly if set during invocation of the wrong compiler. It is also
3052 very important that the linker uses the proper @file{libgcc.a} GCC
3053 library -- that is, the one from the C++ compiler installation. The
3054 implicit link command as suggested in the @command{gnatmake} command
3055 from the former example can be replaced by an explicit link command with
3056 the full-verbosity option in order to verify which library is used:
3059 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=c++
3061 If there is a problem due to interfering environment variables, it can
3062 be worked around by using an intermediate script. The following example
3063 shows the proper script to use when GNAT has not been installed at its
3064 default location and g++ has been installed at its default location:
3072 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=./my_script
3076 Using a non-GNU C++ compiler: The commands previously described can be
3077 used to insure that the C++ linker is used. Nonetheless, you need to add
3078 a few more parameters to the link command line, depending on the exception
3081 If the @code{setjmp/longjmp} exception mechanism is used, only the paths
3082 to the libgcc libraries are required:
3087 CC $* `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a`
3088 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3091 Where CC is the name of the non-GNU C++ compiler.
3093 If the @code{zero cost} exception mechanism is used, and the platform
3094 supports automatic registration of exception tables (e.g.@: Solaris or IRIX),
3095 paths to more objects are required:
3100 CC `gcc -print-file-name=crtbegin.o` $* \
3101 `gcc -print-file-name=libgcc.a` `gcc -print-file-name=libgcc_eh.a` \
3102 `gcc -print-file-name=crtend.o`
3103 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
3106 If the @code{zero cost} exception mechanism is used, and the platform
3107 doesn't support automatic registration of exception tables (e.g.@: HP-UX,
3108 Tru64 or AIX), the simple approach described above will not work and
3109 a pre-linking phase using GNAT will be necessary.
3113 Another alternative is to use the @command{gprbuild} multi-language builder
3114 which has a large knowledge base and knows how to link Ada and C++ code
3115 together automatically in most cases.
3117 @node A Simple Example
3118 @subsection A Simple Example
3120 The following example, provided as part of the GNAT examples, shows how
3121 to achieve procedural interfacing between Ada and C++ in both
3122 directions. The C++ class A has two methods. The first method is exported
3123 to Ada by the means of an extern C wrapper function. The second method
3124 calls an Ada subprogram. On the Ada side, The C++ calls are modelled by
3125 a limited record with a layout comparable to the C++ class. The Ada
3126 subprogram, in turn, calls the C++ method. So, starting from the C++
3127 main program, the process passes back and forth between the two
3131 Here are the compilation commands:
3133 $ gnatmake -c simple_cpp_interface
3136 $ gnatbind -n simple_cpp_interface
3137 $ gnatlink simple_cpp_interface -o cpp_main --LINK=g++
3138 -lstdc++ ex7.o cpp_main.o
3142 Here are the corresponding sources:
3150 void adainit (void);
3151 void adafinal (void);
3152 void method1 (A *t);
3174 class A : public Origin @{
3176 void method1 (void);
3177 void method2 (int v);
3187 extern "C" @{ void ada_method2 (A *t, int v);@}
3189 void A::method1 (void)
3192 printf ("in A::method1, a_value = %d \n",a_value);
3196 void A::method2 (int v)
3198 ada_method2 (this, v);
3199 printf ("in A::method2, a_value = %d \n",a_value);
3206 printf ("in A::A, a_value = %d \n",a_value);
3210 @smallexample @c ada
3212 package body Simple_Cpp_Interface is
3214 procedure Ada_Method2 (This : in out A; V : Integer) is
3220 end Simple_Cpp_Interface;
3223 package Simple_Cpp_Interface is
3226 Vptr : System.Address;
3230 pragma Convention (C, A);
3232 procedure Method1 (This : in out A);
3233 pragma Import (C, Method1);
3235 procedure Ada_Method2 (This : in out A; V : Integer);
3236 pragma Export (C, Ada_Method2);
3238 end Simple_Cpp_Interface;
3241 @node Interfacing with C++ constructors
3242 @subsection Interfacing with C++ constructors
3245 In order to interface with C++ constructors GNAT provides the
3246 @code{pragma CPP_Constructor} (@xref{Interfacing to C++,,,
3247 gnat_rm, GNAT Reference Manual}, for additional information).
3248 In this section we present some common uses of C++ constructors
3249 in mixed-languages programs in GNAT.
3251 Let us assume that we need to interface with the following
3259 @b{virtual} int Get_Value ();
3260 Root(); // Default constructor
3261 Root(int v); // 1st non-default constructor
3262 Root(int v, int w); // 2nd non-default constructor
3266 For this purpose we can write the following package spec (further
3267 information on how to build this spec is available in
3268 @ref{Interfacing with C++ at the Class Level} and
3269 @ref{Generating Ada Bindings for C and C++ headers}).
3271 @smallexample @c ada
3272 with Interfaces.C; use Interfaces.C;
3274 type Root is tagged limited record
3278 pragma Import (CPP, Root);
3280 function Get_Value (Obj : Root) return int;
3281 pragma Import (CPP, Get_Value);
3283 function Constructor return Root;
3284 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ev");
3286 function Constructor (v : Integer) return Root;
3287 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ei");
3289 function Constructor (v, w : Integer) return Root;
3290 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Eii");
3294 On the Ada side the constructor is represented by a function (whose
3295 name is arbitrary) that returns the classwide type corresponding to
3296 the imported C++ class. Although the constructor is described as a
3297 function, it is typically a procedure with an extra implicit argument
3298 (the object being initialized) at the implementation level. GNAT
3299 issues the appropriate call, whatever it is, to get the object
3300 properly initialized.
3302 Constructors can only appear in the following contexts:
3306 On the right side of an initialization of an object of type @var{T}.
3308 On the right side of an initialization of a record component of type @var{T}.
3310 In an Ada 2005 limited aggregate.
3312 In an Ada 2005 nested limited aggregate.
3314 In an Ada 2005 limited aggregate that initializes an object built in
3315 place by an extended return statement.
3319 In a declaration of an object whose type is a class imported from C++,
3320 either the default C++ constructor is implicitly called by GNAT, or
3321 else the required C++ constructor must be explicitly called in the
3322 expression that initializes the object. For example:
3324 @smallexample @c ada
3326 Obj2 : Root := Constructor;
3327 Obj3 : Root := Constructor (v => 10);
3328 Obj4 : Root := Constructor (30, 40);
3331 The first two declarations are equivalent: in both cases the default C++
3332 constructor is invoked (in the former case the call to the constructor is
3333 implicit, and in the latter case the call is explicit in the object
3334 declaration). @code{Obj3} is initialized by the C++ non-default constructor
3335 that takes an integer argument, and @code{Obj4} is initialized by the
3336 non-default C++ constructor that takes two integers.
3338 Let us derive the imported C++ class in the Ada side. For example:
3340 @smallexample @c ada
3341 type DT is new Root with record
3342 C_Value : Natural := 2009;
3346 In this case the components DT inherited from the C++ side must be
3347 initialized by a C++ constructor, and the additional Ada components
3348 of type DT are initialized by GNAT. The initialization of such an
3349 object is done either by default, or by means of a function returning
3350 an aggregate of type DT, or by means of an extension aggregate.
3352 @smallexample @c ada
3354 Obj6 : DT := Function_Returning_DT (50);
3355 Obj7 : DT := (Constructor (30,40) with C_Value => 50);
3358 The declaration of @code{Obj5} invokes the default constructors: the
3359 C++ default constructor of the parent type takes care of the initialization
3360 of the components inherited from Root, and GNAT takes care of the default
3361 initialization of the additional Ada components of type DT (that is,
3362 @code{C_Value} is initialized to value 2009). The order of invocation of
3363 the constructors is consistent with the order of elaboration required by
3364 Ada and C++. That is, the constructor of the parent type is always called
3365 before the constructor of the derived type.
3367 Let us now consider a record that has components whose type is imported
3368 from C++. For example:
3370 @smallexample @c ada
3371 type Rec1 is limited record
3372 Data1 : Root := Constructor (10);
3373 Value : Natural := 1000;
3376 type Rec2 (D : Integer := 20) is limited record
3378 Data2 : Root := Constructor (D, 30);
3382 The initialization of an object of type @code{Rec2} will call the
3383 non-default C++ constructors specified for the imported components.
3386 @smallexample @c ada
3390 Using Ada 2005 we can use limited aggregates to initialize an object
3391 invoking C++ constructors that differ from those specified in the type
3392 declarations. For example:
3394 @smallexample @c ada
3395 Obj9 : Rec2 := (Rec => (Data1 => Constructor (15, 16),
3400 The above declaration uses an Ada 2005 limited aggregate to
3401 initialize @code{Obj9}, and the C++ constructor that has two integer
3402 arguments is invoked to initialize the @code{Data1} component instead
3403 of the constructor specified in the declaration of type @code{Rec1}. In
3404 Ada 2005 the box in the aggregate indicates that unspecified components
3405 are initialized using the expression (if any) available in the component
3406 declaration. That is, in this case discriminant @code{D} is initialized
3407 to value @code{20}, @code{Value} is initialized to value 1000, and the
3408 non-default C++ constructor that handles two integers takes care of
3409 initializing component @code{Data2} with values @code{20,30}.
3411 In Ada 2005 we can use the extended return statement to build the Ada
3412 equivalent to C++ non-default constructors. For example:
3414 @smallexample @c ada
3415 function Constructor (V : Integer) return Rec2 is
3417 return Obj : Rec2 := (Rec => (Data1 => Constructor (V, 20),
3420 -- Further actions required for construction of
3421 -- objects of type Rec2
3427 In this example the extended return statement construct is used to
3428 build in place the returned object whose components are initialized
3429 by means of a limited aggregate. Any further action associated with
3430 the constructor can be placed inside the construct.
3432 @node Interfacing with C++ at the Class Level
3433 @subsection Interfacing with C++ at the Class Level
3435 In this section we demonstrate the GNAT features for interfacing with
3436 C++ by means of an example making use of Ada 2005 abstract interface
3437 types. This example consists of a classification of animals; classes
3438 have been used to model our main classification of animals, and
3439 interfaces provide support for the management of secondary
3440 classifications. We first demonstrate a case in which the types and
3441 constructors are defined on the C++ side and imported from the Ada
3442 side, and latter the reverse case.
3444 The root of our derivation will be the @code{Animal} class, with a
3445 single private attribute (the @code{Age} of the animal) and two public
3446 primitives to set and get the value of this attribute.
3451 @b{virtual} void Set_Age (int New_Age);
3452 @b{virtual} int Age ();
3458 Abstract interface types are defined in C++ by means of classes with pure
3459 virtual functions and no data members. In our example we will use two
3460 interfaces that provide support for the common management of @code{Carnivore}
3461 and @code{Domestic} animals:
3464 @b{class} Carnivore @{
3466 @b{virtual} int Number_Of_Teeth () = 0;
3469 @b{class} Domestic @{
3471 @b{virtual void} Set_Owner (char* Name) = 0;
3475 Using these declarations, we can now say that a @code{Dog} is an animal that is
3476 both Carnivore and Domestic, that is:
3479 @b{class} Dog : Animal, Carnivore, Domestic @{
3481 @b{virtual} int Number_Of_Teeth ();
3482 @b{virtual} void Set_Owner (char* Name);
3484 Dog(); // Constructor
3491 In the following examples we will assume that the previous declarations are
3492 located in a file named @code{animals.h}. The following package demonstrates
3493 how to import these C++ declarations from the Ada side:
3495 @smallexample @c ada
3496 with Interfaces.C.Strings; use Interfaces.C.Strings;
3498 type Carnivore is interface;
3499 pragma Convention (C_Plus_Plus, Carnivore);
3500 function Number_Of_Teeth (X : Carnivore)
3501 return Natural is abstract;
3503 type Domestic is interface;
3504 pragma Convention (C_Plus_Plus, Set_Owner);
3506 (X : in out Domestic;
3507 Name : Chars_Ptr) is abstract;
3509 type Animal is tagged record
3512 pragma Import (C_Plus_Plus, Animal);
3514 procedure Set_Age (X : in out Animal; Age : Integer);
3515 pragma Import (C_Plus_Plus, Set_Age);
3517 function Age (X : Animal) return Integer;
3518 pragma Import (C_Plus_Plus, Age);
3520 type Dog is new Animal and Carnivore and Domestic with record
3521 Tooth_Count : Natural;
3522 Owner : String (1 .. 30);
3524 pragma Import (C_Plus_Plus, Dog);
3526 function Number_Of_Teeth (A : Dog) return Integer;
3527 pragma Import (C_Plus_Plus, Number_Of_Teeth);
3529 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3530 pragma Import (C_Plus_Plus, Set_Owner);
3532 function New_Dog return Dog;
3533 pragma CPP_Constructor (New_Dog);
3534 pragma Import (CPP, New_Dog, "_ZN3DogC2Ev");
3538 Thanks to the compatibility between GNAT run-time structures and the C++ ABI,
3539 interfacing with these C++ classes is easy. The only requirement is that all
3540 the primitives and components must be declared exactly in the same order in
3543 Regarding the abstract interfaces, we must indicate to the GNAT compiler by
3544 means of a @code{pragma Convention (C_Plus_Plus)}, the convention used to pass
3545 the arguments to the called primitives will be the same as for C++. For the
3546 imported classes we use @code{pragma Import} with convention @code{C_Plus_Plus}
3547 to indicate that they have been defined on the C++ side; this is required
3548 because the dispatch table associated with these tagged types will be built
3549 in the C++ side and therefore will not contain the predefined Ada primitives
3550 which Ada would otherwise expect.
3552 As the reader can see there is no need to indicate the C++ mangled names
3553 associated with each subprogram because it is assumed that all the calls to
3554 these primitives will be dispatching calls. The only exception is the
3555 constructor, which must be registered with the compiler by means of
3556 @code{pragma CPP_Constructor} and needs to provide its associated C++
3557 mangled name because the Ada compiler generates direct calls to it.
3559 With the above packages we can now declare objects of type Dog on the Ada side
3560 and dispatch calls to the corresponding subprograms on the C++ side. We can
3561 also extend the tagged type Dog with further fields and primitives, and
3562 override some of its C++ primitives on the Ada side. For example, here we have
3563 a type derivation defined on the Ada side that inherits all the dispatching
3564 primitives of the ancestor from the C++ side.
3567 @b{with} Animals; @b{use} Animals;
3568 @b{package} Vaccinated_Animals @b{is}
3569 @b{type} Vaccinated_Dog @b{is new} Dog @b{with null record};
3570 @b{function} Vaccination_Expired (A : Vaccinated_Dog) @b{return} Boolean;
3571 @b{end} Vaccinated_Animals;
3574 It is important to note that, because of the ABI compatibility, the programmer
3575 does not need to add any further information to indicate either the object
3576 layout or the dispatch table entry associated with each dispatching operation.
3578 Now let us define all the types and constructors on the Ada side and export
3579 them to C++, using the same hierarchy of our previous example:
3581 @smallexample @c ada
3582 with Interfaces.C.Strings;
3583 use Interfaces.C.Strings;
3585 type Carnivore is interface;
3586 pragma Convention (C_Plus_Plus, Carnivore);
3587 function Number_Of_Teeth (X : Carnivore)
3588 return Natural is abstract;
3590 type Domestic is interface;
3591 pragma Convention (C_Plus_Plus, Set_Owner);
3593 (X : in out Domestic;
3594 Name : Chars_Ptr) is abstract;
3596 type Animal is tagged record
3599 pragma Convention (C_Plus_Plus, Animal);
3601 procedure Set_Age (X : in out Animal; Age : Integer);
3602 pragma Export (C_Plus_Plus, Set_Age);
3604 function Age (X : Animal) return Integer;
3605 pragma Export (C_Plus_Plus, Age);
3607 type Dog is new Animal and Carnivore and Domestic with record
3608 Tooth_Count : Natural;
3609 Owner : String (1 .. 30);
3611 pragma Convention (C_Plus_Plus, Dog);
3613 function Number_Of_Teeth (A : Dog) return Integer;
3614 pragma Export (C_Plus_Plus, Number_Of_Teeth);
3616 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
3617 pragma Export (C_Plus_Plus, Set_Owner);
3619 function New_Dog return Dog'Class;
3620 pragma Export (C_Plus_Plus, New_Dog);
3624 Compared with our previous example the only difference is the use of
3625 @code{pragma Export} to indicate to the GNAT compiler that the primitives will
3626 be available to C++. Thanks to the ABI compatibility, on the C++ side there is
3627 nothing else to be done; as explained above, the only requirement is that all
3628 the primitives and components are declared in exactly the same order.
3630 For completeness, let us see a brief C++ main program that uses the
3631 declarations available in @code{animals.h} (presented in our first example) to
3632 import and use the declarations from the Ada side, properly initializing and
3633 finalizing the Ada run-time system along the way:
3636 @b{#include} "animals.h"
3637 @b{#include} <iostream>
3638 @b{using namespace} std;
3640 void Check_Carnivore (Carnivore *obj) @{@dots{}@}
3641 void Check_Domestic (Domestic *obj) @{@dots{}@}
3642 void Check_Animal (Animal *obj) @{@dots{}@}
3643 void Check_Dog (Dog *obj) @{@dots{}@}
3646 void adainit (void);
3647 void adafinal (void);
3653 Dog *obj = new_dog(); // Ada constructor
3654 Check_Carnivore (obj); // Check secondary DT
3655 Check_Domestic (obj); // Check secondary DT
3656 Check_Animal (obj); // Check primary DT
3657 Check_Dog (obj); // Check primary DT
3662 adainit (); test(); adafinal ();
3667 @node Comparison between GNAT and C/C++ Compilation Models
3668 @section Comparison between GNAT and C/C++ Compilation Models
3671 The GNAT model of compilation is close to the C and C++ models. You can
3672 think of Ada specs as corresponding to header files in C. As in C, you
3673 don't need to compile specs; they are compiled when they are used. The
3674 Ada @code{with} is similar in effect to the @code{#include} of a C
3677 One notable difference is that, in Ada, you may compile specs separately
3678 to check them for semantic and syntactic accuracy. This is not always
3679 possible with C headers because they are fragments of programs that have
3680 less specific syntactic or semantic rules.
3682 The other major difference is the requirement for running the binder,
3683 which performs two important functions. First, it checks for
3684 consistency. In C or C++, the only defense against assembling
3685 inconsistent programs lies outside the compiler, in a makefile, for
3686 example. The binder satisfies the Ada requirement that it be impossible
3687 to construct an inconsistent program when the compiler is used in normal
3690 @cindex Elaboration order control
3691 The other important function of the binder is to deal with elaboration
3692 issues. There are also elaboration issues in C++ that are handled
3693 automatically. This automatic handling has the advantage of being
3694 simpler to use, but the C++ programmer has no control over elaboration.
3695 Where @code{gnatbind} might complain there was no valid order of
3696 elaboration, a C++ compiler would simply construct a program that
3697 malfunctioned at run time.
3700 @node Comparison between GNAT and Conventional Ada Library Models
3701 @section Comparison between GNAT and Conventional Ada Library Models
3704 This section is intended for Ada programmers who have
3705 used an Ada compiler implementing the traditional Ada library
3706 model, as described in the Ada Reference Manual.
3708 @cindex GNAT library
3709 In GNAT, there is no ``library'' in the normal sense. Instead, the set of
3710 source files themselves acts as the library. Compiling Ada programs does
3711 not generate any centralized information, but rather an object file and
3712 a ALI file, which are of interest only to the binder and linker.
3713 In a traditional system, the compiler reads information not only from
3714 the source file being compiled, but also from the centralized library.
3715 This means that the effect of a compilation depends on what has been
3716 previously compiled. In particular:
3720 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3721 to the version of the unit most recently compiled into the library.
3724 Inlining is effective only if the necessary body has already been
3725 compiled into the library.
3728 Compiling a unit may obsolete other units in the library.
3732 In GNAT, compiling one unit never affects the compilation of any other
3733 units because the compiler reads only source files. Only changes to source
3734 files can affect the results of a compilation. In particular:
3738 When a unit is @code{with}'ed, the unit seen by the compiler corresponds
3739 to the source version of the unit that is currently accessible to the
3744 Inlining requires the appropriate source files for the package or
3745 subprogram bodies to be available to the compiler. Inlining is always
3746 effective, independent of the order in which units are complied.
3749 Compiling a unit never affects any other compilations. The editing of
3750 sources may cause previous compilations to be out of date if they
3751 depended on the source file being modified.
3755 The most important result of these differences is that order of compilation
3756 is never significant in GNAT. There is no situation in which one is
3757 required to do one compilation before another. What shows up as order of
3758 compilation requirements in the traditional Ada library becomes, in
3759 GNAT, simple source dependencies; in other words, there is only a set
3760 of rules saying what source files must be present when a file is
3764 @node Placement of temporary files
3765 @section Placement of temporary files
3766 @cindex Temporary files (user control over placement)
3769 GNAT creates temporary files in the directory designated by the environment
3770 variable @env{TMPDIR}.
3771 (See the HP @emph{C RTL Reference Manual} on the function @code{getenv()}
3772 for detailed information on how environment variables are resolved.
3773 For most users the easiest way to make use of this feature is to simply
3774 define @env{TMPDIR} as a job level logical name).
3775 For example, if you wish to use a Ramdisk (assuming DECRAM is installed)
3776 for compiler temporary files, then you can include something like the
3777 following command in your @file{LOGIN.COM} file:
3780 $ define/job TMPDIR "/disk$scratchram/000000/temp/"
3784 If @env{TMPDIR} is not defined, then GNAT uses the directory designated by
3785 @env{TMP}; if @env{TMP} is not defined, then GNAT uses the directory
3786 designated by @env{TEMP}.
3787 If none of these environment variables are defined then GNAT uses the
3788 directory designated by the logical name @code{SYS$SCRATCH:}
3789 (by default the user's home directory). If all else fails
3790 GNAT uses the current directory for temporary files.
3793 @c *************************
3794 @node Compiling Using gcc
3795 @chapter Compiling Using @command{gcc}
3798 This chapter discusses how to compile Ada programs using the @command{gcc}
3799 command. It also describes the set of switches
3800 that can be used to control the behavior of the compiler.
3802 * Compiling Programs::
3803 * Switches for gcc::
3804 * Search Paths and the Run-Time Library (RTL)::
3805 * Order of Compilation Issues::
3809 @node Compiling Programs
3810 @section Compiling Programs
3813 The first step in creating an executable program is to compile the units
3814 of the program using the @command{gcc} command. You must compile the
3819 the body file (@file{.adb}) for a library level subprogram or generic
3823 the spec file (@file{.ads}) for a library level package or generic
3824 package that has no body
3827 the body file (@file{.adb}) for a library level package
3828 or generic package that has a body
3833 You need @emph{not} compile the following files
3838 the spec of a library unit which has a body
3845 because they are compiled as part of compiling related units. GNAT
3847 when the corresponding body is compiled, and subunits when the parent is
3850 @cindex cannot generate code
3851 If you attempt to compile any of these files, you will get one of the
3852 following error messages (where @var{fff} is the name of the file you compiled):
3855 cannot generate code for file @var{fff} (package spec)
3856 to check package spec, use -gnatc
3858 cannot generate code for file @var{fff} (missing subunits)
3859 to check parent unit, use -gnatc
3861 cannot generate code for file @var{fff} (subprogram spec)
3862 to check subprogram spec, use -gnatc
3864 cannot generate code for file @var{fff} (subunit)
3865 to check subunit, use -gnatc
3869 As indicated by the above error messages, if you want to submit
3870 one of these files to the compiler to check for correct semantics
3871 without generating code, then use the @option{-gnatc} switch.
3873 The basic command for compiling a file containing an Ada unit is
3876 $ gcc -c @ovar{switches} @file{file name}
3880 where @var{file name} is the name of the Ada file (usually
3882 @file{.ads} for a spec or @file{.adb} for a body).
3885 @option{-c} switch to tell @command{gcc} to compile, but not link, the file.
3887 The result of a successful compilation is an object file, which has the
3888 same name as the source file but an extension of @file{.o} and an Ada
3889 Library Information (ALI) file, which also has the same name as the
3890 source file, but with @file{.ali} as the extension. GNAT creates these
3891 two output files in the current directory, but you may specify a source
3892 file in any directory using an absolute or relative path specification
3893 containing the directory information.
3896 @command{gcc} is actually a driver program that looks at the extensions of
3897 the file arguments and loads the appropriate compiler. For example, the
3898 GNU C compiler is @file{cc1}, and the Ada compiler is @file{gnat1}.
3899 These programs are in directories known to the driver program (in some
3900 configurations via environment variables you set), but need not be in
3901 your path. The @command{gcc} driver also calls the assembler and any other
3902 utilities needed to complete the generation of the required object
3905 It is possible to supply several file names on the same @command{gcc}
3906 command. This causes @command{gcc} to call the appropriate compiler for
3907 each file. For example, the following command lists three separate
3908 files to be compiled:
3911 $ gcc -c x.adb y.adb z.c
3915 calls @code{gnat1} (the Ada compiler) twice to compile @file{x.adb} and
3916 @file{y.adb}, and @code{cc1} (the C compiler) once to compile @file{z.c}.
3917 The compiler generates three object files @file{x.o}, @file{y.o} and
3918 @file{z.o} and the two ALI files @file{x.ali} and @file{y.ali} from the
3919 Ada compilations. Any switches apply to all the files ^listed,^listed.^
3922 @option{-gnat@var{x}} switches, which apply only to Ada compilations.
3925 @node Switches for gcc
3926 @section Switches for @command{gcc}
3929 The @command{gcc} command accepts switches that control the
3930 compilation process. These switches are fully described in this section.
3931 First we briefly list all the switches, in alphabetical order, then we
3932 describe the switches in more detail in functionally grouped sections.
3934 More switches exist for GCC than those documented here, especially
3935 for specific targets. However, their use is not recommended as
3936 they may change code generation in ways that are incompatible with
3937 the Ada run-time library, or can cause inconsistencies between
3941 * Output and Error Message Control::
3942 * Warning Message Control::
3943 * Debugging and Assertion Control::
3944 * Validity Checking::
3947 * Using gcc for Syntax Checking::
3948 * Using gcc for Semantic Checking::
3949 * Compiling Different Versions of Ada::
3950 * Character Set Control::
3951 * File Naming Control::
3952 * Subprogram Inlining Control::
3953 * Auxiliary Output Control::
3954 * Debugging Control::
3955 * Exception Handling Control::
3956 * Units to Sources Mapping Files::
3957 * Integrated Preprocessing::
3958 * Code Generation Control::
3967 @cindex @option{-b} (@command{gcc})
3968 @item -b @var{target}
3969 Compile your program to run on @var{target}, which is the name of a
3970 system configuration. You must have a GNAT cross-compiler built if
3971 @var{target} is not the same as your host system.
3974 @cindex @option{-B} (@command{gcc})
3975 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
3976 from @var{dir} instead of the default location. Only use this switch
3977 when multiple versions of the GNAT compiler are available.
3978 @xref{Directory Options,, Options for Directory Search, gcc, Using the
3979 GNU Compiler Collection (GCC)}, for further details. You would normally
3980 use the @option{-b} or @option{-V} switch instead.
3983 @cindex @option{-c} (@command{gcc})
3984 Compile. Always use this switch when compiling Ada programs.
3986 Note: for some other languages when using @command{gcc}, notably in
3987 the case of C and C++, it is possible to use
3988 use @command{gcc} without a @option{-c} switch to
3989 compile and link in one step. In the case of GNAT, you
3990 cannot use this approach, because the binder must be run
3991 and @command{gcc} cannot be used to run the GNAT binder.
3995 @cindex @option{-fno-inline} (@command{gcc})
3996 Suppresses all back-end inlining, even if other optimization or inlining
3998 This includes suppression of inlining that results
3999 from the use of the pragma @code{Inline_Always}.
4000 Any occurrences of pragma @code{Inline} or @code{Inline_Always}
4001 are ignored, and @option{-gnatn} and @option{-gnatN} have no
4002 effect if this switch is present.
4004 @item -fno-inline-functions
4005 @cindex @option{-fno-inline-functions} (@command{gcc})
4006 Suppresses automatic inlining of simple subprograms, which is enabled
4007 if @option{-O3} is used.
4009 @item -fno-inline-small-functions
4010 @cindex @option{-fno-inline-small-functions} (@command{gcc})
4011 Suppresses automatic inlining of small subprograms, which is enabled
4012 if @option{-O2} is used.
4014 @item -fno-inline-functions-called-once
4015 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
4016 Suppresses inlining of subprograms local to the unit and called once
4017 from within it, which is enabled if @option{-O1} is used.
4020 @cindex @option{-fno-ivopts} (@command{gcc})
4021 Suppresses high-level loop induction variable optimizations, which are
4022 enabled if @option{-O1} is used. These optimizations are generally
4023 profitable but, for some specific cases of loops with numerous uses
4024 of the iteration variable that follow a common pattern, they may end
4025 up destroying the regularity that could be exploited at a lower level
4026 and thus producing inferior code.
4028 @item -fno-strict-aliasing
4029 @cindex @option{-fno-strict-aliasing} (@command{gcc})
4030 Causes the compiler to avoid assumptions regarding non-aliasing
4031 of objects of different types. See
4032 @ref{Optimization and Strict Aliasing} for details.
4035 @cindex @option{-fstack-check} (@command{gcc})
4036 Activates stack checking.
4037 See @ref{Stack Overflow Checking} for details.
4040 @cindex @option{-fstack-usage} (@command{gcc})
4041 Makes the compiler output stack usage information for the program, on a
4042 per-function basis. See @ref{Static Stack Usage Analysis} for details.
4044 @item -fcallgraph-info@r{[}=su@r{]}
4045 @cindex @option{-fcallgraph-info} (@command{gcc})
4046 Makes the compiler output callgraph information for the program, on a
4047 per-file basis. The information is generated in the VCG format. It can
4048 be decorated with stack-usage per-node information.
4051 @cindex @option{^-g^/DEBUG^} (@command{gcc})
4052 Generate debugging information. This information is stored in the object
4053 file and copied from there to the final executable file by the linker,
4054 where it can be read by the debugger. You must use the
4055 @option{^-g^/DEBUG^} switch if you plan on using the debugger.
4058 @cindex @option{-gnat83} (@command{gcc})
4059 Enforce Ada 83 restrictions.
4062 @cindex @option{-gnat95} (@command{gcc})
4063 Enforce Ada 95 restrictions.
4066 @cindex @option{-gnat05} (@command{gcc})
4067 Allow full Ada 2005 features.
4070 @cindex @option{-gnata} (@command{gcc})
4071 Assertions enabled. @code{Pragma Assert} and @code{pragma Debug} to be
4072 activated. Note that these pragmas can also be controlled using the
4073 configuration pragmas @code{Assertion_Policy} and @code{Debug_Policy}.
4074 It also activates pragmas @code{Check}, @code{Precondition}, and
4075 @code{Postcondition}. Note that these pragmas can also be controlled
4076 using the configuration pragma @code{Check_Policy}.
4079 @cindex @option{-gnatA} (@command{gcc})
4080 Avoid processing @file{gnat.adc}. If a @file{gnat.adc} file is present,
4084 @cindex @option{-gnatb} (@command{gcc})
4085 Generate brief messages to @file{stderr} even if verbose mode set.
4088 @cindex @option{-gnatB} (@command{gcc})
4089 Assume no invalid (bad) values except for 'Valid attribute use
4090 (@pxref{Validity Checking}).
4093 @cindex @option{-gnatc} (@command{gcc})
4094 Check syntax and semantics only (no code generation attempted).
4097 @cindex @option{-gnatC} (@command{gcc})
4098 Generate CodePeer information (no code generation attempted).
4099 This switch will generate an intermediate representation suitable for
4100 use by CodePeer (@file{.scil} files). This switch is not compatible with
4101 code generation (it will, among other things, disable some switches such
4102 as -gnatn, and enable others such as -gnata).
4105 @cindex @option{-gnatd} (@command{gcc})
4106 Specify debug options for the compiler. The string of characters after
4107 the @option{-gnatd} specify the specific debug options. The possible
4108 characters are 0-9, a-z, A-Z, optionally preceded by a dot. See
4109 compiler source file @file{debug.adb} for details of the implemented
4110 debug options. Certain debug options are relevant to applications
4111 programmers, and these are documented at appropriate points in this
4116 @cindex @option{-gnatD[nn]} (@command{gcc})
4119 @item /XDEBUG /LXDEBUG=nnn
4121 Create expanded source files for source level debugging. This switch
4122 also suppress generation of cross-reference information
4123 (see @option{-gnatx}).
4125 @item -gnatec=@var{path}
4126 @cindex @option{-gnatec} (@command{gcc})
4127 Specify a configuration pragma file
4129 (the equal sign is optional)
4131 (@pxref{The Configuration Pragmas Files}).
4133 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=@var{value}@r{]}
4134 @cindex @option{-gnateD} (@command{gcc})
4135 Defines a symbol, associated with @var{value}, for preprocessing.
4136 (@pxref{Integrated Preprocessing}).
4139 @cindex @option{-gnatef} (@command{gcc})
4140 Display full source path name in brief error messages.
4143 @cindex @option{-gnateG} (@command{gcc})
4144 Save result of preprocessing in a text file.
4146 @item -gnatem=@var{path}
4147 @cindex @option{-gnatem} (@command{gcc})
4148 Specify a mapping file
4150 (the equal sign is optional)
4152 (@pxref{Units to Sources Mapping Files}).
4154 @item -gnatep=@var{file}
4155 @cindex @option{-gnatep} (@command{gcc})
4156 Specify a preprocessing data file
4158 (the equal sign is optional)
4160 (@pxref{Integrated Preprocessing}).
4163 @cindex @option{-gnateS} (@command{gcc})
4164 Generate SCO (Source Coverage Obligation) information in the ALI
4165 file. This information is used by advanced coverage tools. See
4166 unit @file{SCOs} in the compiler sources for details in files
4167 @file{scos.ads} and @file{scos.adb}.
4170 @cindex @option{-gnatE} (@command{gcc})
4171 Full dynamic elaboration checks.
4174 @cindex @option{-gnatf} (@command{gcc})
4175 Full errors. Multiple errors per line, all undefined references, do not
4176 attempt to suppress cascaded errors.
4179 @cindex @option{-gnatF} (@command{gcc})
4180 Externals names are folded to all uppercase.
4182 @item ^-gnatg^/GNAT_INTERNAL^
4183 @cindex @option{^-gnatg^/GNAT_INTERNAL^} (@command{gcc})
4184 Internal GNAT implementation mode. This should not be used for
4185 applications programs, it is intended only for use by the compiler
4186 and its run-time library. For documentation, see the GNAT sources.
4187 Note that @option{^-gnatg^/GNAT_INTERNAL^} implies
4188 @option{^-gnatwae^/WARNINGS=ALL,ERRORS^} and
4189 @option{^-gnatyg^/STYLE_CHECKS=GNAT^}
4190 so that all standard warnings and all standard style options are turned on.
4191 All warnings and style error messages are treated as errors.
4195 @cindex @option{-gnatG[nn]} (@command{gcc})
4198 @item /EXPAND_SOURCE, /LEXPAND_SOURCE=nnn
4200 List generated expanded code in source form.
4202 @item ^-gnath^/HELP^
4203 @cindex @option{^-gnath^/HELP^} (@command{gcc})
4204 Output usage information. The output is written to @file{stdout}.
4206 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
4207 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
4208 Identifier character set
4210 (@var{c}=1/2/3/4/8/9/p/f/n/w).
4212 For details of the possible selections for @var{c},
4213 see @ref{Character Set Control}.
4215 @item ^-gnatI^/IGNORE_REP_CLAUSES^
4216 @cindex @option{^-gnatI^IGNORE_REP_CLAUSES^} (@command{gcc})
4217 Ignore representation clauses. When this switch is used,
4218 representation clauses are treated as comments. This is useful
4219 when initially porting code where you want to ignore rep clause
4220 problems, and also for compiling foreign code (particularly
4221 for use with ASIS). The representation clauses that are ignored
4222 are: enumeration_representation_clause, record_representation_clause,
4223 and attribute_definition_clause for the following attributes:
4224 Address, Alignment, Bit_Order, Component_Size, Machine_Radix,
4225 Object_Size, Size, Small, Stream_Size, and Value_Size.
4226 Note that this option should be used only for compiling -- the
4227 code is likely to malfunction at run time.
4230 @cindex @option{-gnatjnn} (@command{gcc})
4231 Reformat error messages to fit on nn character lines
4233 @item -gnatk=@var{n}
4234 @cindex @option{-gnatk} (@command{gcc})
4235 Limit file names to @var{n} (1-999) characters ^(@code{k} = krunch)^^.
4238 @cindex @option{-gnatl} (@command{gcc})
4239 Output full source listing with embedded error messages.
4242 @cindex @option{-gnatL} (@command{gcc})
4243 Used in conjunction with -gnatG or -gnatD to intersperse original
4244 source lines (as comment lines with line numbers) in the expanded
4247 @item -gnatm=@var{n}
4248 @cindex @option{-gnatm} (@command{gcc})
4249 Limit number of detected error or warning messages to @var{n}
4250 where @var{n} is in the range 1..999999. The default setting if
4251 no switch is given is 9999. If the number of warnings reaches this
4252 limit, then a message is output and further warnings are suppressed,
4253 but the compilation is continued. If the number of error messages
4254 reaches this limit, then a message is output and the compilation
4255 is abandoned. The equal sign here is optional. A value of zero
4256 means that no limit applies.
4259 @cindex @option{-gnatn} (@command{gcc})
4260 Activate inlining for subprograms for which
4261 pragma @code{inline} is specified. This inlining is performed
4262 by the GCC back-end.
4265 @cindex @option{-gnatN} (@command{gcc})
4266 Activate front end inlining for subprograms for which
4267 pragma @code{Inline} is specified. This inlining is performed
4268 by the front end and will be visible in the
4269 @option{-gnatG} output.
4271 When using a gcc-based back end (in practice this means using any version
4272 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
4273 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
4274 Historically front end inlining was more extensive than the gcc back end
4275 inlining, but that is no longer the case.
4278 @cindex @option{-gnato} (@command{gcc})
4279 Enable numeric overflow checking (which is not normally enabled by
4280 default). Note that division by zero is a separate check that is not
4281 controlled by this switch (division by zero checking is on by default).
4284 @cindex @option{-gnatp} (@command{gcc})
4285 Suppress all checks. See @ref{Run-Time Checks} for details.
4288 @cindex @option{-gnatP} (@command{gcc})
4289 Enable polling. This is required on some systems (notably Windows NT) to
4290 obtain asynchronous abort and asynchronous transfer of control capability.
4291 @xref{Pragma Polling,,, gnat_rm, GNAT Reference Manual}, for full
4295 @cindex @option{-gnatq} (@command{gcc})
4296 Don't quit. Try semantics, even if parse errors.
4299 @cindex @option{-gnatQ} (@command{gcc})
4300 Don't quit. Generate @file{ALI} and tree files even if illegalities.
4303 @cindex @option{-gnatr} (@command{gcc})
4304 Treat pragma Restrictions as Restriction_Warnings.
4306 @item ^-gnatR@r{[}0@r{/}1@r{/}2@r{/}3@r{[}s@r{]]}^/REPRESENTATION_INFO^
4307 @cindex @option{-gnatR} (@command{gcc})
4308 Output representation information for declared types and objects.
4311 @cindex @option{-gnats} (@command{gcc})
4315 @cindex @option{-gnatS} (@command{gcc})
4316 Print package Standard.
4319 @cindex @option{-gnatt} (@command{gcc})
4320 Generate tree output file.
4322 @item ^-gnatT^/TABLE_MULTIPLIER=^@var{nnn}
4323 @cindex @option{^-gnatT^/TABLE_MULTIPLIER^} (@command{gcc})
4324 All compiler tables start at @var{nnn} times usual starting size.
4327 @cindex @option{-gnatu} (@command{gcc})
4328 List units for this compilation.
4331 @cindex @option{-gnatU} (@command{gcc})
4332 Tag all error messages with the unique string ``error:''
4335 @cindex @option{-gnatv} (@command{gcc})
4336 Verbose mode. Full error output with source lines to @file{stdout}.
4339 @cindex @option{-gnatV} (@command{gcc})
4340 Control level of validity checking (@pxref{Validity Checking}).
4342 @item ^-gnatw@var{xxx}^/WARNINGS=(@var{option}@r{[},@dots{}@r{]})^
4343 @cindex @option{^-gnatw^/WARNINGS^} (@command{gcc})
4345 ^@var{xxx} is a string of option letters that^the list of options^ denotes
4346 the exact warnings that
4347 are enabled or disabled (@pxref{Warning Message Control}).
4349 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
4350 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
4351 Wide character encoding method
4353 (@var{e}=n/h/u/s/e/8).
4356 (@var{e}=@code{BRACKETS, NONE, HEX, UPPER, SHIFT_JIS, EUC, UTF8})
4360 @cindex @option{-gnatx} (@command{gcc})
4361 Suppress generation of cross-reference information.
4363 @item ^-gnaty^/STYLE_CHECKS=(option,option@dots{})^
4364 @cindex @option{^-gnaty^/STYLE_CHECKS^} (@command{gcc})
4365 Enable built-in style checks (@pxref{Style Checking}).
4367 @item ^-gnatz^/DISTRIBUTION_STUBS=^@var{m}
4368 @cindex @option{^-gnatz^/DISTRIBUTION_STUBS^} (@command{gcc})
4369 Distribution stub generation and compilation
4371 (@var{m}=r/c for receiver/caller stubs).
4374 (@var{m}=@code{RECEIVER} or @code{CALLER} to specify the type of stubs
4375 to be generated and compiled).
4378 @item ^-I^/SEARCH=^@var{dir}
4379 @cindex @option{^-I^/SEARCH^} (@command{gcc})
4381 Direct GNAT to search the @var{dir} directory for source files needed by
4382 the current compilation
4383 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4385 @item ^-I-^/NOCURRENT_DIRECTORY^
4386 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gcc})
4388 Except for the source file named in the command line, do not look for source
4389 files in the directory containing the source file named in the command line
4390 (@pxref{Search Paths and the Run-Time Library (RTL)}).
4394 @cindex @option{-mbig-switch} (@command{gcc})
4395 @cindex @code{case} statement (effect of @option{-mbig-switch} option)
4396 This standard gcc switch causes the compiler to use larger offsets in its
4397 jump table representation for @code{case} statements.
4398 This may result in less efficient code, but is sometimes necessary
4399 (for example on HP-UX targets)
4400 @cindex HP-UX and @option{-mbig-switch} option
4401 in order to compile large and/or nested @code{case} statements.
4404 @cindex @option{-o} (@command{gcc})
4405 This switch is used in @command{gcc} to redirect the generated object file
4406 and its associated ALI file. Beware of this switch with GNAT, because it may
4407 cause the object file and ALI file to have different names which in turn
4408 may confuse the binder and the linker.
4412 @cindex @option{-nostdinc} (@command{gcc})
4413 Inhibit the search of the default location for the GNAT Run Time
4414 Library (RTL) source files.
4417 @cindex @option{-nostdlib} (@command{gcc})
4418 Inhibit the search of the default location for the GNAT Run Time
4419 Library (RTL) ALI files.
4423 @cindex @option{-O} (@command{gcc})
4424 @var{n} controls the optimization level.
4428 No optimization, the default setting if no @option{-O} appears
4431 Normal optimization, the default if you specify @option{-O} without
4432 an operand. A good compromise between code quality and compilation
4436 Extensive optimization, may improve execution time, possibly at the cost of
4437 substantially increased compilation time.
4440 Same as @option{-O2}, and also includes inline expansion for small subprograms
4444 Optimize space usage
4448 See also @ref{Optimization Levels}.
4453 @cindex @option{/NOOPTIMIZE} (@code{GNAT COMPILE})
4454 Equivalent to @option{/OPTIMIZE=NONE}.
4455 This is the default behavior in the absence of an @option{/OPTIMIZE}
4458 @item /OPTIMIZE@r{[}=(keyword@r{[},@dots{}@r{]})@r{]}
4459 @cindex @option{/OPTIMIZE} (@code{GNAT COMPILE})
4460 Selects the level of optimization for your program. The supported
4461 keywords are as follows:
4464 Perform most optimizations, including those that
4466 This is the default if the @option{/OPTIMIZE} qualifier is supplied
4467 without keyword options.
4470 Do not do any optimizations. Same as @code{/NOOPTIMIZE}.
4473 Perform some optimizations, but omit ones that are costly.
4476 Same as @code{SOME}.
4479 Full optimization as in @option{/OPTIMIZE=ALL}, and also attempts
4480 automatic inlining of small subprograms within a unit
4483 Try to unroll loops. This keyword may be specified together with
4484 any keyword above other than @code{NONE}. Loop unrolling
4485 usually, but not always, improves the performance of programs.
4488 Optimize space usage
4492 See also @ref{Optimization Levels}.
4496 @item -pass-exit-codes
4497 @cindex @option{-pass-exit-codes} (@command{gcc})
4498 Catch exit codes from the compiler and use the most meaningful as
4502 @item --RTS=@var{rts-path}
4503 @cindex @option{--RTS} (@command{gcc})
4504 Specifies the default location of the runtime library. Same meaning as the
4505 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
4508 @cindex @option{^-S^/ASM^} (@command{gcc})
4509 ^Used in place of @option{-c} to^Used to^
4510 cause the assembler source file to be
4511 generated, using @file{^.s^.S^} as the extension,
4512 instead of the object file.
4513 This may be useful if you need to examine the generated assembly code.
4515 @item ^-fverbose-asm^/VERBOSE_ASM^
4516 @cindex @option{^-fverbose-asm^/VERBOSE_ASM^} (@command{gcc})
4517 ^Used in conjunction with @option{-S}^Used in place of @option{/ASM}^
4518 to cause the generated assembly code file to be annotated with variable
4519 names, making it significantly easier to follow.
4522 @cindex @option{^-v^/VERBOSE^} (@command{gcc})
4523 Show commands generated by the @command{gcc} driver. Normally used only for
4524 debugging purposes or if you need to be sure what version of the
4525 compiler you are executing.
4529 @cindex @option{-V} (@command{gcc})
4530 Execute @var{ver} version of the compiler. This is the @command{gcc}
4531 version, not the GNAT version.
4534 @item ^-w^/NO_BACK_END_WARNINGS^
4535 @cindex @option{-w} (@command{gcc})
4536 Turn off warnings generated by the back end of the compiler. Use of
4537 this switch also causes the default for front end warnings to be set
4538 to suppress (as though @option{-gnatws} had appeared at the start of
4544 @c Combining qualifiers does not work on VMS
4545 You may combine a sequence of GNAT switches into a single switch. For
4546 example, the combined switch
4548 @cindex Combining GNAT switches
4554 is equivalent to specifying the following sequence of switches:
4557 -gnato -gnatf -gnati3
4562 The following restrictions apply to the combination of switches
4567 The switch @option{-gnatc} if combined with other switches must come
4568 first in the string.
4571 The switch @option{-gnats} if combined with other switches must come
4572 first in the string.
4576 ^^@option{/DISTRIBUTION_STUBS=},^
4577 @option{-gnatzc} and @option{-gnatzr} may not be combined with any other
4578 switches, and only one of them may appear in the command line.
4582 Once a ``y'' appears in the string (that is a use of the @option{-gnaty}
4583 switch), then all further characters in the switch are interpreted
4584 as style modifiers (see description of @option{-gnaty}).
4587 Once a ``d'' appears in the string (that is a use of the @option{-gnatd}
4588 switch), then all further characters in the switch are interpreted
4589 as debug flags (see description of @option{-gnatd}).
4592 Once a ``w'' appears in the string (that is a use of the @option{-gnatw}
4593 switch), then all further characters in the switch are interpreted
4594 as warning mode modifiers (see description of @option{-gnatw}).
4597 Once a ``V'' appears in the string (that is a use of the @option{-gnatV}
4598 switch), then all further characters in the switch are interpreted
4599 as validity checking options (@pxref{Validity Checking}).
4602 Option ``em'', ``ec'', ``ep'', ``l='' and ``R'' must be the last options in
4603 a combined list of options.
4607 @node Output and Error Message Control
4608 @subsection Output and Error Message Control
4612 The standard default format for error messages is called ``brief format''.
4613 Brief format messages are written to @file{stderr} (the standard error
4614 file) and have the following form:
4617 e.adb:3:04: Incorrect spelling of keyword "function"
4618 e.adb:4:20: ";" should be "is"
4622 The first integer after the file name is the line number in the file,
4623 and the second integer is the column number within the line.
4625 @code{GPS} can parse the error messages
4626 and point to the referenced character.
4628 The following switches provide control over the error message
4634 @cindex @option{-gnatv} (@command{gcc})
4637 The v stands for verbose.
4639 The effect of this setting is to write long-format error
4640 messages to @file{stdout} (the standard output file.
4641 The same program compiled with the
4642 @option{-gnatv} switch would generate:
4646 3. funcion X (Q : Integer)
4648 >>> Incorrect spelling of keyword "function"
4651 >>> ";" should be "is"
4656 The vertical bar indicates the location of the error, and the @samp{>>>}
4657 prefix can be used to search for error messages. When this switch is
4658 used the only source lines output are those with errors.
4661 @cindex @option{-gnatl} (@command{gcc})
4663 The @code{l} stands for list.
4665 This switch causes a full listing of
4666 the file to be generated. In the case where a body is
4667 compiled, the corresponding spec is also listed, along
4668 with any subunits. Typical output from compiling a package
4669 body @file{p.adb} might look like:
4671 @smallexample @c ada
4675 1. package body p is
4677 3. procedure a is separate;
4688 2. pragma Elaborate_Body
4712 When you specify the @option{-gnatv} or @option{-gnatl} switches and
4713 standard output is redirected, a brief summary is written to
4714 @file{stderr} (standard error) giving the number of error messages and
4715 warning messages generated.
4717 @item -^gnatl^OUTPUT_FILE^=file
4718 @cindex @option{^-gnatl^OUTPUT_FILE^=fname} (@command{gcc})
4719 This has the same effect as @option{-gnatl} except that the output is
4720 written to a file instead of to standard output. If the given name
4721 @file{fname} does not start with a period, then it is the full name
4722 of the file to be written. If @file{fname} is an extension, it is
4723 appended to the name of the file being compiled. For example, if
4724 file @file{xyz.adb} is compiled with @option{^-gnatl^OUTPUT_FILE^=.lst},
4725 then the output is written to file ^xyz.adb.lst^xyz.adb_lst^.
4728 @cindex @option{-gnatU} (@command{gcc})
4729 This switch forces all error messages to be preceded by the unique
4730 string ``error:''. This means that error messages take a few more
4731 characters in space, but allows easy searching for and identification
4735 @cindex @option{-gnatb} (@command{gcc})
4737 The @code{b} stands for brief.
4739 This switch causes GNAT to generate the
4740 brief format error messages to @file{stderr} (the standard error
4741 file) as well as the verbose
4742 format message or full listing (which as usual is written to
4743 @file{stdout} (the standard output file).
4745 @item -gnatm=@var{n}
4746 @cindex @option{-gnatm} (@command{gcc})
4748 The @code{m} stands for maximum.
4750 @var{n} is a decimal integer in the
4751 range of 1 to 999999 and limits the number of error or warning
4752 messages to be generated. For example, using
4753 @option{-gnatm2} might yield
4756 e.adb:3:04: Incorrect spelling of keyword "function"
4757 e.adb:5:35: missing ".."
4758 fatal error: maximum number of errors detected
4759 compilation abandoned
4763 The default setting if
4764 no switch is given is 9999. If the number of warnings reaches this
4765 limit, then a message is output and further warnings are suppressed,
4766 but the compilation is continued. If the number of error messages
4767 reaches this limit, then a message is output and the compilation
4768 is abandoned. A value of zero means that no limit applies.
4771 Note that the equal sign is optional, so the switches
4772 @option{-gnatm2} and @option{-gnatm=2} are equivalent.
4775 @cindex @option{-gnatf} (@command{gcc})
4776 @cindex Error messages, suppressing
4778 The @code{f} stands for full.
4780 Normally, the compiler suppresses error messages that are likely to be
4781 redundant. This switch causes all error
4782 messages to be generated. In particular, in the case of
4783 references to undefined variables. If a given variable is referenced
4784 several times, the normal format of messages is
4786 e.adb:7:07: "V" is undefined (more references follow)
4790 where the parenthetical comment warns that there are additional
4791 references to the variable @code{V}. Compiling the same program with the
4792 @option{-gnatf} switch yields
4795 e.adb:7:07: "V" is undefined
4796 e.adb:8:07: "V" is undefined
4797 e.adb:8:12: "V" is undefined
4798 e.adb:8:16: "V" is undefined
4799 e.adb:9:07: "V" is undefined
4800 e.adb:9:12: "V" is undefined
4804 The @option{-gnatf} switch also generates additional information for
4805 some error messages. Some examples are:
4809 Details on possibly non-portable unchecked conversion
4811 List possible interpretations for ambiguous calls
4813 Additional details on incorrect parameters
4817 @cindex @option{-gnatjnn} (@command{gcc})
4818 In normal operation mode (or if @option{-gnatj0} is used, then error messages
4819 with continuation lines are treated as though the continuation lines were
4820 separate messages (and so a warning with two continuation lines counts as
4821 three warnings, and is listed as three separate messages).
4823 If the @option{-gnatjnn} switch is used with a positive value for nn, then
4824 messages are output in a different manner. A message and all its continuation
4825 lines are treated as a unit, and count as only one warning or message in the
4826 statistics totals. Furthermore, the message is reformatted so that no line
4827 is longer than nn characters.
4830 @cindex @option{-gnatq} (@command{gcc})
4832 The @code{q} stands for quit (really ``don't quit'').
4834 In normal operation mode, the compiler first parses the program and
4835 determines if there are any syntax errors. If there are, appropriate
4836 error messages are generated and compilation is immediately terminated.
4838 GNAT to continue with semantic analysis even if syntax errors have been
4839 found. This may enable the detection of more errors in a single run. On
4840 the other hand, the semantic analyzer is more likely to encounter some
4841 internal fatal error when given a syntactically invalid tree.
4844 @cindex @option{-gnatQ} (@command{gcc})
4845 In normal operation mode, the @file{ALI} file is not generated if any
4846 illegalities are detected in the program. The use of @option{-gnatQ} forces
4847 generation of the @file{ALI} file. This file is marked as being in
4848 error, so it cannot be used for binding purposes, but it does contain
4849 reasonably complete cross-reference information, and thus may be useful
4850 for use by tools (e.g., semantic browsing tools or integrated development
4851 environments) that are driven from the @file{ALI} file. This switch
4852 implies @option{-gnatq}, since the semantic phase must be run to get a
4853 meaningful ALI file.
4855 In addition, if @option{-gnatt} is also specified, then the tree file is
4856 generated even if there are illegalities. It may be useful in this case
4857 to also specify @option{-gnatq} to ensure that full semantic processing
4858 occurs. The resulting tree file can be processed by ASIS, for the purpose
4859 of providing partial information about illegal units, but if the error
4860 causes the tree to be badly malformed, then ASIS may crash during the
4863 When @option{-gnatQ} is used and the generated @file{ALI} file is marked as
4864 being in error, @command{gnatmake} will attempt to recompile the source when it
4865 finds such an @file{ALI} file, including with switch @option{-gnatc}.
4867 Note that @option{-gnatQ} has no effect if @option{-gnats} is specified,
4868 since ALI files are never generated if @option{-gnats} is set.
4872 @node Warning Message Control
4873 @subsection Warning Message Control
4874 @cindex Warning messages
4876 In addition to error messages, which correspond to illegalities as defined
4877 in the Ada Reference Manual, the compiler detects two kinds of warning
4880 First, the compiler considers some constructs suspicious and generates a
4881 warning message to alert you to a possible error. Second, if the
4882 compiler detects a situation that is sure to raise an exception at
4883 run time, it generates a warning message. The following shows an example
4884 of warning messages:
4886 e.adb:4:24: warning: creation of object may raise Storage_Error
4887 e.adb:10:17: warning: static value out of range
4888 e.adb:10:17: warning: "Constraint_Error" will be raised at run time
4892 GNAT considers a large number of situations as appropriate
4893 for the generation of warning messages. As always, warnings are not
4894 definite indications of errors. For example, if you do an out-of-range
4895 assignment with the deliberate intention of raising a
4896 @code{Constraint_Error} exception, then the warning that may be
4897 issued does not indicate an error. Some of the situations for which GNAT
4898 issues warnings (at least some of the time) are given in the following
4899 list. This list is not complete, and new warnings are often added to
4900 subsequent versions of GNAT. The list is intended to give a general idea
4901 of the kinds of warnings that are generated.
4905 Possible infinitely recursive calls
4908 Out-of-range values being assigned
4911 Possible order of elaboration problems
4914 Assertions (pragma Assert) that are sure to fail
4920 Address clauses with possibly unaligned values, or where an attempt is
4921 made to overlay a smaller variable with a larger one.
4924 Fixed-point type declarations with a null range
4927 Direct_IO or Sequential_IO instantiated with a type that has access values
4930 Variables that are never assigned a value
4933 Variables that are referenced before being initialized
4936 Task entries with no corresponding @code{accept} statement
4939 Duplicate accepts for the same task entry in a @code{select}
4942 Objects that take too much storage
4945 Unchecked conversion between types of differing sizes
4948 Missing @code{return} statement along some execution path in a function
4951 Incorrect (unrecognized) pragmas
4954 Incorrect external names
4957 Allocation from empty storage pool
4960 Potentially blocking operation in protected type
4963 Suspicious parenthesization of expressions
4966 Mismatching bounds in an aggregate
4969 Attempt to return local value by reference
4972 Premature instantiation of a generic body
4975 Attempt to pack aliased components
4978 Out of bounds array subscripts
4981 Wrong length on string assignment
4984 Violations of style rules if style checking is enabled
4987 Unused @code{with} clauses
4990 @code{Bit_Order} usage that does not have any effect
4993 @code{Standard.Duration} used to resolve universal fixed expression
4996 Dereference of possibly null value
4999 Declaration that is likely to cause storage error
5002 Internal GNAT unit @code{with}'ed by application unit
5005 Values known to be out of range at compile time
5008 Unreferenced labels and variables
5011 Address overlays that could clobber memory
5014 Unexpected initialization when address clause present
5017 Bad alignment for address clause
5020 Useless type conversions
5023 Redundant assignment statements and other redundant constructs
5026 Useless exception handlers
5029 Accidental hiding of name by child unit
5032 Access before elaboration detected at compile time
5035 A range in a @code{for} loop that is known to be null or might be null
5040 The following section lists compiler switches that are available
5041 to control the handling of warning messages. It is also possible
5042 to exercise much finer control over what warnings are issued and
5043 suppressed using the GNAT pragma Warnings, @xref{Pragma Warnings,,,
5044 gnat_rm, GNAT Reference manual}.
5049 @emph{Activate all optional errors.}
5050 @cindex @option{-gnatwa} (@command{gcc})
5051 This switch activates most optional warning messages, see remaining list
5052 in this section for details on optional warning messages that can be
5053 individually controlled. The warnings that are not turned on by this
5055 @option{-gnatwd} (implicit dereferencing),
5056 @option{-gnatwh} (hiding),
5057 @option{-gnatwl} (elaboration warnings),
5058 @option{-gnatw.o} (warn on values set by out parameters ignored)
5059 and @option{-gnatwt} (tracking of deleted conditional code).
5060 All other optional warnings are turned on.
5063 @emph{Suppress all optional errors.}
5064 @cindex @option{-gnatwA} (@command{gcc})
5065 This switch suppresses all optional warning messages, see remaining list
5066 in this section for details on optional warning messages that can be
5067 individually controlled.
5070 @emph{Activate warnings on failing assertions.}
5071 @cindex @option{-gnatw.a} (@command{gcc})
5072 @cindex Assert failures
5073 This switch activates warnings for assertions where the compiler can tell at
5074 compile time that the assertion will fail. Note that this warning is given
5075 even if assertions are disabled. The default is that such warnings are
5079 @emph{Suppress warnings on failing assertions.}
5080 @cindex @option{-gnatw.A} (@command{gcc})
5081 @cindex Assert failures
5082 This switch suppresses warnings for assertions where the compiler can tell at
5083 compile time that the assertion will fail.
5086 @emph{Activate warnings on bad fixed values.}
5087 @cindex @option{-gnatwb} (@command{gcc})
5088 @cindex Bad fixed values
5089 @cindex Fixed-point Small value
5091 This switch activates warnings for static fixed-point expressions whose
5092 value is not an exact multiple of Small. Such values are implementation
5093 dependent, since an implementation is free to choose either of the multiples
5094 that surround the value. GNAT always chooses the closer one, but this is not
5095 required behavior, and it is better to specify a value that is an exact
5096 multiple, ensuring predictable execution. The default is that such warnings
5100 @emph{Suppress warnings on bad fixed values.}
5101 @cindex @option{-gnatwB} (@command{gcc})
5102 This switch suppresses warnings for static fixed-point expressions whose
5103 value is not an exact multiple of Small.
5106 @emph{Activate warnings on biased representation.}
5107 @cindex @option{-gnatw.b} (@command{gcc})
5108 @cindex Biased representation
5109 This switch activates warnings when a size clause, value size clause, component
5110 clause, or component size clause forces the use of biased representation for an
5111 integer type (e.g. representing a range of 10..11 in a single bit by using 0/1
5112 to represent 10/11). The default is that such warnings are generated.
5115 @emph{Suppress warnings on biased representation.}
5116 @cindex @option{-gnatwB} (@command{gcc})
5117 This switch suppresses warnings for representation clauses that force the use
5118 of biased representation.
5121 @emph{Activate warnings on conditionals.}
5122 @cindex @option{-gnatwc} (@command{gcc})
5123 @cindex Conditionals, constant
5124 This switch activates warnings for conditional expressions used in
5125 tests that are known to be True or False at compile time. The default
5126 is that such warnings are not generated.
5127 Note that this warning does
5128 not get issued for the use of boolean variables or constants whose
5129 values are known at compile time, since this is a standard technique
5130 for conditional compilation in Ada, and this would generate too many
5131 false positive warnings.
5133 This warning option also activates a special test for comparisons using
5134 the operators ``>='' and`` <=''.
5135 If the compiler can tell that only the equality condition is possible,
5136 then it will warn that the ``>'' or ``<'' part of the test
5137 is useless and that the operator could be replaced by ``=''.
5138 An example would be comparing a @code{Natural} variable <= 0.
5140 This warning option also generates warnings if
5141 one or both tests is optimized away in a membership test for integer
5142 values if the result can be determined at compile time. Range tests on
5143 enumeration types are not included, since it is common for such tests
5144 to include an end point.
5146 This warning can also be turned on using @option{-gnatwa}.
5149 @emph{Suppress warnings on conditionals.}
5150 @cindex @option{-gnatwC} (@command{gcc})
5151 This switch suppresses warnings for conditional expressions used in
5152 tests that are known to be True or False at compile time.
5155 @emph{Activate warnings on missing component clauses.}
5156 @cindex @option{-gnatw.c} (@command{gcc})
5157 @cindex Component clause, missing
5158 This switch activates warnings for record components where a record
5159 representation clause is present and has component clauses for the
5160 majority, but not all, of the components. A warning is given for each
5161 component for which no component clause is present.
5163 This warning can also be turned on using @option{-gnatwa}.
5166 @emph{Suppress warnings on missing component clauses.}
5167 @cindex @option{-gnatwC} (@command{gcc})
5168 This switch suppresses warnings for record components that are
5169 missing a component clause in the situation described above.
5172 @emph{Activate warnings on implicit dereferencing.}
5173 @cindex @option{-gnatwd} (@command{gcc})
5174 If this switch is set, then the use of a prefix of an access type
5175 in an indexed component, slice, or selected component without an
5176 explicit @code{.all} will generate a warning. With this warning
5177 enabled, access checks occur only at points where an explicit
5178 @code{.all} appears in the source code (assuming no warnings are
5179 generated as a result of this switch). The default is that such
5180 warnings are not generated.
5181 Note that @option{-gnatwa} does not affect the setting of
5182 this warning option.
5185 @emph{Suppress warnings on implicit dereferencing.}
5186 @cindex @option{-gnatwD} (@command{gcc})
5187 @cindex Implicit dereferencing
5188 @cindex Dereferencing, implicit
5189 This switch suppresses warnings for implicit dereferences in
5190 indexed components, slices, and selected components.
5193 @emph{Treat warnings as errors.}
5194 @cindex @option{-gnatwe} (@command{gcc})
5195 @cindex Warnings, treat as error
5196 This switch causes warning messages to be treated as errors.
5197 The warning string still appears, but the warning messages are counted
5198 as errors, and prevent the generation of an object file.
5201 @emph{Activate every optional warning}
5202 @cindex @option{-gnatw.e} (@command{gcc})
5203 @cindex Warnings, activate every optional warning
5204 This switch activates all optional warnings, including those which
5205 are not activated by @code{-gnatwa}.
5208 @emph{Activate warnings on unreferenced formals.}
5209 @cindex @option{-gnatwf} (@command{gcc})
5210 @cindex Formals, unreferenced
5211 This switch causes a warning to be generated if a formal parameter
5212 is not referenced in the body of the subprogram. This warning can
5213 also be turned on using @option{-gnatwa} or @option{-gnatwu}. The
5214 default is that these warnings are not generated.
5217 @emph{Suppress warnings on unreferenced formals.}
5218 @cindex @option{-gnatwF} (@command{gcc})
5219 This switch suppresses warnings for unreferenced formal
5220 parameters. Note that the
5221 combination @option{-gnatwu} followed by @option{-gnatwF} has the
5222 effect of warning on unreferenced entities other than subprogram
5226 @emph{Activate warnings on unrecognized pragmas.}
5227 @cindex @option{-gnatwg} (@command{gcc})
5228 @cindex Pragmas, unrecognized
5229 This switch causes a warning to be generated if an unrecognized
5230 pragma is encountered. Apart from issuing this warning, the
5231 pragma is ignored and has no effect. This warning can
5232 also be turned on using @option{-gnatwa}. The default
5233 is that such warnings are issued (satisfying the Ada Reference
5234 Manual requirement that such warnings appear).
5237 @emph{Suppress warnings on unrecognized pragmas.}
5238 @cindex @option{-gnatwG} (@command{gcc})
5239 This switch suppresses warnings for unrecognized pragmas.
5242 @emph{Activate warnings on hiding.}
5243 @cindex @option{-gnatwh} (@command{gcc})
5244 @cindex Hiding of Declarations
5245 This switch activates warnings on hiding declarations.
5246 A declaration is considered hiding
5247 if it is for a non-overloadable entity, and it declares an entity with the
5248 same name as some other entity that is directly or use-visible. The default
5249 is that such warnings are not generated.
5250 Note that @option{-gnatwa} does not affect the setting of this warning option.
5253 @emph{Suppress warnings on hiding.}
5254 @cindex @option{-gnatwH} (@command{gcc})
5255 This switch suppresses warnings on hiding declarations.
5258 @emph{Activate warnings on implementation units.}
5259 @cindex @option{-gnatwi} (@command{gcc})
5260 This switch activates warnings for a @code{with} of an internal GNAT
5261 implementation unit, defined as any unit from the @code{Ada},
5262 @code{Interfaces}, @code{GNAT},
5263 ^^@code{DEC},^ or @code{System}
5264 hierarchies that is not
5265 documented in either the Ada Reference Manual or the GNAT
5266 Programmer's Reference Manual. Such units are intended only
5267 for internal implementation purposes and should not be @code{with}'ed
5268 by user programs. The default is that such warnings are generated
5269 This warning can also be turned on using @option{-gnatwa}.
5272 @emph{Disable warnings on implementation units.}
5273 @cindex @option{-gnatwI} (@command{gcc})
5274 This switch disables warnings for a @code{with} of an internal GNAT
5275 implementation unit.
5278 @emph{Activate warnings on overlapping actuals.}
5279 @cindex @option{-gnatw.i} (@command{gcc})
5280 This switch enables a warning on statically detectable overlapping actuals in
5281 a subprogram call, when one of the actuals is an in-out parameter, and the
5282 types of the actuals are not by-copy types. The warning is off by default,
5283 and is not included under -gnatwa.
5286 @emph{Disable warnings on overlapping actuals.}
5287 @cindex @option{-gnatw.I} (@command{gcc})
5288 This switch disables warnings on overlapping actuals in a call..
5291 @emph{Activate warnings on obsolescent features (Annex J).}
5292 @cindex @option{-gnatwj} (@command{gcc})
5293 @cindex Features, obsolescent
5294 @cindex Obsolescent features
5295 If this warning option is activated, then warnings are generated for
5296 calls to subprograms marked with @code{pragma Obsolescent} and
5297 for use of features in Annex J of the Ada Reference Manual. In the
5298 case of Annex J, not all features are flagged. In particular use
5299 of the renamed packages (like @code{Text_IO}) and use of package
5300 @code{ASCII} are not flagged, since these are very common and
5301 would generate many annoying positive warnings. The default is that
5302 such warnings are not generated. This warning is also turned on by
5303 the use of @option{-gnatwa}.
5305 In addition to the above cases, warnings are also generated for
5306 GNAT features that have been provided in past versions but which
5307 have been superseded (typically by features in the new Ada standard).
5308 For example, @code{pragma Ravenscar} will be flagged since its
5309 function is replaced by @code{pragma Profile(Ravenscar)}.
5311 Note that this warning option functions differently from the
5312 restriction @code{No_Obsolescent_Features} in two respects.
5313 First, the restriction applies only to annex J features.
5314 Second, the restriction does flag uses of package @code{ASCII}.
5317 @emph{Suppress warnings on obsolescent features (Annex J).}
5318 @cindex @option{-gnatwJ} (@command{gcc})
5319 This switch disables warnings on use of obsolescent features.
5322 @emph{Activate warnings on variables that could be constants.}
5323 @cindex @option{-gnatwk} (@command{gcc})
5324 This switch activates warnings for variables that are initialized but
5325 never modified, and then could be declared constants. The default is that
5326 such warnings are not given.
5327 This warning can also be turned on using @option{-gnatwa}.
5330 @emph{Suppress warnings on variables that could be constants.}
5331 @cindex @option{-gnatwK} (@command{gcc})
5332 This switch disables warnings on variables that could be declared constants.
5335 @emph{Activate warnings for elaboration pragmas.}
5336 @cindex @option{-gnatwl} (@command{gcc})
5337 @cindex Elaboration, warnings
5338 This switch activates warnings on missing
5339 @code{Elaborate_All} and @code{Elaborate} pragmas.
5340 See the section in this guide on elaboration checking for details on
5341 when such pragmas should be used. In dynamic elaboration mode, this switch
5342 generations warnings about the need to add elaboration pragmas. Note however,
5343 that if you blindly follow these warnings, and add @code{Elaborate_All}
5344 warnings wherever they are recommended, you basically end up with the
5345 equivalent of the static elaboration model, which may not be what you want for
5346 legacy code for which the static model does not work.
5348 For the static model, the messages generated are labeled "info:" (for
5349 information messages). They are not warnings to add elaboration pragmas,
5350 merely informational messages showing what implicit elaboration pragmas
5351 have been added, for use in analyzing elaboration circularity problems.
5353 Warnings are also generated if you
5354 are using the static mode of elaboration, and a @code{pragma Elaborate}
5355 is encountered. The default is that such warnings
5357 This warning is not automatically turned on by the use of @option{-gnatwa}.
5360 @emph{Suppress warnings for elaboration pragmas.}
5361 @cindex @option{-gnatwL} (@command{gcc})
5362 This switch suppresses warnings on missing Elaborate and Elaborate_All pragmas.
5363 See the section in this guide on elaboration checking for details on
5364 when such pragmas should be used.
5367 @emph{Activate warnings on modified but unreferenced variables.}
5368 @cindex @option{-gnatwm} (@command{gcc})
5369 This switch activates warnings for variables that are assigned (using
5370 an initialization value or with one or more assignment statements) but
5371 whose value is never read. The warning is suppressed for volatile
5372 variables and also for variables that are renamings of other variables
5373 or for which an address clause is given.
5374 This warning can also be turned on using @option{-gnatwa}.
5375 The default is that these warnings are not given.
5378 @emph{Disable warnings on modified but unreferenced variables.}
5379 @cindex @option{-gnatwM} (@command{gcc})
5380 This switch disables warnings for variables that are assigned or
5381 initialized, but never read.
5384 @emph{Activate warnings on suspicious modulus values.}
5385 @cindex @option{-gnatw.m} (@command{gcc})
5386 This switch activates warnings for modulus values that seem suspicious.
5387 The cases caught are where the size is the same as the modulus (e.g.
5388 a modulus of 7 with a size of 7 bits), and modulus values of 32 or 64
5389 with no size clause. The guess in both cases is that 2**x was intended
5390 rather than x. The default is that these warnings are given.
5393 @emph{Disable warnings on suspicious modulus values.}
5394 @cindex @option{-gnatw.M} (@command{gcc})
5395 This switch disables warnings for suspicious modulus values.
5398 @emph{Set normal warnings mode.}
5399 @cindex @option{-gnatwn} (@command{gcc})
5400 This switch sets normal warning mode, in which enabled warnings are
5401 issued and treated as warnings rather than errors. This is the default
5402 mode. the switch @option{-gnatwn} can be used to cancel the effect of
5403 an explicit @option{-gnatws} or
5404 @option{-gnatwe}. It also cancels the effect of the
5405 implicit @option{-gnatwe} that is activated by the
5406 use of @option{-gnatg}.
5409 @emph{Activate warnings on address clause overlays.}
5410 @cindex @option{-gnatwo} (@command{gcc})
5411 @cindex Address Clauses, warnings
5412 This switch activates warnings for possibly unintended initialization
5413 effects of defining address clauses that cause one variable to overlap
5414 another. The default is that such warnings are generated.
5415 This warning can also be turned on using @option{-gnatwa}.
5418 @emph{Suppress warnings on address clause overlays.}
5419 @cindex @option{-gnatwO} (@command{gcc})
5420 This switch suppresses warnings on possibly unintended initialization
5421 effects of defining address clauses that cause one variable to overlap
5425 @emph{Activate warnings on modified but unreferenced out parameters.}
5426 @cindex @option{-gnatw.o} (@command{gcc})
5427 This switch activates warnings for variables that are modified by using
5428 them as actuals for a call to a procedure with an out mode formal, where
5429 the resulting assigned value is never read. It is applicable in the case
5430 where there is more than one out mode formal. If there is only one out
5431 mode formal, the warning is issued by default (controlled by -gnatwu).
5432 The warning is suppressed for volatile
5433 variables and also for variables that are renamings of other variables
5434 or for which an address clause is given.
5435 The default is that these warnings are not given. Note that this warning
5436 is not included in -gnatwa, it must be activated explicitly.
5439 @emph{Disable warnings on modified but unreferenced out parameters.}
5440 @cindex @option{-gnatw.O} (@command{gcc})
5441 This switch suppresses warnings for variables that are modified by using
5442 them as actuals for a call to a procedure with an out mode formal, where
5443 the resulting assigned value is never read.
5446 @emph{Activate warnings on ineffective pragma Inlines.}
5447 @cindex @option{-gnatwp} (@command{gcc})
5448 @cindex Inlining, warnings
5449 This switch activates warnings for failure of front end inlining
5450 (activated by @option{-gnatN}) to inline a particular call. There are
5451 many reasons for not being able to inline a call, including most
5452 commonly that the call is too complex to inline. The default is
5453 that such warnings are not given.
5454 This warning can also be turned on using @option{-gnatwa}.
5455 Warnings on ineffective inlining by the gcc back-end can be activated
5456 separately, using the gcc switch -Winline.
5459 @emph{Suppress warnings on ineffective pragma Inlines.}
5460 @cindex @option{-gnatwP} (@command{gcc})
5461 This switch suppresses warnings on ineffective pragma Inlines. If the
5462 inlining mechanism cannot inline a call, it will simply ignore the
5466 @emph{Activate warnings on parameter ordering.}
5467 @cindex @option{-gnatw.p} (@command{gcc})
5468 @cindex Parameter order, warnings
5469 This switch activates warnings for cases of suspicious parameter
5470 ordering when the list of arguments are all simple identifiers that
5471 match the names of the formals, but are in a different order. The
5472 warning is suppressed if any use of named parameter notation is used,
5473 so this is the appropriate way to suppress a false positive (and
5474 serves to emphasize that the "misordering" is deliberate). The
5476 that such warnings are not given.
5477 This warning can also be turned on using @option{-gnatwa}.
5480 @emph{Suppress warnings on parameter ordering.}
5481 @cindex @option{-gnatw.P} (@command{gcc})
5482 This switch suppresses warnings on cases of suspicious parameter
5486 @emph{Activate warnings on questionable missing parentheses.}
5487 @cindex @option{-gnatwq} (@command{gcc})
5488 @cindex Parentheses, warnings
5489 This switch activates warnings for cases where parentheses are not used and
5490 the result is potential ambiguity from a readers point of view. For example
5491 (not a > b) when a and b are modular means ((not a) > b) and very likely the
5492 programmer intended (not (a > b)). Similarly (-x mod 5) means (-(x mod 5)) and
5493 quite likely ((-x) mod 5) was intended. In such situations it seems best to
5494 follow the rule of always parenthesizing to make the association clear, and
5495 this warning switch warns if such parentheses are not present. The default
5496 is that these warnings are given.
5497 This warning can also be turned on using @option{-gnatwa}.
5500 @emph{Suppress warnings on questionable missing parentheses.}
5501 @cindex @option{-gnatwQ} (@command{gcc})
5502 This switch suppresses warnings for cases where the association is not
5503 clear and the use of parentheses is preferred.
5506 @emph{Activate warnings on redundant constructs.}
5507 @cindex @option{-gnatwr} (@command{gcc})
5508 This switch activates warnings for redundant constructs. The following
5509 is the current list of constructs regarded as redundant:
5513 Assignment of an item to itself.
5515 Type conversion that converts an expression to its own type.
5517 Use of the attribute @code{Base} where @code{typ'Base} is the same
5520 Use of pragma @code{Pack} when all components are placed by a record
5521 representation clause.
5523 Exception handler containing only a reraise statement (raise with no
5524 operand) which has no effect.
5526 Use of the operator abs on an operand that is known at compile time
5529 Comparison of boolean expressions to an explicit True value.
5532 This warning can also be turned on using @option{-gnatwa}.
5533 The default is that warnings for redundant constructs are not given.
5536 @emph{Suppress warnings on redundant constructs.}
5537 @cindex @option{-gnatwR} (@command{gcc})
5538 This switch suppresses warnings for redundant constructs.
5541 @emph{Activate warnings for object renaming function.}
5542 @cindex @option{-gnatw.r} (@command{gcc})
5543 This switch activates warnings for an object renaming that renames a
5544 function call, which is equivalent to a constant declaration (as
5545 opposed to renaming the function itself). The default is that these
5546 warnings are given. This warning can also be turned on using
5550 @emph{Suppress warnings for object renaming function.}
5551 @cindex @option{-gnatwT} (@command{gcc})
5552 This switch suppresses warnings for object renaming function.
5555 @emph{Suppress all warnings.}
5556 @cindex @option{-gnatws} (@command{gcc})
5557 This switch completely suppresses the
5558 output of all warning messages from the GNAT front end.
5559 Note that it does not suppress warnings from the @command{gcc} back end.
5560 To suppress these back end warnings as well, use the switch @option{-w}
5561 in addition to @option{-gnatws}.
5564 @emph{Activate warnings for tracking of deleted conditional code.}
5565 @cindex @option{-gnatwt} (@command{gcc})
5566 @cindex Deactivated code, warnings
5567 @cindex Deleted code, warnings
5568 This switch activates warnings for tracking of code in conditionals (IF and
5569 CASE statements) that is detected to be dead code which cannot be executed, and
5570 which is removed by the front end. This warning is off by default, and is not
5571 turned on by @option{-gnatwa}, it has to be turned on explicitly. This may be
5572 useful for detecting deactivated code in certified applications.
5575 @emph{Suppress warnings for tracking of deleted conditional code.}
5576 @cindex @option{-gnatwT} (@command{gcc})
5577 This switch suppresses warnings for tracking of deleted conditional code.
5580 @emph{Activate warnings on unused entities.}
5581 @cindex @option{-gnatwu} (@command{gcc})
5582 This switch activates warnings to be generated for entities that
5583 are declared but not referenced, and for units that are @code{with}'ed
5585 referenced. In the case of packages, a warning is also generated if
5586 no entities in the package are referenced. This means that if the package
5587 is referenced but the only references are in @code{use}
5588 clauses or @code{renames}
5589 declarations, a warning is still generated. A warning is also generated
5590 for a generic package that is @code{with}'ed but never instantiated.
5591 In the case where a package or subprogram body is compiled, and there
5592 is a @code{with} on the corresponding spec
5593 that is only referenced in the body,
5594 a warning is also generated, noting that the
5595 @code{with} can be moved to the body. The default is that
5596 such warnings are not generated.
5597 This switch also activates warnings on unreferenced formals
5598 (it includes the effect of @option{-gnatwf}).
5599 This warning can also be turned on using @option{-gnatwa}.
5602 @emph{Suppress warnings on unused entities.}
5603 @cindex @option{-gnatwU} (@command{gcc})
5604 This switch suppresses warnings for unused entities and packages.
5605 It also turns off warnings on unreferenced formals (and thus includes
5606 the effect of @option{-gnatwF}).
5609 @emph{Activate warnings on unassigned variables.}
5610 @cindex @option{-gnatwv} (@command{gcc})
5611 @cindex Unassigned variable warnings
5612 This switch activates warnings for access to variables which
5613 may not be properly initialized. The default is that
5614 such warnings are generated.
5615 This warning can also be turned on using @option{-gnatwa}.
5618 @emph{Suppress warnings on unassigned variables.}
5619 @cindex @option{-gnatwV} (@command{gcc})
5620 This switch suppresses warnings for access to variables which
5621 may not be properly initialized.
5622 For variables of a composite type, the warning can also be suppressed in
5623 Ada 2005 by using a default initialization with a box. For example, if
5624 Table is an array of records whose components are only partially uninitialized,
5625 then the following code:
5627 @smallexample @c ada
5628 Tab : Table := (others => <>);
5631 will suppress warnings on subsequent statements that access components
5635 @emph{Activate warnings on wrong low bound assumption.}
5636 @cindex @option{-gnatww} (@command{gcc})
5637 @cindex String indexing warnings
5638 This switch activates warnings for indexing an unconstrained string parameter
5639 with a literal or S'Length. This is a case where the code is assuming that the
5640 low bound is one, which is in general not true (for example when a slice is
5641 passed). The default is that such warnings are generated.
5642 This warning can also be turned on using @option{-gnatwa}.
5645 @emph{Suppress warnings on wrong low bound assumption.}
5646 @cindex @option{-gnatwW} (@command{gcc})
5647 This switch suppresses warnings for indexing an unconstrained string parameter
5648 with a literal or S'Length. Note that this warning can also be suppressed
5649 in a particular case by adding an
5650 assertion that the lower bound is 1,
5651 as shown in the following example.
5653 @smallexample @c ada
5654 procedure K (S : String) is
5655 pragma Assert (S'First = 1);
5660 @emph{Activate warnings on unnecessary Warnings Off pragmas}
5661 @cindex @option{-gnatw.w} (@command{gcc})
5662 @cindex Warnings Off control
5663 This switch activates warnings for use of @code{pragma Warnings (Off, entity}
5664 where either the pragma is entirely useless (because it suppresses no
5665 warnings), or it could be replaced by @code{pragma Unreferenced} or
5666 @code{pragma Unmodified}.The default is that these warnings are not given.
5667 Note that this warning is not included in -gnatwa, it must be
5668 activated explicitly.
5671 @emph{Suppress warnings on unnecessary Warnings Off pragmas}
5672 @cindex @option{-gnatw.W} (@command{gcc})
5673 This switch suppresses warnings for use of @code{pragma Warnings (Off, entity}.
5676 @emph{Activate warnings on Export/Import pragmas.}
5677 @cindex @option{-gnatwx} (@command{gcc})
5678 @cindex Export/Import pragma warnings
5679 This switch activates warnings on Export/Import pragmas when
5680 the compiler detects a possible conflict between the Ada and
5681 foreign language calling sequences. For example, the use of
5682 default parameters in a convention C procedure is dubious
5683 because the C compiler cannot supply the proper default, so
5684 a warning is issued. The default is that such warnings are
5686 This warning can also be turned on using @option{-gnatwa}.
5689 @emph{Suppress warnings on Export/Import pragmas.}
5690 @cindex @option{-gnatwX} (@command{gcc})
5691 This switch suppresses warnings on Export/Import pragmas.
5692 The sense of this is that you are telling the compiler that
5693 you know what you are doing in writing the pragma, and it
5694 should not complain at you.
5697 @emph{Activate warnings for No_Exception_Propagation mode.}
5698 @cindex @option{-gnatwm} (@command{gcc})
5699 This switch activates warnings for exception usage when pragma Restrictions
5700 (No_Exception_Propagation) is in effect. Warnings are given for implicit or
5701 explicit exception raises which are not covered by a local handler, and for
5702 exception handlers which do not cover a local raise. The default is that these
5703 warnings are not given.
5706 @emph{Disable warnings for No_Exception_Propagation mode.}
5707 This switch disables warnings for exception usage when pragma Restrictions
5708 (No_Exception_Propagation) is in effect.
5711 @emph{Activate warnings for Ada 2005 compatibility issues.}
5712 @cindex @option{-gnatwy} (@command{gcc})
5713 @cindex Ada 2005 compatibility issues warnings
5714 For the most part Ada 2005 is upwards compatible with Ada 95,
5715 but there are some exceptions (for example the fact that
5716 @code{interface} is now a reserved word in Ada 2005). This
5717 switch activates several warnings to help in identifying
5718 and correcting such incompatibilities. The default is that
5719 these warnings are generated. Note that at one point Ada 2005
5720 was called Ada 0Y, hence the choice of character.
5721 This warning can also be turned on using @option{-gnatwa}.
5724 @emph{Disable warnings for Ada 2005 compatibility issues.}
5725 @cindex @option{-gnatwY} (@command{gcc})
5726 @cindex Ada 2005 compatibility issues warnings
5727 This switch suppresses several warnings intended to help in identifying
5728 incompatibilities between Ada 95 and Ada 2005.
5731 @emph{Activate warnings on unchecked conversions.}
5732 @cindex @option{-gnatwz} (@command{gcc})
5733 @cindex Unchecked_Conversion warnings
5734 This switch activates warnings for unchecked conversions
5735 where the types are known at compile time to have different
5737 is that such warnings are generated. Warnings are also
5738 generated for subprogram pointers with different conventions,
5739 and, on VMS only, for data pointers with different conventions.
5740 This warning can also be turned on using @option{-gnatwa}.
5743 @emph{Suppress warnings on unchecked conversions.}
5744 @cindex @option{-gnatwZ} (@command{gcc})
5745 This switch suppresses warnings for unchecked conversions
5746 where the types are known at compile time to have different
5747 sizes or conventions.
5749 @item ^-Wunused^WARNINGS=UNUSED^
5750 @cindex @option{-Wunused}
5751 The warnings controlled by the @option{-gnatw} switch are generated by
5752 the front end of the compiler. The @option{GCC} back end can provide
5753 additional warnings and they are controlled by the @option{-W} switch.
5754 For example, @option{^-Wunused^WARNINGS=UNUSED^} activates back end
5755 warnings for entities that are declared but not referenced.
5757 @item ^-Wuninitialized^WARNINGS=UNINITIALIZED^
5758 @cindex @option{-Wuninitialized}
5759 Similarly, @option{^-Wuninitialized^WARNINGS=UNINITIALIZED^} activates
5760 the back end warning for uninitialized variables. This switch must be
5761 used in conjunction with an optimization level greater than zero.
5763 @item ^-Wall^/ALL_BACK_END_WARNINGS^
5764 @cindex @option{-Wall}
5765 This switch enables all the above warnings from the @option{GCC} back end.
5766 The code generator detects a number of warning situations that are missed
5767 by the @option{GNAT} front end, and this switch can be used to activate them.
5768 The use of this switch also sets the default front end warning mode to
5769 @option{-gnatwa}, that is, most front end warnings activated as well.
5771 @item ^-w^/NO_BACK_END_WARNINGS^
5773 Conversely, this switch suppresses warnings from the @option{GCC} back end.
5774 The use of this switch also sets the default front end warning mode to
5775 @option{-gnatws}, that is, front end warnings suppressed as well.
5781 A string of warning parameters can be used in the same parameter. For example:
5788 will turn on all optional warnings except for elaboration pragma warnings,
5789 and also specify that warnings should be treated as errors.
5791 When no switch @option{^-gnatw^/WARNINGS^} is used, this is equivalent to:
5816 @node Debugging and Assertion Control
5817 @subsection Debugging and Assertion Control
5821 @cindex @option{-gnata} (@command{gcc})
5827 The pragmas @code{Assert} and @code{Debug} normally have no effect and
5828 are ignored. This switch, where @samp{a} stands for assert, causes
5829 @code{Assert} and @code{Debug} pragmas to be activated.
5831 The pragmas have the form:
5835 @b{pragma} Assert (@var{Boolean-expression} @r{[},
5836 @var{static-string-expression}@r{]})
5837 @b{pragma} Debug (@var{procedure call})
5842 The @code{Assert} pragma causes @var{Boolean-expression} to be tested.
5843 If the result is @code{True}, the pragma has no effect (other than
5844 possible side effects from evaluating the expression). If the result is
5845 @code{False}, the exception @code{Assert_Failure} declared in the package
5846 @code{System.Assertions} is
5847 raised (passing @var{static-string-expression}, if present, as the
5848 message associated with the exception). If no string expression is
5849 given the default is a string giving the file name and line number
5852 The @code{Debug} pragma causes @var{procedure} to be called. Note that
5853 @code{pragma Debug} may appear within a declaration sequence, allowing
5854 debugging procedures to be called between declarations.
5857 @item /DEBUG@r{[}=debug-level@r{]}
5859 Specifies how much debugging information is to be included in
5860 the resulting object file where 'debug-level' is one of the following:
5863 Include both debugger symbol records and traceback
5865 This is the default setting.
5867 Include both debugger symbol records and traceback in
5870 Excludes both debugger symbol records and traceback
5871 the object file. Same as /NODEBUG.
5873 Includes only debugger symbol records in the object
5874 file. Note that this doesn't include traceback information.
5879 @node Validity Checking
5880 @subsection Validity Checking
5881 @findex Validity Checking
5884 The Ada Reference Manual defines the concept of invalid values (see
5885 RM 13.9.1). The primary source of invalid values is uninitialized
5886 variables. A scalar variable that is left uninitialized may contain
5887 an invalid value; the concept of invalid does not apply to access or
5890 It is an error to read an invalid value, but the RM does not require
5891 run-time checks to detect such errors, except for some minimal
5892 checking to prevent erroneous execution (i.e. unpredictable
5893 behavior). This corresponds to the @option{-gnatVd} switch below,
5894 which is the default. For example, by default, if the expression of a
5895 case statement is invalid, it will raise Constraint_Error rather than
5896 causing a wild jump, and if an array index on the left-hand side of an
5897 assignment is invalid, it will raise Constraint_Error rather than
5898 overwriting an arbitrary memory location.
5900 The @option{-gnatVa} may be used to enable additional validity checks,
5901 which are not required by the RM. These checks are often very
5902 expensive (which is why the RM does not require them). These checks
5903 are useful in tracking down uninitialized variables, but they are
5904 not usually recommended for production builds.
5906 The other @option{-gnatV^@var{x}^^} switches below allow finer-grained
5907 control; you can enable whichever validity checks you desire. However,
5908 for most debugging purposes, @option{-gnatVa} is sufficient, and the
5909 default @option{-gnatVd} (i.e. standard Ada behavior) is usually
5910 sufficient for non-debugging use.
5912 The @option{-gnatB} switch tells the compiler to assume that all
5913 values are valid (that is, within their declared subtype range)
5914 except in the context of a use of the Valid attribute. This means
5915 the compiler can generate more efficient code, since the range
5916 of values is better known at compile time. However, an uninitialized
5917 variable can cause wild jumps and memory corruption in this mode.
5919 The @option{-gnatV^@var{x}^^} switch allows control over the validity
5920 checking mode as described below.
5922 The @code{x} argument is a string of letters that
5923 indicate validity checks that are performed or not performed in addition
5924 to the default checks required by Ada as described above.
5927 The options allowed for this qualifier
5928 indicate validity checks that are performed or not performed in addition
5929 to the default checks required by Ada as described above.
5935 @emph{All validity checks.}
5936 @cindex @option{-gnatVa} (@command{gcc})
5937 All validity checks are turned on.
5939 That is, @option{-gnatVa} is
5940 equivalent to @option{gnatVcdfimorst}.
5944 @emph{Validity checks for copies.}
5945 @cindex @option{-gnatVc} (@command{gcc})
5946 The right hand side of assignments, and the initializing values of
5947 object declarations are validity checked.
5950 @emph{Default (RM) validity checks.}
5951 @cindex @option{-gnatVd} (@command{gcc})
5952 Some validity checks are done by default following normal Ada semantics
5954 A check is done in case statements that the expression is within the range
5955 of the subtype. If it is not, Constraint_Error is raised.
5956 For assignments to array components, a check is done that the expression used
5957 as index is within the range. If it is not, Constraint_Error is raised.
5958 Both these validity checks may be turned off using switch @option{-gnatVD}.
5959 They are turned on by default. If @option{-gnatVD} is specified, a subsequent
5960 switch @option{-gnatVd} will leave the checks turned on.
5961 Switch @option{-gnatVD} should be used only if you are sure that all such
5962 expressions have valid values. If you use this switch and invalid values
5963 are present, then the program is erroneous, and wild jumps or memory
5964 overwriting may occur.
5967 @emph{Validity checks for elementary components.}
5968 @cindex @option{-gnatVe} (@command{gcc})
5969 In the absence of this switch, assignments to record or array components are
5970 not validity checked, even if validity checks for assignments generally
5971 (@option{-gnatVc}) are turned on. In Ada, assignment of composite values do not
5972 require valid data, but assignment of individual components does. So for
5973 example, there is a difference between copying the elements of an array with a
5974 slice assignment, compared to assigning element by element in a loop. This
5975 switch allows you to turn off validity checking for components, even when they
5976 are assigned component by component.
5979 @emph{Validity checks for floating-point values.}
5980 @cindex @option{-gnatVf} (@command{gcc})
5981 In the absence of this switch, validity checking occurs only for discrete
5982 values. If @option{-gnatVf} is specified, then validity checking also applies
5983 for floating-point values, and NaNs and infinities are considered invalid,
5984 as well as out of range values for constrained types. Note that this means
5985 that standard IEEE infinity mode is not allowed. The exact contexts
5986 in which floating-point values are checked depends on the setting of other
5987 options. For example,
5988 @option{^-gnatVif^VALIDITY_CHECKING=(IN_PARAMS,FLOATS)^} or
5989 @option{^-gnatVfi^VALIDITY_CHECKING=(FLOATS,IN_PARAMS)^}
5990 (the order does not matter) specifies that floating-point parameters of mode
5991 @code{in} should be validity checked.
5994 @emph{Validity checks for @code{in} mode parameters}
5995 @cindex @option{-gnatVi} (@command{gcc})
5996 Arguments for parameters of mode @code{in} are validity checked in function
5997 and procedure calls at the point of call.
6000 @emph{Validity checks for @code{in out} mode parameters.}
6001 @cindex @option{-gnatVm} (@command{gcc})
6002 Arguments for parameters of mode @code{in out} are validity checked in
6003 procedure calls at the point of call. The @code{'m'} here stands for
6004 modify, since this concerns parameters that can be modified by the call.
6005 Note that there is no specific option to test @code{out} parameters,
6006 but any reference within the subprogram will be tested in the usual
6007 manner, and if an invalid value is copied back, any reference to it
6008 will be subject to validity checking.
6011 @emph{No validity checks.}
6012 @cindex @option{-gnatVn} (@command{gcc})
6013 This switch turns off all validity checking, including the default checking
6014 for case statements and left hand side subscripts. Note that the use of
6015 the switch @option{-gnatp} suppresses all run-time checks, including
6016 validity checks, and thus implies @option{-gnatVn}. When this switch
6017 is used, it cancels any other @option{-gnatV} previously issued.
6020 @emph{Validity checks for operator and attribute operands.}
6021 @cindex @option{-gnatVo} (@command{gcc})
6022 Arguments for predefined operators and attributes are validity checked.
6023 This includes all operators in package @code{Standard},
6024 the shift operators defined as intrinsic in package @code{Interfaces}
6025 and operands for attributes such as @code{Pos}. Checks are also made
6026 on individual component values for composite comparisons, and on the
6027 expressions in type conversions and qualified expressions. Checks are
6028 also made on explicit ranges using @samp{..} (e.g.@: slices, loops etc).
6031 @emph{Validity checks for parameters.}
6032 @cindex @option{-gnatVp} (@command{gcc})
6033 This controls the treatment of parameters within a subprogram (as opposed
6034 to @option{-gnatVi} and @option{-gnatVm} which control validity testing
6035 of parameters on a call. If either of these call options is used, then
6036 normally an assumption is made within a subprogram that the input arguments
6037 have been validity checking at the point of call, and do not need checking
6038 again within a subprogram). If @option{-gnatVp} is set, then this assumption
6039 is not made, and parameters are not assumed to be valid, so their validity
6040 will be checked (or rechecked) within the subprogram.
6043 @emph{Validity checks for function returns.}
6044 @cindex @option{-gnatVr} (@command{gcc})
6045 The expression in @code{return} statements in functions is validity
6049 @emph{Validity checks for subscripts.}
6050 @cindex @option{-gnatVs} (@command{gcc})
6051 All subscripts expressions are checked for validity, whether they appear
6052 on the right side or left side (in default mode only left side subscripts
6053 are validity checked).
6056 @emph{Validity checks for tests.}
6057 @cindex @option{-gnatVt} (@command{gcc})
6058 Expressions used as conditions in @code{if}, @code{while} or @code{exit}
6059 statements are checked, as well as guard expressions in entry calls.
6064 The @option{-gnatV} switch may be followed by
6065 ^a string of letters^a list of options^
6066 to turn on a series of validity checking options.
6068 @option{^-gnatVcr^/VALIDITY_CHECKING=(COPIES, RETURNS)^}
6069 specifies that in addition to the default validity checking, copies and
6070 function return expressions are to be validity checked.
6071 In order to make it easier
6072 to specify the desired combination of effects,
6074 the upper case letters @code{CDFIMORST} may
6075 be used to turn off the corresponding lower case option.
6078 the prefix @code{NO} on an option turns off the corresponding validity
6081 @item @code{NOCOPIES}
6082 @item @code{NODEFAULT}
6083 @item @code{NOFLOATS}
6084 @item @code{NOIN_PARAMS}
6085 @item @code{NOMOD_PARAMS}
6086 @item @code{NOOPERANDS}
6087 @item @code{NORETURNS}
6088 @item @code{NOSUBSCRIPTS}
6089 @item @code{NOTESTS}
6093 @option{^-gnatVaM^/VALIDITY_CHECKING=(ALL, NOMOD_PARAMS)^}
6094 turns on all validity checking options except for
6095 checking of @code{@b{in out}} procedure arguments.
6097 The specification of additional validity checking generates extra code (and
6098 in the case of @option{-gnatVa} the code expansion can be substantial).
6099 However, these additional checks can be very useful in detecting
6100 uninitialized variables, incorrect use of unchecked conversion, and other
6101 errors leading to invalid values. The use of pragma @code{Initialize_Scalars}
6102 is useful in conjunction with the extra validity checking, since this
6103 ensures that wherever possible uninitialized variables have invalid values.
6105 See also the pragma @code{Validity_Checks} which allows modification of
6106 the validity checking mode at the program source level, and also allows for
6107 temporary disabling of validity checks.
6109 @node Style Checking
6110 @subsection Style Checking
6111 @findex Style checking
6114 The @option{-gnaty^x^(option,option,@dots{})^} switch
6115 @cindex @option{-gnaty} (@command{gcc})
6116 causes the compiler to
6117 enforce specified style rules. A limited set of style rules has been used
6118 in writing the GNAT sources themselves. This switch allows user programs
6119 to activate all or some of these checks. If the source program fails a
6120 specified style check, an appropriate warning message is given, preceded by
6121 the character sequence ``(style)''.
6123 @code{(option,option,@dots{})} is a sequence of keywords
6126 The string @var{x} is a sequence of letters or digits
6128 indicating the particular style
6129 checks to be performed. The following checks are defined:
6134 @emph{Specify indentation level.}
6135 If a digit from 1-9 appears
6136 ^in the string after @option{-gnaty}^as an option for /STYLE_CHECKS^
6137 then proper indentation is checked, with the digit indicating the
6138 indentation level required. A value of zero turns off this style check.
6139 The general style of required indentation is as specified by
6140 the examples in the Ada Reference Manual. Full line comments must be
6141 aligned with the @code{--} starting on a column that is a multiple of
6142 the alignment level, or they may be aligned the same way as the following
6143 non-blank line (this is useful when full line comments appear in the middle
6147 @emph{Check attribute casing.}
6148 Attribute names, including the case of keywords such as @code{digits}
6149 used as attributes names, must be written in mixed case, that is, the
6150 initial letter and any letter following an underscore must be uppercase.
6151 All other letters must be lowercase.
6153 @item ^A^ARRAY_INDEXES^
6154 @emph{Use of array index numbers in array attributes.}
6155 When using the array attributes First, Last, Range,
6156 or Length, the index number must be omitted for one-dimensional arrays
6157 and is required for multi-dimensional arrays.
6160 @emph{Blanks not allowed at statement end.}
6161 Trailing blanks are not allowed at the end of statements. The purpose of this
6162 rule, together with h (no horizontal tabs), is to enforce a canonical format
6163 for the use of blanks to separate source tokens.
6165 @item ^B^BOOLEAN_OPERATORS^
6166 @emph{Check Boolean operators.}
6167 The use of AND/OR operators is not permitted except in the cases of modular
6168 operands, array operands, and simple stand-alone boolean variables or
6169 boolean constants. In all other cases AND THEN/OR ELSE are required.
6172 @emph{Check comments.}
6173 Comments must meet the following set of rules:
6178 The ``@code{--}'' that starts the column must either start in column one,
6179 or else at least one blank must precede this sequence.
6182 Comments that follow other tokens on a line must have at least one blank
6183 following the ``@code{--}'' at the start of the comment.
6186 Full line comments must have two blanks following the ``@code{--}'' that
6187 starts the comment, with the following exceptions.
6190 A line consisting only of the ``@code{--}'' characters, possibly preceded
6191 by blanks is permitted.
6194 A comment starting with ``@code{--x}'' where @code{x} is a special character
6196 This allows proper processing of the output generated by specialized tools
6197 including @command{gnatprep} (where ``@code{--!}'' is used) and the SPARK
6199 language (where ``@code{--#}'' is used). For the purposes of this rule, a
6200 special character is defined as being in one of the ASCII ranges
6201 @code{16#21#@dots{}16#2F#} or @code{16#3A#@dots{}16#3F#}.
6202 Note that this usage is not permitted
6203 in GNAT implementation units (i.e., when @option{-gnatg} is used).
6206 A line consisting entirely of minus signs, possibly preceded by blanks, is
6207 permitted. This allows the construction of box comments where lines of minus
6208 signs are used to form the top and bottom of the box.
6211 A comment that starts and ends with ``@code{--}'' is permitted as long as at
6212 least one blank follows the initial ``@code{--}''. Together with the preceding
6213 rule, this allows the construction of box comments, as shown in the following
6216 ---------------------------
6217 -- This is a box comment --
6218 -- with two text lines. --
6219 ---------------------------
6223 @item ^d^DOS_LINE_ENDINGS^
6224 @emph{Check no DOS line terminators present.}
6225 All lines must be terminated by a single ASCII.LF
6226 character (in particular the DOS line terminator sequence CR/LF is not
6230 @emph{Check end/exit labels.}
6231 Optional labels on @code{end} statements ending subprograms and on
6232 @code{exit} statements exiting named loops, are required to be present.
6235 @emph{No form feeds or vertical tabs.}
6236 Neither form feeds nor vertical tab characters are permitted
6240 @emph{GNAT style mode}
6241 The set of style check switches is set to match that used by the GNAT sources.
6242 This may be useful when developing code that is eventually intended to be
6243 incorporated into GNAT. For further details, see GNAT sources.
6246 @emph{No horizontal tabs.}
6247 Horizontal tab characters are not permitted in the source text.
6248 Together with the b (no blanks at end of line) check, this
6249 enforces a canonical form for the use of blanks to separate
6253 @emph{Check if-then layout.}
6254 The keyword @code{then} must appear either on the same
6255 line as corresponding @code{if}, or on a line on its own, lined
6256 up under the @code{if} with at least one non-blank line in between
6257 containing all or part of the condition to be tested.
6260 @emph{check mode IN keywords}
6261 Mode @code{in} (the default mode) is not
6262 allowed to be given explicitly. @code{in out} is fine,
6263 but not @code{in} on its own.
6266 @emph{Check keyword casing.}
6267 All keywords must be in lower case (with the exception of keywords
6268 such as @code{digits} used as attribute names to which this check
6272 @emph{Check layout.}
6273 Layout of statement and declaration constructs must follow the
6274 recommendations in the Ada Reference Manual, as indicated by the
6275 form of the syntax rules. For example an @code{else} keyword must
6276 be lined up with the corresponding @code{if} keyword.
6278 There are two respects in which the style rule enforced by this check
6279 option are more liberal than those in the Ada Reference Manual. First
6280 in the case of record declarations, it is permissible to put the
6281 @code{record} keyword on the same line as the @code{type} keyword, and
6282 then the @code{end} in @code{end record} must line up under @code{type}.
6283 This is also permitted when the type declaration is split on two lines.
6284 For example, any of the following three layouts is acceptable:
6286 @smallexample @c ada
6309 Second, in the case of a block statement, a permitted alternative
6310 is to put the block label on the same line as the @code{declare} or
6311 @code{begin} keyword, and then line the @code{end} keyword up under
6312 the block label. For example both the following are permitted:
6314 @smallexample @c ada
6332 The same alternative format is allowed for loops. For example, both of
6333 the following are permitted:
6335 @smallexample @c ada
6337 Clear : while J < 10 loop
6348 @item ^Lnnn^MAX_NESTING=nnn^
6349 @emph{Set maximum nesting level}
6350 The maximum level of nesting of constructs (including subprograms, loops,
6351 blocks, packages, and conditionals) may not exceed the given value
6352 @option{nnn}. A value of zero disconnects this style check.
6354 @item ^m^LINE_LENGTH^
6355 @emph{Check maximum line length.}
6356 The length of source lines must not exceed 79 characters, including
6357 any trailing blanks. The value of 79 allows convenient display on an
6358 80 character wide device or window, allowing for possible special
6359 treatment of 80 character lines. Note that this count is of
6360 characters in the source text. This means that a tab character counts
6361 as one character in this count but a wide character sequence counts as
6362 a single character (however many bytes are needed in the encoding).
6364 @item ^Mnnn^MAX_LENGTH=nnn^
6365 @emph{Set maximum line length.}
6366 The length of lines must not exceed the
6367 given value @option{nnn}. The maximum value that can be specified is 32767.
6369 @item ^n^STANDARD_CASING^
6370 @emph{Check casing of entities in Standard.}
6371 Any identifier from Standard must be cased
6372 to match the presentation in the Ada Reference Manual (for example,
6373 @code{Integer} and @code{ASCII.NUL}).
6376 @emph{Turn off all style checks}
6377 All style check options are turned off.
6379 @item ^o^ORDERED_SUBPROGRAMS^
6380 @emph{Check order of subprogram bodies.}
6381 All subprogram bodies in a given scope
6382 (e.g.@: a package body) must be in alphabetical order. The ordering
6383 rule uses normal Ada rules for comparing strings, ignoring casing
6384 of letters, except that if there is a trailing numeric suffix, then
6385 the value of this suffix is used in the ordering (e.g.@: Junk2 comes
6388 @item ^O^OVERRIDING_INDICATORS^
6389 @emph{Check that overriding subprograms are explicitly marked as such.}
6390 The declaration of a primitive operation of a type extension that overrides
6391 an inherited operation must carry an overriding indicator.
6394 @emph{Check pragma casing.}
6395 Pragma names must be written in mixed case, that is, the
6396 initial letter and any letter following an underscore must be uppercase.
6397 All other letters must be lowercase.
6399 @item ^r^REFERENCES^
6400 @emph{Check references.}
6401 All identifier references must be cased in the same way as the
6402 corresponding declaration. No specific casing style is imposed on
6403 identifiers. The only requirement is for consistency of references
6406 @item ^S^STATEMENTS_AFTER_THEN_ELSE^
6407 @emph{Check no statements after THEN/ELSE.}
6408 No statements are allowed
6409 on the same line as a THEN or ELSE keyword following the
6410 keyword in an IF statement. OR ELSE and AND THEN are not affected,
6411 and a special exception allows a pragma to appear after ELSE.
6414 @emph{Check separate specs.}
6415 Separate declarations (``specs'') are required for subprograms (a
6416 body is not allowed to serve as its own declaration). The only
6417 exception is that parameterless library level procedures are
6418 not required to have a separate declaration. This exception covers
6419 the most frequent form of main program procedures.
6422 @emph{Check token spacing.}
6423 The following token spacing rules are enforced:
6428 The keywords @code{@b{abs}} and @code{@b{not}} must be followed by a space.
6431 The token @code{=>} must be surrounded by spaces.
6434 The token @code{<>} must be preceded by a space or a left parenthesis.
6437 Binary operators other than @code{**} must be surrounded by spaces.
6438 There is no restriction on the layout of the @code{**} binary operator.
6441 Colon must be surrounded by spaces.
6444 Colon-equal (assignment, initialization) must be surrounded by spaces.
6447 Comma must be the first non-blank character on the line, or be
6448 immediately preceded by a non-blank character, and must be followed
6452 If the token preceding a left parenthesis ends with a letter or digit, then
6453 a space must separate the two tokens.
6456 if the token following a right parenthesis starts with a letter or digit, then
6457 a space must separate the two tokens.
6460 A right parenthesis must either be the first non-blank character on
6461 a line, or it must be preceded by a non-blank character.
6464 A semicolon must not be preceded by a space, and must not be followed by
6465 a non-blank character.
6468 A unary plus or minus may not be followed by a space.
6471 A vertical bar must be surrounded by spaces.
6474 @item ^u^UNNECESSARY_BLANK_LINES^
6475 @emph{Check unnecessary blank lines.}
6476 Unnecessary blank lines are not allowed. A blank line is considered
6477 unnecessary if it appears at the end of the file, or if more than
6478 one blank line occurs in sequence.
6480 @item ^x^XTRA_PARENS^
6481 @emph{Check extra parentheses.}
6482 Unnecessary extra level of parentheses (C-style) are not allowed
6483 around conditions in @code{if} statements, @code{while} statements and
6484 @code{exit} statements.
6486 @item ^y^ALL_BUILTIN^
6487 @emph{Set all standard style check options}
6488 This is equivalent to @code{gnaty3aAbcefhiklmnprst}, that is all checking
6489 options enabled with the exception of @option{-gnatyo}, @option{-gnatyI},
6490 @option{-gnatyS}, @option{-gnatyLnnn},
6491 @option{-gnatyd}, @option{-gnatyu}, and @option{-gnatyx}.
6495 @emph{Remove style check options}
6496 This causes any subsequent options in the string to act as canceling the
6497 corresponding style check option. To cancel maximum nesting level control,
6498 use @option{L} parameter witout any integer value after that, because any
6499 digit following @option{-} in the parameter string of the @option{-gnaty}
6500 option will be threated as canceling indentation check. The same is true
6501 for @option{M} parameter. @option{y} and @option{N} parameters are not
6502 allowed after @option{-}.
6505 This causes any subsequent options in the string to enable the corresponding
6506 style check option. That is, it cancels the effect of a previous ^-^REMOVE^,
6512 @emph{Removing style check options}
6513 If the name of a style check is preceded by @option{NO} then the corresponding
6514 style check is turned off. For example @option{NOCOMMENTS} turns off style
6515 checking for comments.
6520 In the above rules, appearing in column one is always permitted, that is,
6521 counts as meeting either a requirement for a required preceding space,
6522 or as meeting a requirement for no preceding space.
6524 Appearing at the end of a line is also always permitted, that is, counts
6525 as meeting either a requirement for a following space, or as meeting
6526 a requirement for no following space.
6529 If any of these style rules is violated, a message is generated giving
6530 details on the violation. The initial characters of such messages are
6531 always ``@code{(style)}''. Note that these messages are treated as warning
6532 messages, so they normally do not prevent the generation of an object
6533 file. The @option{-gnatwe} switch can be used to treat warning messages,
6534 including style messages, as fatal errors.
6538 @option{-gnaty} on its own (that is not
6539 followed by any letters or digits), then the effect is equivalent
6540 to the use of @option{-gnatyy}, as described above, that is all
6541 built-in standard style check options are enabled.
6545 /STYLE_CHECKS=ALL_BUILTIN enables all checking options with
6546 the exception of ORDERED_SUBPROGRAMS, UNNECESSARY_BLANK_LINES,
6547 XTRA_PARENS, and DOS_LINE_ENDINGS. In addition
6557 clears any previously set style checks.
6559 @node Run-Time Checks
6560 @subsection Run-Time Checks
6561 @cindex Division by zero
6562 @cindex Access before elaboration
6563 @cindex Checks, division by zero
6564 @cindex Checks, access before elaboration
6565 @cindex Checks, stack overflow checking
6568 By default, the following checks are suppressed: integer overflow
6569 checks, stack overflow checks, and checks for access before
6570 elaboration on subprogram calls. All other checks, including range
6571 checks and array bounds checks, are turned on by default. The
6572 following @command{gcc} switches refine this default behavior.
6577 @cindex @option{-gnatp} (@command{gcc})
6578 @cindex Suppressing checks
6579 @cindex Checks, suppressing
6581 This switch causes the unit to be compiled
6582 as though @code{pragma Suppress (All_checks)}
6583 had been present in the source. Validity checks are also eliminated (in
6584 other words @option{-gnatp} also implies @option{-gnatVn}.
6585 Use this switch to improve the performance
6586 of the code at the expense of safety in the presence of invalid data or
6589 Note that when checks are suppressed, the compiler is allowed, but not
6590 required, to omit the checking code. If the run-time cost of the
6591 checking code is zero or near-zero, the compiler will generate it even
6592 if checks are suppressed. In particular, if the compiler can prove
6593 that a certain check will necessarily fail, it will generate code to
6594 do an unconditional ``raise'', even if checks are suppressed. The
6595 compiler warns in this case. Another case in which checks may not be
6596 eliminated is when they are embedded in certain run time routines such
6597 as math library routines.
6599 Of course, run-time checks are omitted whenever the compiler can prove
6600 that they will not fail, whether or not checks are suppressed.
6602 Note that if you suppress a check that would have failed, program
6603 execution is erroneous, which means the behavior is totally
6604 unpredictable. The program might crash, or print wrong answers, or
6605 do anything else. It might even do exactly what you wanted it to do
6606 (and then it might start failing mysteriously next week or next
6607 year). The compiler will generate code based on the assumption that
6608 the condition being checked is true, which can result in disaster if
6609 that assumption is wrong.
6612 @cindex @option{-gnato} (@command{gcc})
6613 @cindex Overflow checks
6614 @cindex Check, overflow
6615 Enables overflow checking for integer operations.
6616 This causes GNAT to generate slower and larger executable
6617 programs by adding code to check for overflow (resulting in raising
6618 @code{Constraint_Error} as required by standard Ada
6619 semantics). These overflow checks correspond to situations in which
6620 the true value of the result of an operation may be outside the base
6621 range of the result type. The following example shows the distinction:
6623 @smallexample @c ada
6624 X1 : Integer := "Integer'Last";
6625 X2 : Integer range 1 .. 5 := "5";
6626 X3 : Integer := "Integer'Last";
6627 X4 : Integer range 1 .. 5 := "5";
6628 F : Float := "2.0E+20";
6637 Note that if explicit values are assigned at compile time, the
6638 compiler may be able to detect overflow at compile time, in which case
6639 no actual run-time checking code is required, and Constraint_Error
6640 will be raised unconditionally, with or without
6641 @option{-gnato}. That's why the assigned values in the above fragment
6642 are in quotes, the meaning is "assign a value not known to the
6643 compiler that happens to be equal to ...". The remaining discussion
6644 assumes that the compiler cannot detect the values at compile time.
6646 Here the first addition results in a value that is outside the base range
6647 of Integer, and hence requires an overflow check for detection of the
6648 constraint error. Thus the first assignment to @code{X1} raises a
6649 @code{Constraint_Error} exception only if @option{-gnato} is set.
6651 The second increment operation results in a violation of the explicit
6652 range constraint; such range checks are performed by default, and are
6653 unaffected by @option{-gnato}.
6655 The two conversions of @code{F} both result in values that are outside
6656 the base range of type @code{Integer} and thus will raise
6657 @code{Constraint_Error} exceptions only if @option{-gnato} is used.
6658 The fact that the result of the second conversion is assigned to
6659 variable @code{X4} with a restricted range is irrelevant, since the problem
6660 is in the conversion, not the assignment.
6662 Basically the rule is that in the default mode (@option{-gnato} not
6663 used), the generated code assures that all integer variables stay
6664 within their declared ranges, or within the base range if there is
6665 no declared range. This prevents any serious problems like indexes
6666 out of range for array operations.
6668 What is not checked in default mode is an overflow that results in
6669 an in-range, but incorrect value. In the above example, the assignments
6670 to @code{X1}, @code{X2}, @code{X3} all give results that are within the
6671 range of the target variable, but the result is wrong in the sense that
6672 it is too large to be represented correctly. Typically the assignment
6673 to @code{X1} will result in wrap around to the largest negative number.
6674 The conversions of @code{F} will result in some @code{Integer} value
6675 and if that integer value is out of the @code{X4} range then the
6676 subsequent assignment would generate an exception.
6678 @findex Machine_Overflows
6679 Note that the @option{-gnato} switch does not affect the code generated
6680 for any floating-point operations; it applies only to integer
6682 For floating-point, GNAT has the @code{Machine_Overflows}
6683 attribute set to @code{False} and the normal mode of operation is to
6684 generate IEEE NaN and infinite values on overflow or invalid operations
6685 (such as dividing 0.0 by 0.0).
6687 The reason that we distinguish overflow checking from other kinds of
6688 range constraint checking is that a failure of an overflow check, unlike
6689 for example the failure of a range check, can result in an incorrect
6690 value, but cannot cause random memory destruction (like an out of range
6691 subscript), or a wild jump (from an out of range case value). Overflow
6692 checking is also quite expensive in time and space, since in general it
6693 requires the use of double length arithmetic.
6695 Note again that @option{-gnato} is off by default, so overflow checking is
6696 not performed in default mode. This means that out of the box, with the
6697 default settings, GNAT does not do all the checks expected from the
6698 language description in the Ada Reference Manual. If you want all constraint
6699 checks to be performed, as described in this Manual, then you must
6700 explicitly use the -gnato switch either on the @command{gnatmake} or
6701 @command{gcc} command.
6704 @cindex @option{-gnatE} (@command{gcc})
6705 @cindex Elaboration checks
6706 @cindex Check, elaboration
6707 Enables dynamic checks for access-before-elaboration
6708 on subprogram calls and generic instantiations.
6709 Note that @option{-gnatE} is not necessary for safety, because in the
6710 default mode, GNAT ensures statically that the checks would not fail.
6711 For full details of the effect and use of this switch,
6712 @xref{Compiling Using gcc}.
6715 @cindex @option{-fstack-check} (@command{gcc})
6716 @cindex Stack Overflow Checking
6717 @cindex Checks, stack overflow checking
6718 Activates stack overflow checking. For full details of the effect and use of
6719 this switch see @ref{Stack Overflow Checking}.
6724 The setting of these switches only controls the default setting of the
6725 checks. You may modify them using either @code{Suppress} (to remove
6726 checks) or @code{Unsuppress} (to add back suppressed checks) pragmas in
6729 @node Using gcc for Syntax Checking
6730 @subsection Using @command{gcc} for Syntax Checking
6733 @cindex @option{-gnats} (@command{gcc})
6737 The @code{s} stands for ``syntax''.
6740 Run GNAT in syntax checking only mode. For
6741 example, the command
6744 $ gcc -c -gnats x.adb
6748 compiles file @file{x.adb} in syntax-check-only mode. You can check a
6749 series of files in a single command
6751 , and can use wild cards to specify such a group of files.
6752 Note that you must specify the @option{-c} (compile
6753 only) flag in addition to the @option{-gnats} flag.
6756 You may use other switches in conjunction with @option{-gnats}. In
6757 particular, @option{-gnatl} and @option{-gnatv} are useful to control the
6758 format of any generated error messages.
6760 When the source file is empty or contains only empty lines and/or comments,
6761 the output is a warning:
6764 $ gcc -c -gnats -x ada toto.txt
6765 toto.txt:1:01: warning: empty file, contains no compilation units
6769 Otherwise, the output is simply the error messages, if any. No object file or
6770 ALI file is generated by a syntax-only compilation. Also, no units other
6771 than the one specified are accessed. For example, if a unit @code{X}
6772 @code{with}'s a unit @code{Y}, compiling unit @code{X} in syntax
6773 check only mode does not access the source file containing unit
6776 @cindex Multiple units, syntax checking
6777 Normally, GNAT allows only a single unit in a source file. However, this
6778 restriction does not apply in syntax-check-only mode, and it is possible
6779 to check a file containing multiple compilation units concatenated
6780 together. This is primarily used by the @code{gnatchop} utility
6781 (@pxref{Renaming Files Using gnatchop}).
6784 @node Using gcc for Semantic Checking
6785 @subsection Using @command{gcc} for Semantic Checking
6788 @cindex @option{-gnatc} (@command{gcc})
6792 The @code{c} stands for ``check''.
6794 Causes the compiler to operate in semantic check mode,
6795 with full checking for all illegalities specified in the
6796 Ada Reference Manual, but without generation of any object code
6797 (no object file is generated).
6799 Because dependent files must be accessed, you must follow the GNAT
6800 semantic restrictions on file structuring to operate in this mode:
6804 The needed source files must be accessible
6805 (@pxref{Search Paths and the Run-Time Library (RTL)}).
6808 Each file must contain only one compilation unit.
6811 The file name and unit name must match (@pxref{File Naming Rules}).
6814 The output consists of error messages as appropriate. No object file is
6815 generated. An @file{ALI} file is generated for use in the context of
6816 cross-reference tools, but this file is marked as not being suitable
6817 for binding (since no object file is generated).
6818 The checking corresponds exactly to the notion of
6819 legality in the Ada Reference Manual.
6821 Any unit can be compiled in semantics-checking-only mode, including
6822 units that would not normally be compiled (subunits,
6823 and specifications where a separate body is present).
6826 @node Compiling Different Versions of Ada
6827 @subsection Compiling Different Versions of Ada
6830 The switches described in this section allow you to explicitly specify
6831 the version of the Ada language that your programs are written in.
6832 By default @value{EDITION} assumes @value{DEFAULTLANGUAGEVERSION},
6833 but you can also specify @value{NONDEFAULTLANGUAGEVERSION} or
6834 indicate Ada 83 compatibility mode.
6837 @cindex Compatibility with Ada 83
6839 @item -gnat83 (Ada 83 Compatibility Mode)
6840 @cindex @option{-gnat83} (@command{gcc})
6841 @cindex ACVC, Ada 83 tests
6845 Although GNAT is primarily an Ada 95 / Ada 2005 compiler, this switch
6846 specifies that the program is to be compiled in Ada 83 mode. With
6847 @option{-gnat83}, GNAT rejects most post-Ada 83 extensions and applies Ada 83
6848 semantics where this can be done easily.
6849 It is not possible to guarantee this switch does a perfect
6850 job; some subtle tests, such as are
6851 found in earlier ACVC tests (and that have been removed from the ACATS suite
6852 for Ada 95), might not compile correctly.
6853 Nevertheless, this switch may be useful in some circumstances, for example
6854 where, due to contractual reasons, existing code needs to be maintained
6855 using only Ada 83 features.
6857 With few exceptions (most notably the need to use @code{<>} on
6858 @cindex Generic formal parameters
6859 unconstrained generic formal parameters, the use of the new Ada 95 / Ada 2005
6860 reserved words, and the use of packages
6861 with optional bodies), it is not necessary to specify the
6862 @option{-gnat83} switch when compiling Ada 83 programs, because, with rare
6863 exceptions, Ada 95 and Ada 2005 are upwardly compatible with Ada 83. Thus
6864 a correct Ada 83 program is usually also a correct program
6865 in these later versions of the language standard.
6866 For further information, please refer to @ref{Compatibility and Porting Guide}.
6868 @item -gnat95 (Ada 95 mode)
6869 @cindex @option{-gnat95} (@command{gcc})
6873 This switch directs the compiler to implement the Ada 95 version of the
6875 Since Ada 95 is almost completely upwards
6876 compatible with Ada 83, Ada 83 programs may generally be compiled using
6877 this switch (see the description of the @option{-gnat83} switch for further
6878 information about Ada 83 mode).
6879 If an Ada 2005 program is compiled in Ada 95 mode,
6880 uses of the new Ada 2005 features will cause error
6881 messages or warnings.
6883 This switch also can be used to cancel the effect of a previous
6884 @option{-gnat83} or @option{-gnat05} switch earlier in the command line.
6886 @item -gnat05 (Ada 2005 mode)
6887 @cindex @option{-gnat05} (@command{gcc})
6888 @cindex Ada 2005 mode
6891 This switch directs the compiler to implement the Ada 2005 version of the
6893 Since Ada 2005 is almost completely upwards
6894 compatible with Ada 95 (and thus also with Ada 83), Ada 83 and Ada 95 programs
6895 may generally be compiled using this switch (see the description of the
6896 @option{-gnat83} and @option{-gnat95} switches for further
6899 For information about the approved ``Ada Issues'' that have been incorporated
6900 into Ada 2005, see @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs}.
6901 Included with GNAT releases is a file @file{features-ada0y} that describes
6902 the set of implemented Ada 2005 features.
6906 @node Character Set Control
6907 @subsection Character Set Control
6909 @item ^-gnati^/IDENTIFIER_CHARACTER_SET=^@var{c}
6910 @cindex @option{^-gnati^/IDENTIFIER_CHARACTER_SET^} (@command{gcc})
6913 Normally GNAT recognizes the Latin-1 character set in source program
6914 identifiers, as described in the Ada Reference Manual.
6916 GNAT to recognize alternate character sets in identifiers. @var{c} is a
6917 single character ^^or word^ indicating the character set, as follows:
6921 ISO 8859-1 (Latin-1) identifiers
6924 ISO 8859-2 (Latin-2) letters allowed in identifiers
6927 ISO 8859-3 (Latin-3) letters allowed in identifiers
6930 ISO 8859-4 (Latin-4) letters allowed in identifiers
6933 ISO 8859-5 (Cyrillic) letters allowed in identifiers
6936 ISO 8859-15 (Latin-9) letters allowed in identifiers
6939 IBM PC letters (code page 437) allowed in identifiers
6942 IBM PC letters (code page 850) allowed in identifiers
6944 @item ^f^FULL_UPPER^
6945 Full upper-half codes allowed in identifiers
6948 No upper-half codes allowed in identifiers
6951 Wide-character codes (that is, codes greater than 255)
6952 allowed in identifiers
6955 @xref{Foreign Language Representation}, for full details on the
6956 implementation of these character sets.
6958 @item ^-gnatW^/WIDE_CHARACTER_ENCODING=^@var{e}
6959 @cindex @option{^-gnatW^/WIDE_CHARACTER_ENCODING^} (@command{gcc})
6960 Specify the method of encoding for wide characters.
6961 @var{e} is one of the following:
6966 Hex encoding (brackets coding also recognized)
6969 Upper half encoding (brackets encoding also recognized)
6972 Shift/JIS encoding (brackets encoding also recognized)
6975 EUC encoding (brackets encoding also recognized)
6978 UTF-8 encoding (brackets encoding also recognized)
6981 Brackets encoding only (default value)
6983 For full details on these encoding
6984 methods see @ref{Wide Character Encodings}.
6985 Note that brackets coding is always accepted, even if one of the other
6986 options is specified, so for example @option{-gnatW8} specifies that both
6987 brackets and UTF-8 encodings will be recognized. The units that are
6988 with'ed directly or indirectly will be scanned using the specified
6989 representation scheme, and so if one of the non-brackets scheme is
6990 used, it must be used consistently throughout the program. However,
6991 since brackets encoding is always recognized, it may be conveniently
6992 used in standard libraries, allowing these libraries to be used with
6993 any of the available coding schemes.
6996 If no @option{-gnatW?} parameter is present, then the default
6997 representation is normally Brackets encoding only. However, if the
6998 first three characters of the file are 16#EF# 16#BB# 16#BF# (the standard
6999 byte order mark or BOM for UTF-8), then these three characters are
7000 skipped and the default representation for the file is set to UTF-8.
7002 Note that the wide character representation that is specified (explicitly
7003 or by default) for the main program also acts as the default encoding used
7004 for Wide_Text_IO files if not specifically overridden by a WCEM form
7008 @node File Naming Control
7009 @subsection File Naming Control
7012 @item ^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{n}
7013 @cindex @option{-gnatk} (@command{gcc})
7014 Activates file name ``krunching''. @var{n}, a decimal integer in the range
7015 1-999, indicates the maximum allowable length of a file name (not
7016 including the @file{.ads} or @file{.adb} extension). The default is not
7017 to enable file name krunching.
7019 For the source file naming rules, @xref{File Naming Rules}.
7022 @node Subprogram Inlining Control
7023 @subsection Subprogram Inlining Control
7028 @cindex @option{-gnatn} (@command{gcc})
7030 The @code{n} here is intended to suggest the first syllable of the
7033 GNAT recognizes and processes @code{Inline} pragmas. However, for the
7034 inlining to actually occur, optimization must be enabled. To enable
7035 inlining of subprograms specified by pragma @code{Inline},
7036 you must also specify this switch.
7037 In the absence of this switch, GNAT does not attempt
7038 inlining and does not need to access the bodies of
7039 subprograms for which @code{pragma Inline} is specified if they are not
7040 in the current unit.
7042 If you specify this switch the compiler will access these bodies,
7043 creating an extra source dependency for the resulting object file, and
7044 where possible, the call will be inlined.
7045 For further details on when inlining is possible
7046 see @ref{Inlining of Subprograms}.
7049 @cindex @option{-gnatN} (@command{gcc})
7050 This switch activates front-end inlining which also
7051 generates additional dependencies.
7053 When using a gcc-based back end (in practice this means using any version
7054 of GNAT other than the JGNAT, .NET or GNAAMP versions), then the use of
7055 @option{-gnatN} is deprecated, and the use of @option{-gnatn} is preferred.
7056 Historically front end inlining was more extensive than the gcc back end
7057 inlining, but that is no longer the case.
7060 @node Auxiliary Output Control
7061 @subsection Auxiliary Output Control
7065 @cindex @option{-gnatt} (@command{gcc})
7066 @cindex Writing internal trees
7067 @cindex Internal trees, writing to file
7068 Causes GNAT to write the internal tree for a unit to a file (with the
7069 extension @file{.adt}.
7070 This not normally required, but is used by separate analysis tools.
7072 these tools do the necessary compilations automatically, so you should
7073 not have to specify this switch in normal operation.
7074 Note that the combination of switches @option{-gnatct}
7075 generates a tree in the form required by ASIS applications.
7078 @cindex @option{-gnatu} (@command{gcc})
7079 Print a list of units required by this compilation on @file{stdout}.
7080 The listing includes all units on which the unit being compiled depends
7081 either directly or indirectly.
7084 @item -pass-exit-codes
7085 @cindex @option{-pass-exit-codes} (@command{gcc})
7086 If this switch is not used, the exit code returned by @command{gcc} when
7087 compiling multiple files indicates whether all source files have
7088 been successfully used to generate object files or not.
7090 When @option{-pass-exit-codes} is used, @command{gcc} exits with an extended
7091 exit status and allows an integrated development environment to better
7092 react to a compilation failure. Those exit status are:
7096 There was an error in at least one source file.
7098 At least one source file did not generate an object file.
7100 The compiler died unexpectedly (internal error for example).
7102 An object file has been generated for every source file.
7107 @node Debugging Control
7108 @subsection Debugging Control
7112 @cindex Debugging options
7115 @cindex @option{-gnatd} (@command{gcc})
7116 Activate internal debugging switches. @var{x} is a letter or digit, or
7117 string of letters or digits, which specifies the type of debugging
7118 outputs desired. Normally these are used only for internal development
7119 or system debugging purposes. You can find full documentation for these
7120 switches in the body of the @code{Debug} unit in the compiler source
7121 file @file{debug.adb}.
7125 @cindex @option{-gnatG} (@command{gcc})
7126 This switch causes the compiler to generate auxiliary output containing
7127 a pseudo-source listing of the generated expanded code. Like most Ada
7128 compilers, GNAT works by first transforming the high level Ada code into
7129 lower level constructs. For example, tasking operations are transformed
7130 into calls to the tasking run-time routines. A unique capability of GNAT
7131 is to list this expanded code in a form very close to normal Ada source.
7132 This is very useful in understanding the implications of various Ada
7133 usage on the efficiency of the generated code. There are many cases in
7134 Ada (e.g.@: the use of controlled types), where simple Ada statements can
7135 generate a lot of run-time code. By using @option{-gnatG} you can identify
7136 these cases, and consider whether it may be desirable to modify the coding
7137 approach to improve efficiency.
7139 The optional parameter @code{nn} if present after -gnatG specifies an
7140 alternative maximum line length that overrides the normal default of 72.
7141 This value is in the range 40-999999, values less than 40 being silently
7142 reset to 40. The equal sign is optional.
7144 The format of the output is very similar to standard Ada source, and is
7145 easily understood by an Ada programmer. The following special syntactic
7146 additions correspond to low level features used in the generated code that
7147 do not have any exact analogies in pure Ada source form. The following
7148 is a partial list of these special constructions. See the spec
7149 of package @code{Sprint} in file @file{sprint.ads} for a full list.
7151 If the switch @option{-gnatL} is used in conjunction with
7152 @cindex @option{-gnatL} (@command{gcc})
7153 @option{-gnatG}, then the original source lines are interspersed
7154 in the expanded source (as comment lines with the original line number).
7157 @item new @var{xxx} @r{[}storage_pool = @var{yyy}@r{]}
7158 Shows the storage pool being used for an allocator.
7160 @item at end @var{procedure-name};
7161 Shows the finalization (cleanup) procedure for a scope.
7163 @item (if @var{expr} then @var{expr} else @var{expr})
7164 Conditional expression equivalent to the @code{x?y:z} construction in C.
7166 @item @var{target}^^^(@var{source})
7167 A conversion with floating-point truncation instead of rounding.
7169 @item @var{target}?(@var{source})
7170 A conversion that bypasses normal Ada semantic checking. In particular
7171 enumeration types and fixed-point types are treated simply as integers.
7173 @item @var{target}?^^^(@var{source})
7174 Combines the above two cases.
7176 @item @var{x} #/ @var{y}
7177 @itemx @var{x} #mod @var{y}
7178 @itemx @var{x} #* @var{y}
7179 @itemx @var{x} #rem @var{y}
7180 A division or multiplication of fixed-point values which are treated as
7181 integers without any kind of scaling.
7183 @item free @var{expr} @r{[}storage_pool = @var{xxx}@r{]}
7184 Shows the storage pool associated with a @code{free} statement.
7186 @item [subtype or type declaration]
7187 Used to list an equivalent declaration for an internally generated
7188 type that is referenced elsewhere in the listing.
7190 @item freeze @var{type-name} @ovar{actions}
7191 Shows the point at which @var{type-name} is frozen, with possible
7192 associated actions to be performed at the freeze point.
7194 @item reference @var{itype}
7195 Reference (and hence definition) to internal type @var{itype}.
7197 @item @var{function-name}! (@var{arg}, @var{arg}, @var{arg})
7198 Intrinsic function call.
7200 @item @var{label-name} : label
7201 Declaration of label @var{labelname}.
7203 @item #$ @var{subprogram-name}
7204 An implicit call to a run-time support routine
7205 (to meet the requirement of H.3.1(9) in a
7208 @item @var{expr} && @var{expr} && @var{expr} @dots{} && @var{expr}
7209 A multiple concatenation (same effect as @var{expr} & @var{expr} &
7210 @var{expr}, but handled more efficiently).
7212 @item [constraint_error]
7213 Raise the @code{Constraint_Error} exception.
7215 @item @var{expression}'reference
7216 A pointer to the result of evaluating @var{expression}.
7218 @item @var{target-type}!(@var{source-expression})
7219 An unchecked conversion of @var{source-expression} to @var{target-type}.
7221 @item [@var{numerator}/@var{denominator}]
7222 Used to represent internal real literals (that) have no exact
7223 representation in base 2-16 (for example, the result of compile time
7224 evaluation of the expression 1.0/27.0).
7228 @cindex @option{-gnatD} (@command{gcc})
7229 When used in conjunction with @option{-gnatG}, this switch causes
7230 the expanded source, as described above for
7231 @option{-gnatG} to be written to files with names
7232 @file{^xxx.dg^XXX_DG^}, where @file{xxx} is the normal file name,
7233 instead of to the standard output file. For
7234 example, if the source file name is @file{hello.adb}, then a file
7235 @file{^hello.adb.dg^HELLO.ADB_DG^} will be written. The debugging
7236 information generated by the @command{gcc} @option{^-g^/DEBUG^} switch
7237 will refer to the generated @file{^xxx.dg^XXX_DG^} file. This allows
7238 you to do source level debugging using the generated code which is
7239 sometimes useful for complex code, for example to find out exactly
7240 which part of a complex construction raised an exception. This switch
7241 also suppress generation of cross-reference information (see
7242 @option{-gnatx}) since otherwise the cross-reference information
7243 would refer to the @file{^.dg^.DG^} file, which would cause
7244 confusion since this is not the original source file.
7246 Note that @option{-gnatD} actually implies @option{-gnatG}
7247 automatically, so it is not necessary to give both options.
7248 In other words @option{-gnatD} is equivalent to @option{-gnatDG}).
7250 If the switch @option{-gnatL} is used in conjunction with
7251 @cindex @option{-gnatL} (@command{gcc})
7252 @option{-gnatDG}, then the original source lines are interspersed
7253 in the expanded source (as comment lines with the original line number).
7255 The optional parameter @code{nn} if present after -gnatD specifies an
7256 alternative maximum line length that overrides the normal default of 72.
7257 This value is in the range 40-999999, values less than 40 being silently
7258 reset to 40. The equal sign is optional.
7261 @cindex @option{-gnatr} (@command{gcc})
7262 @cindex pragma Restrictions
7263 This switch causes pragma Restrictions to be treated as Restriction_Warnings
7264 so that violation of restrictions causes warnings rather than illegalities.
7265 This is useful during the development process when new restrictions are added
7266 or investigated. The switch also causes pragma Profile to be treated as
7267 Profile_Warnings, and pragma Restricted_Run_Time and pragma Ravenscar set
7268 restriction warnings rather than restrictions.
7271 @item -gnatR@r{[}0@r{|}1@r{|}2@r{|}3@r{[}s@r{]]}
7272 @cindex @option{-gnatR} (@command{gcc})
7273 This switch controls output from the compiler of a listing showing
7274 representation information for declared types and objects. For
7275 @option{-gnatR0}, no information is output (equivalent to omitting
7276 the @option{-gnatR} switch). For @option{-gnatR1} (which is the default,
7277 so @option{-gnatR} with no parameter has the same effect), size and alignment
7278 information is listed for declared array and record types. For
7279 @option{-gnatR2}, size and alignment information is listed for all
7280 declared types and objects. Finally @option{-gnatR3} includes symbolic
7281 expressions for values that are computed at run time for
7282 variant records. These symbolic expressions have a mostly obvious
7283 format with #n being used to represent the value of the n'th
7284 discriminant. See source files @file{repinfo.ads/adb} in the
7285 @code{GNAT} sources for full details on the format of @option{-gnatR3}
7286 output. If the switch is followed by an s (e.g.@: @option{-gnatR2s}), then
7287 the output is to a file with the name @file{^file.rep^file_REP^} where
7288 file is the name of the corresponding source file.
7291 @item /REPRESENTATION_INFO
7292 @cindex @option{/REPRESENTATION_INFO} (@command{gcc})
7293 This qualifier controls output from the compiler of a listing showing
7294 representation information for declared types and objects. For
7295 @option{/REPRESENTATION_INFO=NONE}, no information is output
7296 (equivalent to omitting the @option{/REPRESENTATION_INFO} qualifier).
7297 @option{/REPRESENTATION_INFO} without option is equivalent to
7298 @option{/REPRESENTATION_INFO=ARRAYS}.
7299 For @option{/REPRESENTATION_INFO=ARRAYS}, size and alignment
7300 information is listed for declared array and record types. For
7301 @option{/REPRESENTATION_INFO=OBJECTS}, size and alignment information
7302 is listed for all expression information for values that are computed
7303 at run time for variant records. These symbolic expressions have a mostly
7304 obvious format with #n being used to represent the value of the n'th
7305 discriminant. See source files @file{REPINFO.ADS/ADB} in the
7306 @code{GNAT} sources for full details on the format of
7307 @option{/REPRESENTATION_INFO=SYMBOLIC} output.
7308 If _FILE is added at the end of an option
7309 (e.g.@: @option{/REPRESENTATION_INFO=ARRAYS_FILE}),
7310 then the output is to a file with the name @file{file_REP} where
7311 file is the name of the corresponding source file.
7313 Note that it is possible for record components to have zero size. In
7314 this case, the component clause uses an obvious extension of permitted
7315 Ada syntax, for example @code{at 0 range 0 .. -1}.
7317 Representation information requires that code be generated (since it is the
7318 code generator that lays out complex data structures). If an attempt is made
7319 to output representation information when no code is generated, for example
7320 when a subunit is compiled on its own, then no information can be generated
7321 and the compiler outputs a message to this effect.
7324 @cindex @option{-gnatS} (@command{gcc})
7325 The use of the switch @option{-gnatS} for an
7326 Ada compilation will cause the compiler to output a
7327 representation of package Standard in a form very
7328 close to standard Ada. It is not quite possible to
7329 do this entirely in standard Ada (since new
7330 numeric base types cannot be created in standard
7331 Ada), but the output is easily
7332 readable to any Ada programmer, and is useful to
7333 determine the characteristics of target dependent
7334 types in package Standard.
7337 @cindex @option{-gnatx} (@command{gcc})
7338 Normally the compiler generates full cross-referencing information in
7339 the @file{ALI} file. This information is used by a number of tools,
7340 including @code{gnatfind} and @code{gnatxref}. The @option{-gnatx} switch
7341 suppresses this information. This saves some space and may slightly
7342 speed up compilation, but means that these tools cannot be used.
7345 @node Exception Handling Control
7346 @subsection Exception Handling Control
7349 GNAT uses two methods for handling exceptions at run-time. The
7350 @code{setjmp/longjmp} method saves the context when entering
7351 a frame with an exception handler. Then when an exception is
7352 raised, the context can be restored immediately, without the
7353 need for tracing stack frames. This method provides very fast
7354 exception propagation, but introduces significant overhead for
7355 the use of exception handlers, even if no exception is raised.
7357 The other approach is called ``zero cost'' exception handling.
7358 With this method, the compiler builds static tables to describe
7359 the exception ranges. No dynamic code is required when entering
7360 a frame containing an exception handler. When an exception is
7361 raised, the tables are used to control a back trace of the
7362 subprogram invocation stack to locate the required exception
7363 handler. This method has considerably poorer performance for
7364 the propagation of exceptions, but there is no overhead for
7365 exception handlers if no exception is raised. Note that in this
7366 mode and in the context of mixed Ada and C/C++ programming,
7367 to propagate an exception through a C/C++ code, the C/C++ code
7368 must be compiled with the @option{-funwind-tables} GCC's
7371 The following switches may be used to control which of the
7372 two exception handling methods is used.
7378 @cindex @option{--RTS=sjlj} (@command{gnatmake})
7379 This switch causes the setjmp/longjmp run-time (when available) to be used
7380 for exception handling. If the default
7381 mechanism for the target is zero cost exceptions, then
7382 this switch can be used to modify this default, and must be
7383 used for all units in the partition.
7384 This option is rarely used. One case in which it may be
7385 advantageous is if you have an application where exception
7386 raising is common and the overall performance of the
7387 application is improved by favoring exception propagation.
7390 @cindex @option{--RTS=zcx} (@command{gnatmake})
7391 @cindex Zero Cost Exceptions
7392 This switch causes the zero cost approach to be used
7393 for exception handling. If this is the default mechanism for the
7394 target (see below), then this switch is unneeded. If the default
7395 mechanism for the target is setjmp/longjmp exceptions, then
7396 this switch can be used to modify this default, and must be
7397 used for all units in the partition.
7398 This option can only be used if the zero cost approach
7399 is available for the target in use, otherwise it will generate an error.
7403 The same option @option{--RTS} must be used both for @command{gcc}
7404 and @command{gnatbind}. Passing this option to @command{gnatmake}
7405 (@pxref{Switches for gnatmake}) will ensure the required consistency
7406 through the compilation and binding steps.
7408 @node Units to Sources Mapping Files
7409 @subsection Units to Sources Mapping Files
7413 @item -gnatem=@var{path}
7414 @cindex @option{-gnatem} (@command{gcc})
7415 A mapping file is a way to communicate to the compiler two mappings:
7416 from unit names to file names (without any directory information) and from
7417 file names to path names (with full directory information). These mappings
7418 are used by the compiler to short-circuit the path search.
7420 The use of mapping files is not required for correct operation of the
7421 compiler, but mapping files can improve efficiency, particularly when
7422 sources are read over a slow network connection. In normal operation,
7423 you need not be concerned with the format or use of mapping files,
7424 and the @option{-gnatem} switch is not a switch that you would use
7425 explicitly. It is intended primarily for use by automatic tools such as
7426 @command{gnatmake} running under the project file facility. The
7427 description here of the format of mapping files is provided
7428 for completeness and for possible use by other tools.
7430 A mapping file is a sequence of sets of three lines. In each set, the
7431 first line is the unit name, in lower case, with @code{%s} appended
7432 for specs and @code{%b} appended for bodies; the second line is the
7433 file name; and the third line is the path name.
7439 /gnat/project1/sources/main.2.ada
7442 When the switch @option{-gnatem} is specified, the compiler will
7443 create in memory the two mappings from the specified file. If there is
7444 any problem (nonexistent file, truncated file or duplicate entries),
7445 no mapping will be created.
7447 Several @option{-gnatem} switches may be specified; however, only the
7448 last one on the command line will be taken into account.
7450 When using a project file, @command{gnatmake} creates a temporary
7451 mapping file and communicates it to the compiler using this switch.
7455 @node Integrated Preprocessing
7456 @subsection Integrated Preprocessing
7459 GNAT sources may be preprocessed immediately before compilation.
7460 In this case, the actual
7461 text of the source is not the text of the source file, but is derived from it
7462 through a process called preprocessing. Integrated preprocessing is specified
7463 through switches @option{-gnatep} and/or @option{-gnateD}. @option{-gnatep}
7464 indicates, through a text file, the preprocessing data to be used.
7465 @option{-gnateD} specifies or modifies the values of preprocessing symbol.
7468 Note that when integrated preprocessing is used, the output from the
7469 preprocessor is not written to any external file. Instead it is passed
7470 internally to the compiler. If you need to preserve the result of
7471 preprocessing in a file, then you should use @command{gnatprep}
7472 to perform the desired preprocessing in stand-alone mode.
7475 It is recommended that @command{gnatmake} switch ^-s^/SWITCH_CHECK^ should be
7476 used when Integrated Preprocessing is used. The reason is that preprocessing
7477 with another Preprocessing Data file without changing the sources will
7478 not trigger recompilation without this switch.
7481 Note that @command{gnatmake} switch ^-m^/MINIMAL_RECOMPILATION^ will almost
7482 always trigger recompilation for sources that are preprocessed,
7483 because @command{gnatmake} cannot compute the checksum of the source after
7487 The actual preprocessing function is described in details in section
7488 @ref{Preprocessing Using gnatprep}. This section only describes how integrated
7489 preprocessing is triggered and parameterized.
7493 @item -gnatep=@var{file}
7494 @cindex @option{-gnatep} (@command{gcc})
7495 This switch indicates to the compiler the file name (without directory
7496 information) of the preprocessor data file to use. The preprocessor data file
7497 should be found in the source directories.
7500 A preprocessing data file is a text file with significant lines indicating
7501 how should be preprocessed either a specific source or all sources not
7502 mentioned in other lines. A significant line is a nonempty, non-comment line.
7503 Comments are similar to Ada comments.
7506 Each significant line starts with either a literal string or the character '*'.
7507 A literal string is the file name (without directory information) of the source
7508 to preprocess. A character '*' indicates the preprocessing for all the sources
7509 that are not specified explicitly on other lines (order of the lines is not
7510 significant). It is an error to have two lines with the same file name or two
7511 lines starting with the character '*'.
7514 After the file name or the character '*', another optional literal string
7515 indicating the file name of the definition file to be used for preprocessing
7516 (@pxref{Form of Definitions File}). The definition files are found by the
7517 compiler in one of the source directories. In some cases, when compiling
7518 a source in a directory other than the current directory, if the definition
7519 file is in the current directory, it may be necessary to add the current
7520 directory as a source directory through switch ^-I.^/SEARCH=[]^, otherwise
7521 the compiler would not find the definition file.
7524 Then, optionally, ^switches^switches^ similar to those of @code{gnatprep} may
7525 be found. Those ^switches^switches^ are:
7530 Causes both preprocessor lines and the lines deleted by
7531 preprocessing to be replaced by blank lines, preserving the line number.
7532 This ^switch^switch^ is always implied; however, if specified after @option{-c}
7533 it cancels the effect of @option{-c}.
7536 Causes both preprocessor lines and the lines deleted
7537 by preprocessing to be retained as comments marked
7538 with the special string ``@code{--! }''.
7540 @item -Dsymbol=value
7541 Define or redefine a symbol, associated with value. A symbol is an Ada
7542 identifier, or an Ada reserved word, with the exception of @code{if},
7543 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7544 @code{value} is either a literal string, an Ada identifier or any Ada reserved
7545 word. A symbol declared with this ^switch^switch^ replaces a symbol with the
7546 same name defined in a definition file.
7549 Causes a sorted list of symbol names and values to be
7550 listed on the standard output file.
7553 Causes undefined symbols to be treated as having the value @code{FALSE}
7555 of a preprocessor test. In the absence of this option, an undefined symbol in
7556 a @code{#if} or @code{#elsif} test will be treated as an error.
7561 Examples of valid lines in a preprocessor data file:
7564 "toto.adb" "prep.def" -u
7565 -- preprocess "toto.adb", using definition file "prep.def",
7566 -- undefined symbol are False.
7569 -- preprocess all other sources without a definition file;
7570 -- suppressed lined are commented; symbol VERSION has the value V101.
7572 "titi.adb" "prep2.def" -s
7573 -- preprocess "titi.adb", using definition file "prep2.def";
7574 -- list all symbols with their values.
7577 @item ^-gnateD^/DATA_PREPROCESSING=^symbol@r{[}=value@r{]}
7578 @cindex @option{-gnateD} (@command{gcc})
7579 Define or redefine a preprocessing symbol, associated with value. If no value
7580 is given on the command line, then the value of the symbol is @code{True}.
7581 A symbol is an identifier, following normal Ada (case-insensitive)
7582 rules for its syntax, and value is any sequence (including an empty sequence)
7583 of characters from the set (letters, digits, period, underline).
7584 Ada reserved words may be used as symbols, with the exceptions of @code{if},
7585 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
7588 A symbol declared with this ^switch^switch^ on the command line replaces a
7589 symbol with the same name either in a definition file or specified with a
7590 ^switch^switch^ -D in the preprocessor data file.
7593 This switch is similar to switch @option{^-D^/ASSOCIATE^} of @code{gnatprep}.
7596 When integrated preprocessing is performed and the preprocessor modifies
7597 the source text, write the result of this preprocessing into a file
7598 <source>^.prep^_prep^.
7602 @node Code Generation Control
7603 @subsection Code Generation Control
7607 The GCC technology provides a wide range of target dependent
7608 @option{-m} switches for controlling
7609 details of code generation with respect to different versions of
7610 architectures. This includes variations in instruction sets (e.g.@:
7611 different members of the power pc family), and different requirements
7612 for optimal arrangement of instructions (e.g.@: different members of
7613 the x86 family). The list of available @option{-m} switches may be
7614 found in the GCC documentation.
7616 Use of these @option{-m} switches may in some cases result in improved
7619 The GNAT Pro technology is tested and qualified without any
7620 @option{-m} switches,
7621 so generally the most reliable approach is to avoid the use of these
7622 switches. However, we generally expect most of these switches to work
7623 successfully with GNAT Pro, and many customers have reported successful
7624 use of these options.
7626 Our general advice is to avoid the use of @option{-m} switches unless
7627 special needs lead to requirements in this area. In particular,
7628 there is no point in using @option{-m} switches to improve performance
7629 unless you actually see a performance improvement.
7633 @subsection Return Codes
7634 @cindex Return Codes
7635 @cindex @option{/RETURN_CODES=VMS}
7638 On VMS, GNAT compiled programs return POSIX-style codes by default,
7639 e.g.@: @option{/RETURN_CODES=POSIX}.
7641 To enable VMS style return codes, use GNAT BIND and LINK with the option
7642 @option{/RETURN_CODES=VMS}. For example:
7645 GNAT BIND MYMAIN.ALI /RETURN_CODES=VMS
7646 GNAT LINK MYMAIN.ALI /RETURN_CODES=VMS
7650 Programs built with /RETURN_CODES=VMS are suitable to be called in
7651 VMS DCL scripts. Programs compiled with the default /RETURN_CODES=POSIX
7652 are suitable for spawning with appropriate GNAT RTL routines.
7656 @node Search Paths and the Run-Time Library (RTL)
7657 @section Search Paths and the Run-Time Library (RTL)
7660 With the GNAT source-based library system, the compiler must be able to
7661 find source files for units that are needed by the unit being compiled.
7662 Search paths are used to guide this process.
7664 The compiler compiles one source file whose name must be given
7665 explicitly on the command line. In other words, no searching is done
7666 for this file. To find all other source files that are needed (the most
7667 common being the specs of units), the compiler examines the following
7668 directories, in the following order:
7672 The directory containing the source file of the main unit being compiled
7673 (the file name on the command line).
7676 Each directory named by an @option{^-I^/SOURCE_SEARCH^} switch given on the
7677 @command{gcc} command line, in the order given.
7680 @findex ADA_PRJ_INCLUDE_FILE
7681 Each of the directories listed in the text file whose name is given
7682 by the @env{ADA_PRJ_INCLUDE_FILE} ^environment variable^logical name^.
7685 @env{ADA_PRJ_INCLUDE_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
7686 driver when project files are used. It should not normally be set
7690 @findex ADA_INCLUDE_PATH
7691 Each of the directories listed in the value of the
7692 @env{ADA_INCLUDE_PATH} ^environment variable^logical name^.
7694 Construct this value
7695 exactly as the @env{PATH} environment variable: a list of directory
7696 names separated by colons (semicolons when working with the NT version).
7699 Normally, define this value as a logical name containing a comma separated
7700 list of directory names.
7702 This variable can also be defined by means of an environment string
7703 (an argument to the HP C exec* set of functions).
7707 DEFINE ANOTHER_PATH FOO:[BAG]
7708 DEFINE ADA_INCLUDE_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
7711 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
7712 first, followed by the standard Ada
7713 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADAINCLUDE].
7714 If this is not redefined, the user will obtain the HP Ada 83 IO packages
7715 (Text_IO, Sequential_IO, etc)
7716 instead of the standard Ada packages. Thus, in order to get the standard Ada
7717 packages by default, ADA_INCLUDE_PATH must be redefined.
7721 The content of the @file{ada_source_path} file which is part of the GNAT
7722 installation tree and is used to store standard libraries such as the
7723 GNAT Run Time Library (RTL) source files.
7725 @ref{Installing a library}
7730 Specifying the switch @option{^-I-^/NOCURRENT_DIRECTORY^}
7731 inhibits the use of the directory
7732 containing the source file named in the command line. You can still
7733 have this directory on your search path, but in this case it must be
7734 explicitly requested with a @option{^-I^/SOURCE_SEARCH^} switch.
7736 Specifying the switch @option{-nostdinc}
7737 inhibits the search of the default location for the GNAT Run Time
7738 Library (RTL) source files.
7740 The compiler outputs its object files and ALI files in the current
7743 Caution: The object file can be redirected with the @option{-o} switch;
7744 however, @command{gcc} and @code{gnat1} have not been coordinated on this
7745 so the @file{ALI} file will not go to the right place. Therefore, you should
7746 avoid using the @option{-o} switch.
7750 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
7751 children make up the GNAT RTL, together with the simple @code{System.IO}
7752 package used in the @code{"Hello World"} example. The sources for these units
7753 are needed by the compiler and are kept together in one directory. Not
7754 all of the bodies are needed, but all of the sources are kept together
7755 anyway. In a normal installation, you need not specify these directory
7756 names when compiling or binding. Either the environment variables or
7757 the built-in defaults cause these files to be found.
7759 In addition to the language-defined hierarchies (@code{System}, @code{Ada} and
7760 @code{Interfaces}), the GNAT distribution provides a fourth hierarchy,
7761 consisting of child units of @code{GNAT}. This is a collection of generally
7762 useful types, subprograms, etc. @xref{Top, GNAT Reference Manual, About
7763 This Guid, gnat_rm, GNAT Reference Manual}, for further details.
7765 Besides simplifying access to the RTL, a major use of search paths is
7766 in compiling sources from multiple directories. This can make
7767 development environments much more flexible.
7769 @node Order of Compilation Issues
7770 @section Order of Compilation Issues
7773 If, in our earlier example, there was a spec for the @code{hello}
7774 procedure, it would be contained in the file @file{hello.ads}; yet this
7775 file would not have to be explicitly compiled. This is the result of the
7776 model we chose to implement library management. Some of the consequences
7777 of this model are as follows:
7781 There is no point in compiling specs (except for package
7782 specs with no bodies) because these are compiled as needed by clients. If
7783 you attempt a useless compilation, you will receive an error message.
7784 It is also useless to compile subunits because they are compiled as needed
7788 There are no order of compilation requirements: performing a
7789 compilation never obsoletes anything. The only way you can obsolete
7790 something and require recompilations is to modify one of the
7791 source files on which it depends.
7794 There is no library as such, apart from the ALI files
7795 (@pxref{The Ada Library Information Files}, for information on the format
7796 of these files). For now we find it convenient to create separate ALI files,
7797 but eventually the information therein may be incorporated into the object
7801 When you compile a unit, the source files for the specs of all units
7802 that it @code{with}'s, all its subunits, and the bodies of any generics it
7803 instantiates must be available (reachable by the search-paths mechanism
7804 described above), or you will receive a fatal error message.
7811 The following are some typical Ada compilation command line examples:
7814 @item $ gcc -c xyz.adb
7815 Compile body in file @file{xyz.adb} with all default options.
7818 @item $ gcc -c -O2 -gnata xyz-def.adb
7821 @item $ GNAT COMPILE /OPTIMIZE=ALL -gnata xyz-def.adb
7824 Compile the child unit package in file @file{xyz-def.adb} with extensive
7825 optimizations, and pragma @code{Assert}/@code{Debug} statements
7828 @item $ gcc -c -gnatc abc-def.adb
7829 Compile the subunit in file @file{abc-def.adb} in semantic-checking-only
7833 @node Binding Using gnatbind
7834 @chapter Binding Using @code{gnatbind}
7838 * Running gnatbind::
7839 * Switches for gnatbind::
7840 * Command-Line Access::
7841 * Search Paths for gnatbind::
7842 * Examples of gnatbind Usage::
7846 This chapter describes the GNAT binder, @code{gnatbind}, which is used
7847 to bind compiled GNAT objects.
7849 Note: to invoke @code{gnatbind} with a project file, use the @code{gnat}
7850 driver (see @ref{The GNAT Driver and Project Files}).
7852 The @code{gnatbind} program performs four separate functions:
7856 Checks that a program is consistent, in accordance with the rules in
7857 Chapter 10 of the Ada Reference Manual. In particular, error
7858 messages are generated if a program uses inconsistent versions of a
7862 Checks that an acceptable order of elaboration exists for the program
7863 and issues an error message if it cannot find an order of elaboration
7864 that satisfies the rules in Chapter 10 of the Ada Language Manual.
7867 Generates a main program incorporating the given elaboration order.
7868 This program is a small Ada package (body and spec) that
7869 must be subsequently compiled
7870 using the GNAT compiler. The necessary compilation step is usually
7871 performed automatically by @command{gnatlink}. The two most important
7872 functions of this program
7873 are to call the elaboration routines of units in an appropriate order
7874 and to call the main program.
7877 Determines the set of object files required by the given main program.
7878 This information is output in the forms of comments in the generated program,
7879 to be read by the @command{gnatlink} utility used to link the Ada application.
7882 @node Running gnatbind
7883 @section Running @code{gnatbind}
7886 The form of the @code{gnatbind} command is
7889 $ gnatbind @ovar{switches} @var{mainprog}@r{[}.ali@r{]} @ovar{switches}
7893 where @file{@var{mainprog}.adb} is the Ada file containing the main program
7894 unit body. If no switches are specified, @code{gnatbind} constructs an Ada
7895 package in two files whose names are
7896 @file{b~@var{mainprog}.ads}, and @file{b~@var{mainprog}.adb}.
7897 For example, if given the
7898 parameter @file{hello.ali}, for a main program contained in file
7899 @file{hello.adb}, the binder output files would be @file{b~hello.ads}
7900 and @file{b~hello.adb}.
7902 When doing consistency checking, the binder takes into consideration
7903 any source files it can locate. For example, if the binder determines
7904 that the given main program requires the package @code{Pack}, whose
7906 file is @file{pack.ali} and whose corresponding source spec file is
7907 @file{pack.ads}, it attempts to locate the source file @file{pack.ads}
7908 (using the same search path conventions as previously described for the
7909 @command{gcc} command). If it can locate this source file, it checks that
7911 or source checksums of the source and its references to in @file{ALI} files
7912 match. In other words, any @file{ALI} files that mentions this spec must have
7913 resulted from compiling this version of the source file (or in the case
7914 where the source checksums match, a version close enough that the
7915 difference does not matter).
7917 @cindex Source files, use by binder
7918 The effect of this consistency checking, which includes source files, is
7919 that the binder ensures that the program is consistent with the latest
7920 version of the source files that can be located at bind time. Editing a
7921 source file without compiling files that depend on the source file cause
7922 error messages to be generated by the binder.
7924 For example, suppose you have a main program @file{hello.adb} and a
7925 package @code{P}, from file @file{p.ads} and you perform the following
7930 Enter @code{gcc -c hello.adb} to compile the main program.
7933 Enter @code{gcc -c p.ads} to compile package @code{P}.
7936 Edit file @file{p.ads}.
7939 Enter @code{gnatbind hello}.
7943 At this point, the file @file{p.ali} contains an out-of-date time stamp
7944 because the file @file{p.ads} has been edited. The attempt at binding
7945 fails, and the binder generates the following error messages:
7948 error: "hello.adb" must be recompiled ("p.ads" has been modified)
7949 error: "p.ads" has been modified and must be recompiled
7953 Now both files must be recompiled as indicated, and then the bind can
7954 succeed, generating a main program. You need not normally be concerned
7955 with the contents of this file, but for reference purposes a sample
7956 binder output file is given in @ref{Example of Binder Output File}.
7958 In most normal usage, the default mode of @command{gnatbind} which is to
7959 generate the main package in Ada, as described in the previous section.
7960 In particular, this means that any Ada programmer can read and understand
7961 the generated main program. It can also be debugged just like any other
7962 Ada code provided the @option{^-g^/DEBUG^} switch is used for
7963 @command{gnatbind} and @command{gnatlink}.
7965 However for some purposes it may be convenient to generate the main
7966 program in C rather than Ada. This may for example be helpful when you
7967 are generating a mixed language program with the main program in C. The
7968 GNAT compiler itself is an example.
7969 The use of the @option{^-C^/BIND_FILE=C^} switch
7970 for both @code{gnatbind} and @command{gnatlink} will cause the program to
7971 be generated in C (and compiled using the gnu C compiler).
7973 @node Switches for gnatbind
7974 @section Switches for @command{gnatbind}
7977 The following switches are available with @code{gnatbind}; details will
7978 be presented in subsequent sections.
7981 * Consistency-Checking Modes::
7982 * Binder Error Message Control::
7983 * Elaboration Control::
7985 * Binding with Non-Ada Main Programs::
7986 * Binding Programs with No Main Subprogram::
7993 @cindex @option{--version} @command{gnatbind}
7994 Display Copyright and version, then exit disregarding all other options.
7997 @cindex @option{--help} @command{gnatbind}
7998 If @option{--version} was not used, display usage, then exit disregarding
8002 @cindex @option{-a} @command{gnatbind}
8003 Indicates that, if supported by the platform, the adainit procedure should
8004 be treated as an initialisation routine by the linker (a constructor). This
8005 is intended to be used by the Project Manager to automatically initialize
8006 shared Stand-Alone Libraries.
8008 @item ^-aO^/OBJECT_SEARCH^
8009 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatbind})
8010 Specify directory to be searched for ALI files.
8012 @item ^-aI^/SOURCE_SEARCH^
8013 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
8014 Specify directory to be searched for source file.
8016 @item ^-A^/BIND_FILE=ADA^
8017 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatbind})
8018 Generate binder program in Ada (default)
8020 @item ^-b^/REPORT_ERRORS=BRIEF^
8021 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@command{gnatbind})
8022 Generate brief messages to @file{stderr} even if verbose mode set.
8024 @item ^-c^/NOOUTPUT^
8025 @cindex @option{^-c^/NOOUTPUT^} (@command{gnatbind})
8026 Check only, no generation of binder output file.
8028 @item ^-C^/BIND_FILE=C^
8029 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatbind})
8030 Generate binder program in C
8032 @item ^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
8033 @cindex @option{^-d^/DEFAULT_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}} (@command{gnatbind})
8034 This switch can be used to change the default task stack size value
8035 to a specified size @var{nn}, which is expressed in bytes by default, or
8036 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
8038 In the absence of a @samp{@r{[}k@r{|}m@r{]}} suffix, this switch is equivalent,
8039 in effect, to completing all task specs with
8040 @smallexample @c ada
8041 pragma Storage_Size (nn);
8043 When they do not already have such a pragma.
8045 @item ^-D^/DEFAULT_SECONDARY_STACK_SIZE=^@var{nn}@r{[}k@r{|}m@r{]}
8046 @cindex @option{^-D^/DEFAULT_SECONDARY_STACK_SIZE=nnnnn^} (@command{gnatbind})
8047 This switch can be used to change the default secondary stack size value
8048 to a specified size @var{nn}, which is expressed in bytes by default, or
8049 in kilobytes when suffixed with @var{k} or in megabytes when suffixed
8052 The secondary stack is used to deal with functions that return a variable
8053 sized result, for example a function returning an unconstrained
8054 String. There are two ways in which this secondary stack is allocated.
8056 For most targets, the secondary stack is growing on demand and is allocated
8057 as a chain of blocks in the heap. The -D option is not very
8058 relevant. It only give some control over the size of the allocated
8059 blocks (whose size is the minimum of the default secondary stack size value,
8060 and the actual size needed for the current allocation request).
8062 For certain targets, notably VxWorks 653,
8063 the secondary stack is allocated by carving off a fixed ratio chunk of the
8064 primary task stack. The -D option is used to define the
8065 size of the environment task's secondary stack.
8067 @item ^-e^/ELABORATION_DEPENDENCIES^
8068 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@command{gnatbind})
8069 Output complete list of elaboration-order dependencies.
8071 @item ^-E^/STORE_TRACEBACKS^
8072 @cindex @option{^-E^/STORE_TRACEBACKS^} (@command{gnatbind})
8073 Store tracebacks in exception occurrences when the target supports it.
8074 This is the default with the zero cost exception mechanism.
8076 @c The following may get moved to an appendix
8077 This option is currently supported on the following targets:
8078 all x86 ports, Solaris, Windows, HP-UX, AIX, PowerPC VxWorks and Alpha VxWorks.
8080 See also the packages @code{GNAT.Traceback} and
8081 @code{GNAT.Traceback.Symbolic} for more information.
8083 Note that on x86 ports, you must not use @option{-fomit-frame-pointer}
8084 @command{gcc} option.
8087 @item ^-F^/FORCE_ELABS_FLAGS^
8088 @cindex @option{^-F^/FORCE_ELABS_FLAGS^} (@command{gnatbind})
8089 Force the checks of elaboration flags. @command{gnatbind} does not normally
8090 generate checks of elaboration flags for the main executable, except when
8091 a Stand-Alone Library is used. However, there are cases when this cannot be
8092 detected by gnatbind. An example is importing an interface of a Stand-Alone
8093 Library through a pragma Import and only specifying through a linker switch
8094 this Stand-Alone Library. This switch is used to guarantee that elaboration
8095 flag checks are generated.
8098 @cindex @option{^-h^/HELP^} (@command{gnatbind})
8099 Output usage (help) information
8102 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8103 Specify directory to be searched for source and ALI files.
8105 @item ^-I-^/NOCURRENT_DIRECTORY^
8106 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatbind})
8107 Do not look for sources in the current directory where @code{gnatbind} was
8108 invoked, and do not look for ALI files in the directory containing the
8109 ALI file named in the @code{gnatbind} command line.
8111 @item ^-l^/ORDER_OF_ELABORATION^
8112 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@command{gnatbind})
8113 Output chosen elaboration order.
8115 @item ^-L@var{xxx}^/BUILD_LIBRARY=@var{xxx}^
8116 @cindex @option{^-L^/BUILD_LIBRARY^} (@command{gnatbind})
8117 Bind the units for library building. In this case the adainit and
8118 adafinal procedures (@pxref{Binding with Non-Ada Main Programs})
8119 are renamed to ^@var{xxx}init^@var{XXX}INIT^ and
8120 ^@var{xxx}final^@var{XXX}FINAL^.
8121 Implies ^-n^/NOCOMPILE^.
8123 (@xref{GNAT and Libraries}, for more details.)
8126 On OpenVMS, these init and final procedures are exported in uppercase
8127 letters. For example if /BUILD_LIBRARY=toto is used, the exported name of
8128 the init procedure will be "TOTOINIT" and the exported name of the final
8129 procedure will be "TOTOFINAL".
8132 @item ^-Mxyz^/RENAME_MAIN=xyz^
8133 @cindex @option{^-M^/RENAME_MAIN^} (@command{gnatbind})
8134 Rename generated main program from main to xyz. This option is
8135 supported on cross environments only.
8137 @item ^-m^/ERROR_LIMIT=^@var{n}
8138 @cindex @option{^-m^/ERROR_LIMIT^} (@command{gnatbind})
8139 Limit number of detected errors or warnings to @var{n}, where @var{n} is
8140 in the range 1..999999. The default value if no switch is
8141 given is 9999. If the number of warnings reaches this limit, then a
8142 message is output and further warnings are suppressed, the bind
8143 continues in this case. If the number of errors reaches this
8144 limit, then a message is output and the bind is abandoned.
8145 A value of zero means that no limit is enforced. The equal
8149 Furthermore, under Windows, the sources pointed to by the libraries path
8150 set in the registry are not searched for.
8154 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8158 @cindex @option{-nostdinc} (@command{gnatbind})
8159 Do not look for sources in the system default directory.
8162 @cindex @option{-nostdlib} (@command{gnatbind})
8163 Do not look for library files in the system default directory.
8165 @item --RTS=@var{rts-path}
8166 @cindex @option{--RTS} (@code{gnatbind})
8167 Specifies the default location of the runtime library. Same meaning as the
8168 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
8170 @item ^-o ^/OUTPUT=^@var{file}
8171 @cindex @option{^-o ^/OUTPUT^} (@command{gnatbind})
8172 Name the output file @var{file} (default is @file{b~@var{xxx}.adb}).
8173 Note that if this option is used, then linking must be done manually,
8174 gnatlink cannot be used.
8176 @item ^-O^/OBJECT_LIST^
8177 @cindex @option{^-O^/OBJECT_LIST^} (@command{gnatbind})
8180 @item ^-p^/PESSIMISTIC_ELABORATION^
8181 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@command{gnatbind})
8182 Pessimistic (worst-case) elaboration order
8185 @cindex @option{^-R^-R^} (@command{gnatbind})
8186 Output closure source list.
8188 @item ^-s^/READ_SOURCES=ALL^
8189 @cindex @option{^-s^/READ_SOURCES=ALL^} (@command{gnatbind})
8190 Require all source files to be present.
8192 @item ^-S@var{xxx}^/INITIALIZE_SCALARS=@var{xxx}^
8193 @cindex @option{^-S^/INITIALIZE_SCALARS^} (@command{gnatbind})
8194 Specifies the value to be used when detecting uninitialized scalar
8195 objects with pragma Initialize_Scalars.
8196 The @var{xxx} ^string specified with the switch^option^ may be either
8198 @item ``@option{^in^INVALID^}'' requesting an invalid value where possible
8199 @item ``@option{^lo^LOW^}'' for the lowest possible value
8200 @item ``@option{^hi^HIGH^}'' for the highest possible value
8201 @item ``@option{@var{xx}}'' for a value consisting of repeated bytes with the
8202 value @code{16#@var{xx}#} (i.e., @var{xx} is a string of two hexadecimal digits).
8205 In addition, you can specify @option{-Sev} to indicate that the value is
8206 to be set at run time. In this case, the program will look for an environment
8207 @cindex GNAT_INIT_SCALARS
8208 variable of the form @env{GNAT_INIT_SCALARS=@var{xx}}, where @var{xx} is one
8209 of @option{in/lo/hi/@var{xx}} with the same meanings as above.
8210 If no environment variable is found, or if it does not have a valid value,
8211 then the default is @option{in} (invalid values).
8215 @cindex @option{-static} (@code{gnatbind})
8216 Link against a static GNAT run time.
8219 @cindex @option{-shared} (@code{gnatbind})
8220 Link against a shared GNAT run time when available.
8223 @item ^-t^/NOTIME_STAMP_CHECK^
8224 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8225 Tolerate time stamp and other consistency errors
8227 @item ^-T@var{n}^/TIME_SLICE=@var{n}^
8228 @cindex @option{^-T^/TIME_SLICE^} (@code{gnatbind})
8229 Set the time slice value to @var{n} milliseconds. If the system supports
8230 the specification of a specific time slice value, then the indicated value
8231 is used. If the system does not support specific time slice values, but
8232 does support some general notion of round-robin scheduling, then any
8233 nonzero value will activate round-robin scheduling.
8235 A value of zero is treated specially. It turns off time
8236 slicing, and in addition, indicates to the tasking run time that the
8237 semantics should match as closely as possible the Annex D
8238 requirements of the Ada RM, and in particular sets the default
8239 scheduling policy to @code{FIFO_Within_Priorities}.
8241 @item ^-u@var{n}^/DYNAMIC_STACK_USAGE=@var{n}^
8242 @cindex @option{^-u^/DYNAMIC_STACK_USAGE^} (@code{gnatbind})
8243 Enable dynamic stack usage, with @var{n} results stored and displayed
8244 at program termination. A result is generated when a task
8245 terminates. Results that can't be stored are displayed on the fly, at
8246 task termination. This option is currently not supported on Itanium
8247 platforms. (See @ref{Dynamic Stack Usage Analysis} for details.)
8249 @item ^-v^/REPORT_ERRORS=VERBOSE^
8250 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8251 Verbose mode. Write error messages, header, summary output to
8256 @cindex @option{-w} (@code{gnatbind})
8257 Warning mode (@var{x}=s/e for suppress/treat as error)
8261 @item /WARNINGS=NORMAL
8262 @cindex @option{/WARNINGS} (@code{gnatbind})
8263 Normal warnings mode. Warnings are issued but ignored
8265 @item /WARNINGS=SUPPRESS
8266 @cindex @option{/WARNINGS} (@code{gnatbind})
8267 All warning messages are suppressed
8269 @item /WARNINGS=ERROR
8270 @cindex @option{/WARNINGS} (@code{gnatbind})
8271 Warning messages are treated as fatal errors
8274 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8275 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8276 Override default wide character encoding for standard Text_IO files.
8278 @item ^-x^/READ_SOURCES=NONE^
8279 @cindex @option{^-x^/READ_SOURCES^} (@code{gnatbind})
8280 Exclude source files (check object consistency only).
8283 @item /READ_SOURCES=AVAILABLE
8284 @cindex @option{/READ_SOURCES} (@code{gnatbind})
8285 Default mode, in which sources are checked for consistency only if
8289 @item ^-y^/ENABLE_LEAP_SECONDS^
8290 @cindex @option{^-y^/ENABLE_LEAP_SECONDS^} (@code{gnatbind})
8291 Enable leap seconds support in @code{Ada.Calendar} and its children.
8293 @item ^-z^/ZERO_MAIN^
8294 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8300 You may obtain this listing of switches by running @code{gnatbind} with
8304 @node Consistency-Checking Modes
8305 @subsection Consistency-Checking Modes
8308 As described earlier, by default @code{gnatbind} checks
8309 that object files are consistent with one another and are consistent
8310 with any source files it can locate. The following switches control binder
8315 @item ^-s^/READ_SOURCES=ALL^
8316 @cindex @option{^-s^/READ_SOURCES=ALL^} (@code{gnatbind})
8317 Require source files to be present. In this mode, the binder must be
8318 able to locate all source files that are referenced, in order to check
8319 their consistency. In normal mode, if a source file cannot be located it
8320 is simply ignored. If you specify this switch, a missing source
8323 @item ^-Wx^/WIDE_CHARACTER_ENCODING=^@var{e}
8324 @cindex @option{^-Wx^/WIDE_CHARACTER_ENCODING^} (@code{gnatbind})
8325 Override default wide character encoding for standard Text_IO files.
8326 Normally the default wide character encoding method used for standard
8327 [Wide_[Wide_]]Text_IO files is taken from the encoding specified for
8328 the main source input (see description of switch
8329 @option{^-gnatWx^/WIDE_CHARACTER_ENCODING^} for the compiler). The
8330 use of this switch for the binder (which has the same set of
8331 possible arguments) overrides this default as specified.
8333 @item ^-x^/READ_SOURCES=NONE^
8334 @cindex @option{^-x^/READ_SOURCES=NONE^} (@code{gnatbind})
8335 Exclude source files. In this mode, the binder only checks that ALI
8336 files are consistent with one another. Source files are not accessed.
8337 The binder runs faster in this mode, and there is still a guarantee that
8338 the resulting program is self-consistent.
8339 If a source file has been edited since it was last compiled, and you
8340 specify this switch, the binder will not detect that the object
8341 file is out of date with respect to the source file. Note that this is the
8342 mode that is automatically used by @command{gnatmake} because in this
8343 case the checking against sources has already been performed by
8344 @command{gnatmake} in the course of compilation (i.e.@: before binding).
8347 @item /READ_SOURCES=AVAILABLE
8348 @cindex @code{/READ_SOURCES=AVAILABLE} (@code{gnatbind})
8349 This is the default mode in which source files are checked if they are
8350 available, and ignored if they are not available.
8354 @node Binder Error Message Control
8355 @subsection Binder Error Message Control
8358 The following switches provide control over the generation of error
8359 messages from the binder:
8363 @item ^-v^/REPORT_ERRORS=VERBOSE^
8364 @cindex @option{^-v^/REPORT_ERRORS=VERBOSE^} (@code{gnatbind})
8365 Verbose mode. In the normal mode, brief error messages are generated to
8366 @file{stderr}. If this switch is present, a header is written
8367 to @file{stdout} and any error messages are directed to @file{stdout}.
8368 All that is written to @file{stderr} is a brief summary message.
8370 @item ^-b^/REPORT_ERRORS=BRIEF^
8371 @cindex @option{^-b^/REPORT_ERRORS=BRIEF^} (@code{gnatbind})
8372 Generate brief error messages to @file{stderr} even if verbose mode is
8373 specified. This is relevant only when used with the
8374 @option{^-v^/REPORT_ERRORS=VERBOSE^} switch.
8378 @cindex @option{-m} (@code{gnatbind})
8379 Limits the number of error messages to @var{n}, a decimal integer in the
8380 range 1-999. The binder terminates immediately if this limit is reached.
8383 @cindex @option{-M} (@code{gnatbind})
8384 Renames the generated main program from @code{main} to @code{xxx}.
8385 This is useful in the case of some cross-building environments, where
8386 the actual main program is separate from the one generated
8390 @item ^-ws^/WARNINGS=SUPPRESS^
8391 @cindex @option{^-ws^/WARNINGS=SUPPRESS^} (@code{gnatbind})
8393 Suppress all warning messages.
8395 @item ^-we^/WARNINGS=ERROR^
8396 @cindex @option{^-we^/WARNINGS=ERROR^} (@code{gnatbind})
8397 Treat any warning messages as fatal errors.
8400 @item /WARNINGS=NORMAL
8401 Standard mode with warnings generated, but warnings do not get treated
8405 @item ^-t^/NOTIME_STAMP_CHECK^
8406 @cindex @option{^-t^/NOTIME_STAMP_CHECK^} (@code{gnatbind})
8407 @cindex Time stamp checks, in binder
8408 @cindex Binder consistency checks
8409 @cindex Consistency checks, in binder
8410 The binder performs a number of consistency checks including:
8414 Check that time stamps of a given source unit are consistent
8416 Check that checksums of a given source unit are consistent
8418 Check that consistent versions of @code{GNAT} were used for compilation
8420 Check consistency of configuration pragmas as required
8424 Normally failure of such checks, in accordance with the consistency
8425 requirements of the Ada Reference Manual, causes error messages to be
8426 generated which abort the binder and prevent the output of a binder
8427 file and subsequent link to obtain an executable.
8429 The @option{^-t^/NOTIME_STAMP_CHECK^} switch converts these error messages
8430 into warnings, so that
8431 binding and linking can continue to completion even in the presence of such
8432 errors. The result may be a failed link (due to missing symbols), or a
8433 non-functional executable which has undefined semantics.
8434 @emph{This means that
8435 @option{^-t^/NOTIME_STAMP_CHECK^} should be used only in unusual situations,
8439 @node Elaboration Control
8440 @subsection Elaboration Control
8443 The following switches provide additional control over the elaboration
8444 order. For full details see @ref{Elaboration Order Handling in GNAT}.
8447 @item ^-p^/PESSIMISTIC_ELABORATION^
8448 @cindex @option{^-p^/PESSIMISTIC_ELABORATION^} (@code{gnatbind})
8449 Normally the binder attempts to choose an elaboration order that is
8450 likely to minimize the likelihood of an elaboration order error resulting
8451 in raising a @code{Program_Error} exception. This switch reverses the
8452 action of the binder, and requests that it deliberately choose an order
8453 that is likely to maximize the likelihood of an elaboration error.
8454 This is useful in ensuring portability and avoiding dependence on
8455 accidental fortuitous elaboration ordering.
8457 Normally it only makes sense to use the @option{^-p^/PESSIMISTIC_ELABORATION^}
8459 elaboration checking is used (@option{-gnatE} switch used for compilation).
8460 This is because in the default static elaboration mode, all necessary
8461 @code{Elaborate} and @code{Elaborate_All} pragmas are implicitly inserted.
8462 These implicit pragmas are still respected by the binder in
8463 @option{^-p^/PESSIMISTIC_ELABORATION^} mode, so a
8464 safe elaboration order is assured.
8467 @node Output Control
8468 @subsection Output Control
8471 The following switches allow additional control over the output
8472 generated by the binder.
8477 @item ^-A^/BIND_FILE=ADA^
8478 @cindex @option{^-A^/BIND_FILE=ADA^} (@code{gnatbind})
8479 Generate binder program in Ada (default). The binder program is named
8480 @file{b~@var{mainprog}.adb} by default. This can be changed with
8481 @option{^-o^/OUTPUT^} @code{gnatbind} option.
8483 @item ^-c^/NOOUTPUT^
8484 @cindex @option{^-c^/NOOUTPUT^} (@code{gnatbind})
8485 Check only. Do not generate the binder output file. In this mode the
8486 binder performs all error checks but does not generate an output file.
8488 @item ^-C^/BIND_FILE=C^
8489 @cindex @option{^-C^/BIND_FILE=C^} (@code{gnatbind})
8490 Generate binder program in C. The binder program is named
8491 @file{b_@var{mainprog}.c}.
8492 This can be changed with @option{^-o^/OUTPUT^} @code{gnatbind}
8495 @item ^-e^/ELABORATION_DEPENDENCIES^
8496 @cindex @option{^-e^/ELABORATION_DEPENDENCIES^} (@code{gnatbind})
8497 Output complete list of elaboration-order dependencies, showing the
8498 reason for each dependency. This output can be rather extensive but may
8499 be useful in diagnosing problems with elaboration order. The output is
8500 written to @file{stdout}.
8503 @cindex @option{^-h^/HELP^} (@code{gnatbind})
8504 Output usage information. The output is written to @file{stdout}.
8506 @item ^-K^/LINKER_OPTION_LIST^
8507 @cindex @option{^-K^/LINKER_OPTION_LIST^} (@code{gnatbind})
8508 Output linker options to @file{stdout}. Includes library search paths,
8509 contents of pragmas Ident and Linker_Options, and libraries added
8512 @item ^-l^/ORDER_OF_ELABORATION^
8513 @cindex @option{^-l^/ORDER_OF_ELABORATION^} (@code{gnatbind})
8514 Output chosen elaboration order. The output is written to @file{stdout}.
8516 @item ^-O^/OBJECT_LIST^
8517 @cindex @option{^-O^/OBJECT_LIST^} (@code{gnatbind})
8518 Output full names of all the object files that must be linked to provide
8519 the Ada component of the program. The output is written to @file{stdout}.
8520 This list includes the files explicitly supplied and referenced by the user
8521 as well as implicitly referenced run-time unit files. The latter are
8522 omitted if the corresponding units reside in shared libraries. The
8523 directory names for the run-time units depend on the system configuration.
8525 @item ^-o ^/OUTPUT=^@var{file}
8526 @cindex @option{^-o^/OUTPUT^} (@code{gnatbind})
8527 Set name of output file to @var{file} instead of the normal
8528 @file{b~@var{mainprog}.adb} default. Note that @var{file} denote the Ada
8529 binder generated body filename. In C mode you would normally give
8530 @var{file} an extension of @file{.c} because it will be a C source program.
8531 Note that if this option is used, then linking must be done manually.
8532 It is not possible to use gnatlink in this case, since it cannot locate
8535 @item ^-r^/RESTRICTION_LIST^
8536 @cindex @option{^-r^/RESTRICTION_LIST^} (@code{gnatbind})
8537 Generate list of @code{pragma Restrictions} that could be applied to
8538 the current unit. This is useful for code audit purposes, and also may
8539 be used to improve code generation in some cases.
8543 @node Binding with Non-Ada Main Programs
8544 @subsection Binding with Non-Ada Main Programs
8547 In our description so far we have assumed that the main
8548 program is in Ada, and that the task of the binder is to generate a
8549 corresponding function @code{main} that invokes this Ada main
8550 program. GNAT also supports the building of executable programs where
8551 the main program is not in Ada, but some of the called routines are
8552 written in Ada and compiled using GNAT (@pxref{Mixed Language Programming}).
8553 The following switch is used in this situation:
8557 @cindex @option{^-n^/NOMAIN^} (@code{gnatbind})
8558 No main program. The main program is not in Ada.
8562 In this case, most of the functions of the binder are still required,
8563 but instead of generating a main program, the binder generates a file
8564 containing the following callable routines:
8569 You must call this routine to initialize the Ada part of the program by
8570 calling the necessary elaboration routines. A call to @code{adainit} is
8571 required before the first call to an Ada subprogram.
8573 Note that it is assumed that the basic execution environment must be setup
8574 to be appropriate for Ada execution at the point where the first Ada
8575 subprogram is called. In particular, if the Ada code will do any
8576 floating-point operations, then the FPU must be setup in an appropriate
8577 manner. For the case of the x86, for example, full precision mode is
8578 required. The procedure GNAT.Float_Control.Reset may be used to ensure
8579 that the FPU is in the right state.
8583 You must call this routine to perform any library-level finalization
8584 required by the Ada subprograms. A call to @code{adafinal} is required
8585 after the last call to an Ada subprogram, and before the program
8590 If the @option{^-n^/NOMAIN^} switch
8591 @cindex @option{^-n^/NOMAIN^} (@command{gnatbind})
8592 @cindex Binder, multiple input files
8593 is given, more than one ALI file may appear on
8594 the command line for @code{gnatbind}. The normal @dfn{closure}
8595 calculation is performed for each of the specified units. Calculating
8596 the closure means finding out the set of units involved by tracing
8597 @code{with} references. The reason it is necessary to be able to
8598 specify more than one ALI file is that a given program may invoke two or
8599 more quite separate groups of Ada units.
8601 The binder takes the name of its output file from the last specified ALI
8602 file, unless overridden by the use of the @option{^-o file^/OUTPUT=file^}.
8603 @cindex @option{^-o^/OUTPUT^} (@command{gnatbind})
8604 The output is an Ada unit in source form that can
8605 be compiled with GNAT unless the -C switch is used in which case the
8606 output is a C source file, which must be compiled using the C compiler.
8607 This compilation occurs automatically as part of the @command{gnatlink}
8610 Currently the GNAT run time requires a FPU using 80 bits mode
8611 precision. Under targets where this is not the default it is required to
8612 call GNAT.Float_Control.Reset before using floating point numbers (this
8613 include float computation, float input and output) in the Ada code. A
8614 side effect is that this could be the wrong mode for the foreign code
8615 where floating point computation could be broken after this call.
8617 @node Binding Programs with No Main Subprogram
8618 @subsection Binding Programs with No Main Subprogram
8621 It is possible to have an Ada program which does not have a main
8622 subprogram. This program will call the elaboration routines of all the
8623 packages, then the finalization routines.
8625 The following switch is used to bind programs organized in this manner:
8628 @item ^-z^/ZERO_MAIN^
8629 @cindex @option{^-z^/ZERO_MAIN^} (@code{gnatbind})
8630 Normally the binder checks that the unit name given on the command line
8631 corresponds to a suitable main subprogram. When this switch is used,
8632 a list of ALI files can be given, and the execution of the program
8633 consists of elaboration of these units in an appropriate order. Note
8634 that the default wide character encoding method for standard Text_IO
8635 files is always set to Brackets if this switch is set (you can use
8637 @option{^-Wx^WIDE_CHARACTER_ENCODING^} to override this default).
8640 @node Command-Line Access
8641 @section Command-Line Access
8644 The package @code{Ada.Command_Line} provides access to the command-line
8645 arguments and program name. In order for this interface to operate
8646 correctly, the two variables
8658 are declared in one of the GNAT library routines. These variables must
8659 be set from the actual @code{argc} and @code{argv} values passed to the
8660 main program. With no @option{^n^/NOMAIN^} present, @code{gnatbind}
8661 generates the C main program to automatically set these variables.
8662 If the @option{^n^/NOMAIN^} switch is used, there is no automatic way to
8663 set these variables. If they are not set, the procedures in
8664 @code{Ada.Command_Line} will not be available, and any attempt to use
8665 them will raise @code{Constraint_Error}. If command line access is
8666 required, your main program must set @code{gnat_argc} and
8667 @code{gnat_argv} from the @code{argc} and @code{argv} values passed to
8670 @node Search Paths for gnatbind
8671 @section Search Paths for @code{gnatbind}
8674 The binder takes the name of an ALI file as its argument and needs to
8675 locate source files as well as other ALI files to verify object consistency.
8677 For source files, it follows exactly the same search rules as @command{gcc}
8678 (@pxref{Search Paths and the Run-Time Library (RTL)}). For ALI files the
8679 directories searched are:
8683 The directory containing the ALI file named in the command line, unless
8684 the switch @option{^-I-^/NOCURRENT_DIRECTORY^} is specified.
8687 All directories specified by @option{^-I^/SEARCH^}
8688 switches on the @code{gnatbind}
8689 command line, in the order given.
8692 @findex ADA_PRJ_OBJECTS_FILE
8693 Each of the directories listed in the text file whose name is given
8694 by the @env{ADA_PRJ_OBJECTS_FILE} ^environment variable^logical name^.
8697 @env{ADA_PRJ_OBJECTS_FILE} is normally set by gnatmake or by the ^gnat^GNAT^
8698 driver when project files are used. It should not normally be set
8702 @findex ADA_OBJECTS_PATH
8703 Each of the directories listed in the value of the
8704 @env{ADA_OBJECTS_PATH} ^environment variable^logical name^.
8706 Construct this value
8707 exactly as the @env{PATH} environment variable: a list of directory
8708 names separated by colons (semicolons when working with the NT version
8712 Normally, define this value as a logical name containing a comma separated
8713 list of directory names.
8715 This variable can also be defined by means of an environment string
8716 (an argument to the HP C exec* set of functions).
8720 DEFINE ANOTHER_PATH FOO:[BAG]
8721 DEFINE ADA_OBJECTS_PATH ANOTHER_PATH,FOO:[BAM],FOO:[BAR]
8724 By default, the path includes GNU:[LIB.OPENVMS7_x.2_8_x.DECLIB]
8725 first, followed by the standard Ada
8726 libraries in GNU:[LIB.OPENVMS7_x.2_8_x.ADALIB].
8727 If this is not redefined, the user will obtain the HP Ada 83 IO packages
8728 (Text_IO, Sequential_IO, etc)
8729 instead of the standard Ada packages. Thus, in order to get the standard Ada
8730 packages by default, ADA_OBJECTS_PATH must be redefined.
8734 The content of the @file{ada_object_path} file which is part of the GNAT
8735 installation tree and is used to store standard libraries such as the
8736 GNAT Run Time Library (RTL) unless the switch @option{-nostdlib} is
8739 @ref{Installing a library}
8744 In the binder the switch @option{^-I^/SEARCH^}
8745 @cindex @option{^-I^/SEARCH^} (@command{gnatbind})
8746 is used to specify both source and
8747 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
8748 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatbind})
8749 instead if you want to specify
8750 source paths only, and @option{^-aO^/LIBRARY_SEARCH^}
8751 @cindex @option{^-aO^/LIBRARY_SEARCH^} (@command{gnatbind})
8752 if you want to specify library paths
8753 only. This means that for the binder
8754 @option{^-I^/SEARCH=^}@var{dir} is equivalent to
8755 @option{^-aI^/SOURCE_SEARCH=^}@var{dir}
8756 @option{^-aO^/OBJECT_SEARCH=^}@var{dir}.
8757 The binder generates the bind file (a C language source file) in the
8758 current working directory.
8764 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
8765 children make up the GNAT Run-Time Library, together with the package
8766 GNAT and its children, which contain a set of useful additional
8767 library functions provided by GNAT. The sources for these units are
8768 needed by the compiler and are kept together in one directory. The ALI
8769 files and object files generated by compiling the RTL are needed by the
8770 binder and the linker and are kept together in one directory, typically
8771 different from the directory containing the sources. In a normal
8772 installation, you need not specify these directory names when compiling
8773 or binding. Either the environment variables or the built-in defaults
8774 cause these files to be found.
8776 Besides simplifying access to the RTL, a major use of search paths is
8777 in compiling sources from multiple directories. This can make
8778 development environments much more flexible.
8780 @node Examples of gnatbind Usage
8781 @section Examples of @code{gnatbind} Usage
8784 This section contains a number of examples of using the GNAT binding
8785 utility @code{gnatbind}.
8788 @item gnatbind hello
8789 The main program @code{Hello} (source program in @file{hello.adb}) is
8790 bound using the standard switch settings. The generated main program is
8791 @file{b~hello.adb}. This is the normal, default use of the binder.
8794 @item gnatbind hello -o mainprog.adb
8797 @item gnatbind HELLO.ALI /OUTPUT=Mainprog.ADB
8799 The main program @code{Hello} (source program in @file{hello.adb}) is
8800 bound using the standard switch settings. The generated main program is
8801 @file{mainprog.adb} with the associated spec in
8802 @file{mainprog.ads}. Note that you must specify the body here not the
8803 spec, in the case where the output is in Ada. Note that if this option
8804 is used, then linking must be done manually, since gnatlink will not
8805 be able to find the generated file.
8808 @item gnatbind main -C -o mainprog.c -x
8811 @item gnatbind MAIN.ALI /BIND_FILE=C /OUTPUT=Mainprog.C /READ_SOURCES=NONE
8813 The main program @code{Main} (source program in
8814 @file{main.adb}) is bound, excluding source files from the
8815 consistency checking, generating
8816 the file @file{mainprog.c}.
8819 @item gnatbind -x main_program -C -o mainprog.c
8820 This command is exactly the same as the previous example. Switches may
8821 appear anywhere in the command line, and single letter switches may be
8822 combined into a single switch.
8826 @item gnatbind -n math dbase -C -o ada-control.c
8829 @item gnatbind /NOMAIN math dbase /BIND_FILE=C /OUTPUT=ada-control.c
8831 The main program is in a language other than Ada, but calls to
8832 subprograms in packages @code{Math} and @code{Dbase} appear. This call
8833 to @code{gnatbind} generates the file @file{ada-control.c} containing
8834 the @code{adainit} and @code{adafinal} routines to be called before and
8835 after accessing the Ada units.
8838 @c ------------------------------------
8839 @node Linking Using gnatlink
8840 @chapter Linking Using @command{gnatlink}
8841 @c ------------------------------------
8845 This chapter discusses @command{gnatlink}, a tool that links
8846 an Ada program and builds an executable file. This utility
8847 invokes the system linker ^(via the @command{gcc} command)^^
8848 with a correct list of object files and library references.
8849 @command{gnatlink} automatically determines the list of files and
8850 references for the Ada part of a program. It uses the binder file
8851 generated by the @command{gnatbind} to determine this list.
8853 Note: to invoke @code{gnatlink} with a project file, use the @code{gnat}
8854 driver (see @ref{The GNAT Driver and Project Files}).
8857 * Running gnatlink::
8858 * Switches for gnatlink::
8861 @node Running gnatlink
8862 @section Running @command{gnatlink}
8865 The form of the @command{gnatlink} command is
8868 $ gnatlink @ovar{switches} @var{mainprog}@r{[}.ali@r{]}
8869 @ovar{non-Ada objects} @ovar{linker options}
8873 The arguments of @command{gnatlink} (switches, main @file{ALI} file,
8875 or linker options) may be in any order, provided that no non-Ada object may
8876 be mistaken for a main @file{ALI} file.
8877 Any file name @file{F} without the @file{.ali}
8878 extension will be taken as the main @file{ALI} file if a file exists
8879 whose name is the concatenation of @file{F} and @file{.ali}.
8882 @file{@var{mainprog}.ali} references the ALI file of the main program.
8883 The @file{.ali} extension of this file can be omitted. From this
8884 reference, @command{gnatlink} locates the corresponding binder file
8885 @file{b~@var{mainprog}.adb} and, using the information in this file along
8886 with the list of non-Ada objects and linker options, constructs a
8887 linker command file to create the executable.
8889 The arguments other than the @command{gnatlink} switches and the main
8890 @file{ALI} file are passed to the linker uninterpreted.
8891 They typically include the names of
8892 object files for units written in other languages than Ada and any library
8893 references required to resolve references in any of these foreign language
8894 units, or in @code{Import} pragmas in any Ada units.
8896 @var{linker options} is an optional list of linker specific
8898 The default linker called by gnatlink is @command{gcc} which in
8899 turn calls the appropriate system linker.
8900 Standard options for the linker such as @option{-lmy_lib} or
8901 @option{-Ldir} can be added as is.
8902 For options that are not recognized by
8903 @command{gcc} as linker options, use the @command{gcc} switches
8904 @option{-Xlinker} or @option{-Wl,}.
8905 Refer to the GCC documentation for
8906 details. Here is an example showing how to generate a linker map:
8909 $ ^gnatlink my_prog -Wl,-Map,MAPFILE^GNAT LINK my_prog.ali /MAP^
8912 Using @var{linker options} it is possible to set the program stack and
8915 See @ref{Setting Stack Size from gnatlink} and
8916 @ref{Setting Heap Size from gnatlink}.
8919 @command{gnatlink} determines the list of objects required by the Ada
8920 program and prepends them to the list of objects passed to the linker.
8921 @command{gnatlink} also gathers any arguments set by the use of
8922 @code{pragma Linker_Options} and adds them to the list of arguments
8923 presented to the linker.
8926 @command{gnatlink} accepts the following types of extra files on the command
8927 line: objects (@file{.OBJ}), libraries (@file{.OLB}), sharable images
8928 (@file{.EXE}), and options files (@file{.OPT}). These are recognized and
8929 handled according to their extension.
8932 @node Switches for gnatlink
8933 @section Switches for @command{gnatlink}
8936 The following switches are available with the @command{gnatlink} utility:
8942 @cindex @option{--version} @command{gnatlink}
8943 Display Copyright and version, then exit disregarding all other options.
8946 @cindex @option{--help} @command{gnatlink}
8947 If @option{--version} was not used, display usage, then exit disregarding
8950 @item ^-A^/BIND_FILE=ADA^
8951 @cindex @option{^-A^/BIND_FILE=ADA^} (@command{gnatlink})
8952 The binder has generated code in Ada. This is the default.
8954 @item ^-C^/BIND_FILE=C^
8955 @cindex @option{^-C^/BIND_FILE=C^} (@command{gnatlink})
8956 If instead of generating a file in Ada, the binder has generated one in
8957 C, then the linker needs to know about it. Use this switch to signal
8958 to @command{gnatlink} that the binder has generated C code rather than
8961 @item ^-f^/FORCE_OBJECT_FILE_LIST^
8962 @cindex Command line length
8963 @cindex @option{^-f^/FORCE_OBJECT_FILE_LIST^} (@command{gnatlink})
8964 On some targets, the command line length is limited, and @command{gnatlink}
8965 will generate a separate file for the linker if the list of object files
8967 The @option{^-f^/FORCE_OBJECT_FILE_LIST^} switch forces this file
8968 to be generated even if
8969 the limit is not exceeded. This is useful in some cases to deal with
8970 special situations where the command line length is exceeded.
8973 @cindex Debugging information, including
8974 @cindex @option{^-g^/DEBUG^} (@command{gnatlink})
8975 The option to include debugging information causes the Ada bind file (in
8976 other words, @file{b~@var{mainprog}.adb}) to be compiled with
8977 @option{^-g^/DEBUG^}.
8978 In addition, the binder does not delete the @file{b~@var{mainprog}.adb},
8979 @file{b~@var{mainprog}.o} and @file{b~@var{mainprog}.ali} files.
8980 Without @option{^-g^/DEBUG^}, the binder removes these files by
8981 default. The same procedure apply if a C bind file was generated using
8982 @option{^-C^/BIND_FILE=C^} @code{gnatbind} option, in this case the filenames
8983 are @file{b_@var{mainprog}.c} and @file{b_@var{mainprog}.o}.
8985 @item ^-n^/NOCOMPILE^
8986 @cindex @option{^-n^/NOCOMPILE^} (@command{gnatlink})
8987 Do not compile the file generated by the binder. This may be used when
8988 a link is rerun with different options, but there is no need to recompile
8992 @cindex @option{^-v^/VERBOSE^} (@command{gnatlink})
8993 Causes additional information to be output, including a full list of the
8994 included object files. This switch option is most useful when you want
8995 to see what set of object files are being used in the link step.
8997 @item ^-v -v^/VERBOSE/VERBOSE^
8998 @cindex @option{^-v -v^/VERBOSE/VERBOSE^} (@command{gnatlink})
8999 Very verbose mode. Requests that the compiler operate in verbose mode when
9000 it compiles the binder file, and that the system linker run in verbose mode.
9002 @item ^-o ^/EXECUTABLE=^@var{exec-name}
9003 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatlink})
9004 @var{exec-name} specifies an alternate name for the generated
9005 executable program. If this switch is omitted, the executable has the same
9006 name as the main unit. For example, @code{gnatlink try.ali} creates
9007 an executable called @file{^try^TRY.EXE^}.
9010 @item -b @var{target}
9011 @cindex @option{-b} (@command{gnatlink})
9012 Compile your program to run on @var{target}, which is the name of a
9013 system configuration. You must have a GNAT cross-compiler built if
9014 @var{target} is not the same as your host system.
9017 @cindex @option{-B} (@command{gnatlink})
9018 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
9019 from @var{dir} instead of the default location. Only use this switch
9020 when multiple versions of the GNAT compiler are available.
9021 @xref{Directory Options,,, gcc, The GNU Compiler Collection},
9022 for further details. You would normally use the @option{-b} or
9023 @option{-V} switch instead.
9025 @item --GCC=@var{compiler_name}
9026 @cindex @option{--GCC=compiler_name} (@command{gnatlink})
9027 Program used for compiling the binder file. The default is
9028 @command{gcc}. You need to use quotes around @var{compiler_name} if
9029 @code{compiler_name} contains spaces or other separator characters.
9030 As an example @option{--GCC="foo -x -y"} will instruct @command{gnatlink} to
9031 use @code{foo -x -y} as your compiler. Note that switch @option{-c} is always
9032 inserted after your command name. Thus in the above example the compiler
9033 command that will be used by @command{gnatlink} will be @code{foo -c -x -y}.
9034 A limitation of this syntax is that the name and path name of the executable
9035 itself must not include any embedded spaces. If the compiler executable is
9036 different from the default one (gcc or <prefix>-gcc), then the back-end
9037 switches in the ALI file are not used to compile the binder generated source.
9038 For example, this is the case with @option{--GCC="foo -x -y"}. But the back end
9039 switches will be used for @option{--GCC="gcc -gnatv"}. If several
9040 @option{--GCC=compiler_name} are used, only the last @var{compiler_name}
9041 is taken into account. However, all the additional switches are also taken
9043 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
9044 @option{--GCC="bar -x -y -z -t"}.
9046 @item --LINK=@var{name}
9047 @cindex @option{--LINK=} (@command{gnatlink})
9048 @var{name} is the name of the linker to be invoked. This is especially
9049 useful in mixed language programs since languages such as C++ require
9050 their own linker to be used. When this switch is omitted, the default
9051 name for the linker is @command{gcc}. When this switch is used, the
9052 specified linker is called instead of @command{gcc} with exactly the same
9053 parameters that would have been passed to @command{gcc} so if the desired
9054 linker requires different parameters it is necessary to use a wrapper
9055 script that massages the parameters before invoking the real linker. It
9056 may be useful to control the exact invocation by using the verbose
9062 @item /DEBUG=TRACEBACK
9063 @cindex @code{/DEBUG=TRACEBACK} (@command{gnatlink})
9064 This qualifier causes sufficient information to be included in the
9065 executable file to allow a traceback, but does not include the full
9066 symbol information needed by the debugger.
9068 @item /IDENTIFICATION="<string>"
9069 @code{"<string>"} specifies the string to be stored in the image file
9070 identification field in the image header.
9071 It overrides any pragma @code{Ident} specified string.
9073 @item /NOINHIBIT-EXEC
9074 Generate the executable file even if there are linker warnings.
9076 @item /NOSTART_FILES
9077 Don't link in the object file containing the ``main'' transfer address.
9078 Used when linking with a foreign language main program compiled with an
9082 Prefer linking with object libraries over sharable images, even without
9088 @node The GNAT Make Program gnatmake
9089 @chapter The GNAT Make Program @command{gnatmake}
9093 * Running gnatmake::
9094 * Switches for gnatmake::
9095 * Mode Switches for gnatmake::
9096 * Notes on the Command Line::
9097 * How gnatmake Works::
9098 * Examples of gnatmake Usage::
9101 A typical development cycle when working on an Ada program consists of
9102 the following steps:
9106 Edit some sources to fix bugs.
9112 Compile all sources affected.
9122 The third step can be tricky, because not only do the modified files
9123 @cindex Dependency rules
9124 have to be compiled, but any files depending on these files must also be
9125 recompiled. The dependency rules in Ada can be quite complex, especially
9126 in the presence of overloading, @code{use} clauses, generics and inlined
9129 @command{gnatmake} automatically takes care of the third and fourth steps
9130 of this process. It determines which sources need to be compiled,
9131 compiles them, and binds and links the resulting object files.
9133 Unlike some other Ada make programs, the dependencies are always
9134 accurately recomputed from the new sources. The source based approach of
9135 the GNAT compilation model makes this possible. This means that if
9136 changes to the source program cause corresponding changes in
9137 dependencies, they will always be tracked exactly correctly by
9140 @node Running gnatmake
9141 @section Running @command{gnatmake}
9144 The usual form of the @command{gnatmake} command is
9147 $ gnatmake @ovar{switches} @var{file_name}
9148 @ovar{file_names} @ovar{mode_switches}
9152 The only required argument is one @var{file_name}, which specifies
9153 a compilation unit that is a main program. Several @var{file_names} can be
9154 specified: this will result in several executables being built.
9155 If @code{switches} are present, they can be placed before the first
9156 @var{file_name}, between @var{file_names} or after the last @var{file_name}.
9157 If @var{mode_switches} are present, they must always be placed after
9158 the last @var{file_name} and all @code{switches}.
9160 If you are using standard file extensions (@file{.adb} and @file{.ads}), then the
9161 extension may be omitted from the @var{file_name} arguments. However, if
9162 you are using non-standard extensions, then it is required that the
9163 extension be given. A relative or absolute directory path can be
9164 specified in a @var{file_name}, in which case, the input source file will
9165 be searched for in the specified directory only. Otherwise, the input
9166 source file will first be searched in the directory where
9167 @command{gnatmake} was invoked and if it is not found, it will be search on
9168 the source path of the compiler as described in
9169 @ref{Search Paths and the Run-Time Library (RTL)}.
9171 All @command{gnatmake} output (except when you specify
9172 @option{^-M^/DEPENDENCIES_LIST^}) is to
9173 @file{stderr}. The output produced by the
9174 @option{^-M^/DEPENDENCIES_LIST^} switch is send to
9177 @node Switches for gnatmake
9178 @section Switches for @command{gnatmake}
9181 You may specify any of the following switches to @command{gnatmake}:
9187 @cindex @option{--version} @command{gnatmake}
9188 Display Copyright and version, then exit disregarding all other options.
9191 @cindex @option{--help} @command{gnatmake}
9192 If @option{--version} was not used, display usage, then exit disregarding
9196 @item --GCC=@var{compiler_name}
9197 @cindex @option{--GCC=compiler_name} (@command{gnatmake})
9198 Program used for compiling. The default is `@command{gcc}'. You need to use
9199 quotes around @var{compiler_name} if @code{compiler_name} contains
9200 spaces or other separator characters. As an example @option{--GCC="foo -x
9201 -y"} will instruct @command{gnatmake} to use @code{foo -x -y} as your
9202 compiler. A limitation of this syntax is that the name and path name of
9203 the executable itself must not include any embedded spaces. Note that
9204 switch @option{-c} is always inserted after your command name. Thus in the
9205 above example the compiler command that will be used by @command{gnatmake}
9206 will be @code{foo -c -x -y}. If several @option{--GCC=compiler_name} are
9207 used, only the last @var{compiler_name} is taken into account. However,
9208 all the additional switches are also taken into account. Thus,
9209 @option{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
9210 @option{--GCC="bar -x -y -z -t"}.
9212 @item --GNATBIND=@var{binder_name}
9213 @cindex @option{--GNATBIND=binder_name} (@command{gnatmake})
9214 Program used for binding. The default is `@code{gnatbind}'. You need to
9215 use quotes around @var{binder_name} if @var{binder_name} contains spaces
9216 or other separator characters. As an example @option{--GNATBIND="bar -x
9217 -y"} will instruct @command{gnatmake} to use @code{bar -x -y} as your
9218 binder. Binder switches that are normally appended by @command{gnatmake}
9219 to `@code{gnatbind}' are now appended to the end of @code{bar -x -y}.
9220 A limitation of this syntax is that the name and path name of the executable
9221 itself must not include any embedded spaces.
9223 @item --GNATLINK=@var{linker_name}
9224 @cindex @option{--GNATLINK=linker_name} (@command{gnatmake})
9225 Program used for linking. The default is `@command{gnatlink}'. You need to
9226 use quotes around @var{linker_name} if @var{linker_name} contains spaces
9227 or other separator characters. As an example @option{--GNATLINK="lan -x
9228 -y"} will instruct @command{gnatmake} to use @code{lan -x -y} as your
9229 linker. Linker switches that are normally appended by @command{gnatmake} to
9230 `@command{gnatlink}' are now appended to the end of @code{lan -x -y}.
9231 A limitation of this syntax is that the name and path name of the executable
9232 itself must not include any embedded spaces.
9236 @item ^-a^/ALL_FILES^
9237 @cindex @option{^-a^/ALL_FILES^} (@command{gnatmake})
9238 Consider all files in the make process, even the GNAT internal system
9239 files (for example, the predefined Ada library files), as well as any
9240 locked files. Locked files are files whose ALI file is write-protected.
9242 @command{gnatmake} does not check these files,
9243 because the assumption is that the GNAT internal files are properly up
9244 to date, and also that any write protected ALI files have been properly
9245 installed. Note that if there is an installation problem, such that one
9246 of these files is not up to date, it will be properly caught by the
9248 You may have to specify this switch if you are working on GNAT
9249 itself. The switch @option{^-a^/ALL_FILES^} is also useful
9250 in conjunction with @option{^-f^/FORCE_COMPILE^}
9251 if you need to recompile an entire application,
9252 including run-time files, using special configuration pragmas,
9253 such as a @code{Normalize_Scalars} pragma.
9256 @code{gnatmake ^-a^/ALL_FILES^} compiles all GNAT
9259 @code{gcc -c -gnatpg} rather than @code{gcc -c}.
9262 the @code{/CHECKS=SUPPRESS_ALL /STYLE_CHECKS=GNAT} switch.
9265 @item ^-b^/ACTIONS=BIND^
9266 @cindex @option{^-b^/ACTIONS=BIND^} (@command{gnatmake})
9267 Bind only. Can be combined with @option{^-c^/ACTIONS=COMPILE^} to do
9268 compilation and binding, but no link.
9269 Can be combined with @option{^-l^/ACTIONS=LINK^}
9270 to do binding and linking. When not combined with
9271 @option{^-c^/ACTIONS=COMPILE^}
9272 all the units in the closure of the main program must have been previously
9273 compiled and must be up to date. The root unit specified by @var{file_name}
9274 may be given without extension, with the source extension or, if no GNAT
9275 Project File is specified, with the ALI file extension.
9277 @item ^-c^/ACTIONS=COMPILE^
9278 @cindex @option{^-c^/ACTIONS=COMPILE^} (@command{gnatmake})
9279 Compile only. Do not perform binding, except when @option{^-b^/ACTIONS=BIND^}
9280 is also specified. Do not perform linking, except if both
9281 @option{^-b^/ACTIONS=BIND^} and
9282 @option{^-l^/ACTIONS=LINK^} are also specified.
9283 If the root unit specified by @var{file_name} is not a main unit, this is the
9284 default. Otherwise @command{gnatmake} will attempt binding and linking
9285 unless all objects are up to date and the executable is more recent than
9289 @cindex @option{^-C^/MAPPING^} (@command{gnatmake})
9290 Use a temporary mapping file. A mapping file is a way to communicate
9291 to the compiler two mappings: from unit names to file names (without
9292 any directory information) and from file names to path names (with
9293 full directory information). A mapping file can make the compiler's
9294 file searches faster, especially if there are many source directories,
9295 or the sources are read over a slow network connection. If
9296 @option{^-P^/PROJECT_FILE^} is used, a mapping file is always used, so
9297 @option{^-C^/MAPPING^} is unnecessary; in this case the mapping file
9298 is initially populated based on the project file. If
9299 @option{^-C^/MAPPING^} is used without
9300 @option{^-P^/PROJECT_FILE^},
9301 the mapping file is initially empty. Each invocation of the compiler
9302 will add any newly accessed sources to the mapping file.
9304 @item ^-C=^/USE_MAPPING_FILE=^@var{file}
9305 @cindex @option{^-C=^/USE_MAPPING^} (@command{gnatmake})
9306 Use a specific mapping file. The file, specified as a path name (absolute or
9307 relative) by this switch, should already exist, otherwise the switch is
9308 ineffective. The specified mapping file will be communicated to the compiler.
9309 This switch is not compatible with a project file
9310 (^-P^/PROJECT_FILE=^@var{file}) or with multiple compiling processes
9311 (^-j^/PROCESSES=^nnn, when nnn is greater than 1).
9313 @item ^-d^/DISPLAY_PROGRESS^
9314 @cindex @option{^-d^/DISPLAY_PROGRESS^} (@command{gnatmake})
9315 Display progress for each source, up to date or not, as a single line
9318 completed x out of y (zz%)
9321 If the file needs to be compiled this is displayed after the invocation of
9322 the compiler. These lines are displayed even in quiet output mode.
9324 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
9325 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@command{gnatmake})
9326 Put all object files and ALI file in directory @var{dir}.
9327 If the @option{^-D^/DIRECTORY_OBJECTS^} switch is not used, all object files
9328 and ALI files go in the current working directory.
9330 This switch cannot be used when using a project file.
9334 @cindex @option{-eL} (@command{gnatmake})
9335 @cindex symbolic links
9336 Follow all symbolic links when processing project files.
9337 This should be used if your project uses symbolic links for files or
9338 directories, but is not needed in other cases.
9340 @cindex naming scheme
9341 This also assumes that no directory matches the naming scheme for files (for
9342 instance that you do not have a directory called "sources.ads" when using the
9343 default GNAT naming scheme).
9345 When you do not have to use this switch (ie by default), gnatmake is able to
9346 save a lot of system calls (several per source file and object file), which
9347 can result in a significant speed up to load and manipulate a project file,
9348 especially when using source files from a remote system.
9352 @item ^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^
9353 @cindex @option{^-eS^/STANDARD_OUTPUT_FOR_COMMANDS^} (@command{gnatmake})
9354 Output the commands for the compiler, the binder and the linker
9355 on ^standard output^SYS$OUTPUT^,
9356 instead of ^standard error^SYS$ERROR^.
9358 @item ^-f^/FORCE_COMPILE^
9359 @cindex @option{^-f^/FORCE_COMPILE^} (@command{gnatmake})
9360 Force recompilations. Recompile all sources, even though some object
9361 files may be up to date, but don't recompile predefined or GNAT internal
9362 files or locked files (files with a write-protected ALI file),
9363 unless the @option{^-a^/ALL_FILES^} switch is also specified.
9365 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
9366 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@command{gnatmake})
9367 When using project files, if some errors or warnings are detected during
9368 parsing and verbose mode is not in effect (no use of switch
9369 ^-v^/VERBOSE^), then error lines start with the full path name of the project
9370 file, rather than its simple file name.
9373 @cindex @option{^-g^/DEBUG^} (@command{gnatmake})
9374 Enable debugging. This switch is simply passed to the compiler and to the
9377 @item ^-i^/IN_PLACE^
9378 @cindex @option{^-i^/IN_PLACE^} (@command{gnatmake})
9379 In normal mode, @command{gnatmake} compiles all object files and ALI files
9380 into the current directory. If the @option{^-i^/IN_PLACE^} switch is used,
9381 then instead object files and ALI files that already exist are overwritten
9382 in place. This means that once a large project is organized into separate
9383 directories in the desired manner, then @command{gnatmake} will automatically
9384 maintain and update this organization. If no ALI files are found on the
9385 Ada object path (@ref{Search Paths and the Run-Time Library (RTL)}),
9386 the new object and ALI files are created in the
9387 directory containing the source being compiled. If another organization
9388 is desired, where objects and sources are kept in different directories,
9389 a useful technique is to create dummy ALI files in the desired directories.
9390 When detecting such a dummy file, @command{gnatmake} will be forced to
9391 recompile the corresponding source file, and it will be put the resulting
9392 object and ALI files in the directory where it found the dummy file.
9394 @item ^-j^/PROCESSES=^@var{n}
9395 @cindex @option{^-j^/PROCESSES^} (@command{gnatmake})
9396 @cindex Parallel make
9397 Use @var{n} processes to carry out the (re)compilations. On a
9398 multiprocessor machine compilations will occur in parallel. In the
9399 event of compilation errors, messages from various compilations might
9400 get interspersed (but @command{gnatmake} will give you the full ordered
9401 list of failing compiles at the end). If this is problematic, rerun
9402 the make process with n set to 1 to get a clean list of messages.
9404 @item ^-k^/CONTINUE_ON_ERROR^
9405 @cindex @option{^-k^/CONTINUE_ON_ERROR^} (@command{gnatmake})
9406 Keep going. Continue as much as possible after a compilation error. To
9407 ease the programmer's task in case of compilation errors, the list of
9408 sources for which the compile fails is given when @command{gnatmake}
9411 If @command{gnatmake} is invoked with several @file{file_names} and with this
9412 switch, if there are compilation errors when building an executable,
9413 @command{gnatmake} will not attempt to build the following executables.
9415 @item ^-l^/ACTIONS=LINK^
9416 @cindex @option{^-l^/ACTIONS=LINK^} (@command{gnatmake})
9417 Link only. Can be combined with @option{^-b^/ACTIONS=BIND^} to binding
9418 and linking. Linking will not be performed if combined with
9419 @option{^-c^/ACTIONS=COMPILE^}
9420 but not with @option{^-b^/ACTIONS=BIND^}.
9421 When not combined with @option{^-b^/ACTIONS=BIND^}
9422 all the units in the closure of the main program must have been previously
9423 compiled and must be up to date, and the main program needs to have been bound.
9424 The root unit specified by @var{file_name}
9425 may be given without extension, with the source extension or, if no GNAT
9426 Project File is specified, with the ALI file extension.
9428 @item ^-m^/MINIMAL_RECOMPILATION^
9429 @cindex @option{^-m^/MINIMAL_RECOMPILATION^} (@command{gnatmake})
9430 Specify that the minimum necessary amount of recompilations
9431 be performed. In this mode @command{gnatmake} ignores time
9432 stamp differences when the only
9433 modifications to a source file consist in adding/removing comments,
9434 empty lines, spaces or tabs. This means that if you have changed the
9435 comments in a source file or have simply reformatted it, using this
9436 switch will tell @command{gnatmake} not to recompile files that depend on it
9437 (provided other sources on which these files depend have undergone no
9438 semantic modifications). Note that the debugging information may be
9439 out of date with respect to the sources if the @option{-m} switch causes
9440 a compilation to be switched, so the use of this switch represents a
9441 trade-off between compilation time and accurate debugging information.
9443 @item ^-M^/DEPENDENCIES_LIST^
9444 @cindex Dependencies, producing list
9445 @cindex @option{^-M^/DEPENDENCIES_LIST^} (@command{gnatmake})
9446 Check if all objects are up to date. If they are, output the object
9447 dependences to @file{stdout} in a form that can be directly exploited in
9448 a @file{Makefile}. By default, each source file is prefixed with its
9449 (relative or absolute) directory name. This name is whatever you
9450 specified in the various @option{^-aI^/SOURCE_SEARCH^}
9451 and @option{^-I^/SEARCH^} switches. If you use
9452 @code{gnatmake ^-M^/DEPENDENCIES_LIST^}
9453 @option{^-q^/QUIET^}
9454 (see below), only the source file names,
9455 without relative paths, are output. If you just specify the
9456 @option{^-M^/DEPENDENCIES_LIST^}
9457 switch, dependencies of the GNAT internal system files are omitted. This
9458 is typically what you want. If you also specify
9459 the @option{^-a^/ALL_FILES^} switch,
9460 dependencies of the GNAT internal files are also listed. Note that
9461 dependencies of the objects in external Ada libraries (see switch
9462 @option{^-aL^/SKIP_MISSING=^}@var{dir} in the following list)
9465 @item ^-n^/DO_OBJECT_CHECK^
9466 @cindex @option{^-n^/DO_OBJECT_CHECK^} (@command{gnatmake})
9467 Don't compile, bind, or link. Checks if all objects are up to date.
9468 If they are not, the full name of the first file that needs to be
9469 recompiled is printed.
9470 Repeated use of this option, followed by compiling the indicated source
9471 file, will eventually result in recompiling all required units.
9473 @item ^-o ^/EXECUTABLE=^@var{exec_name}
9474 @cindex @option{^-o^/EXECUTABLE^} (@command{gnatmake})
9475 Output executable name. The name of the final executable program will be
9476 @var{exec_name}. If the @option{^-o^/EXECUTABLE^} switch is omitted the default
9477 name for the executable will be the name of the input file in appropriate form
9478 for an executable file on the host system.
9480 This switch cannot be used when invoking @command{gnatmake} with several
9483 @item ^-p or --create-missing-dirs^/CREATE_MISSING_DIRS^
9484 @cindex @option{^-p^/CREATE_MISSING_DIRS^} (@command{gnatmake})
9485 When using project files (^-P^/PROJECT_FILE=^@var{project}), create
9486 automatically missing object directories, library directories and exec
9489 @item ^-P^/PROJECT_FILE=^@var{project}
9490 @cindex @option{^-P^/PROJECT_FILE^} (@command{gnatmake})
9491 Use project file @var{project}. Only one such switch can be used.
9492 @xref{gnatmake and Project Files}.
9495 @cindex @option{^-q^/QUIET^} (@command{gnatmake})
9496 Quiet. When this flag is not set, the commands carried out by
9497 @command{gnatmake} are displayed.
9499 @item ^-s^/SWITCH_CHECK/^
9500 @cindex @option{^-s^/SWITCH_CHECK^} (@command{gnatmake})
9501 Recompile if compiler switches have changed since last compilation.
9502 All compiler switches but -I and -o are taken into account in the
9504 orders between different ``first letter'' switches are ignored, but
9505 orders between same switches are taken into account. For example,
9506 @option{-O -O2} is different than @option{-O2 -O}, but @option{-g -O}
9507 is equivalent to @option{-O -g}.
9509 This switch is recommended when Integrated Preprocessing is used.
9512 @cindex @option{^-u^/UNIQUE^} (@command{gnatmake})
9513 Unique. Recompile at most the main files. It implies -c. Combined with
9514 -f, it is equivalent to calling the compiler directly. Note that using
9515 ^-u^/UNIQUE^ with a project file and no main has a special meaning
9516 (@pxref{Project Files and Main Subprograms}).
9518 @item ^-U^/ALL_PROJECTS^
9519 @cindex @option{^-U^/ALL_PROJECTS^} (@command{gnatmake})
9520 When used without a project file or with one or several mains on the command
9521 line, is equivalent to ^-u^/UNIQUE^. When used with a project file and no main
9522 on the command line, all sources of all project files are checked and compiled
9523 if not up to date, and libraries are rebuilt, if necessary.
9526 @cindex @option{^-v^/REASONS^} (@command{gnatmake})
9527 Verbose. Display the reason for all recompilations @command{gnatmake}
9528 decides are necessary, with the highest verbosity level.
9530 @item ^-vl^/LOW_VERBOSITY^
9531 @cindex @option{^-vl^/LOW_VERBOSITY^} (@command{gnatmake})
9532 Verbosity level Low. Display fewer lines than in verbosity Medium.
9534 @item ^-vm^/MEDIUM_VERBOSITY^
9535 @cindex @option{^-vm^/MEDIUM_VERBOSITY^} (@command{gnatmake})
9536 Verbosity level Medium. Potentially display fewer lines than in verbosity High.
9538 @item ^-vh^/HIGH_VERBOSITY^
9539 @cindex @option{^-vm^/HIGH_VERBOSITY^} (@command{gnatmake})
9540 Verbosity level High. Equivalent to ^-v^/REASONS^.
9542 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
9543 Indicate the verbosity of the parsing of GNAT project files.
9544 @xref{Switches Related to Project Files}.
9546 @item ^-x^/NON_PROJECT_UNIT_COMPILATION^
9547 @cindex @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} (@command{gnatmake})
9548 Indicate that sources that are not part of any Project File may be compiled.
9549 Normally, when using Project Files, only sources that are part of a Project
9550 File may be compile. When this switch is used, a source outside of all Project
9551 Files may be compiled. The ALI file and the object file will be put in the
9552 object directory of the main Project. The compilation switches used will only
9553 be those specified on the command line. Even when
9554 @option{^-x^/NON_PROJECT_UNIT_COMPILATION^} is used, mains specified on the
9555 command line need to be sources of a project file.
9557 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
9558 Indicate that external variable @var{name} has the value @var{value}.
9559 The Project Manager will use this value for occurrences of
9560 @code{external(name)} when parsing the project file.
9561 @xref{Switches Related to Project Files}.
9564 @cindex @option{^-z^/NOMAIN^} (@command{gnatmake})
9565 No main subprogram. Bind and link the program even if the unit name
9566 given on the command line is a package name. The resulting executable
9567 will execute the elaboration routines of the package and its closure,
9568 then the finalization routines.
9573 @item @command{gcc} @asis{switches}
9575 Any uppercase or multi-character switch that is not a @command{gnatmake} switch
9576 is passed to @command{gcc} (e.g.@: @option{-O}, @option{-gnato,} etc.)
9579 Any qualifier that cannot be recognized as a qualifier for @code{GNAT MAKE}
9580 but is recognizable as a valid qualifier for @code{GNAT COMPILE} is
9581 automatically treated as a compiler switch, and passed on to all
9582 compilations that are carried out.
9587 Source and library search path switches:
9591 @item ^-aI^/SOURCE_SEARCH=^@var{dir}
9592 @cindex @option{^-aI^/SOURCE_SEARCH^} (@command{gnatmake})
9593 When looking for source files also look in directory @var{dir}.
9594 The order in which source files search is undertaken is
9595 described in @ref{Search Paths and the Run-Time Library (RTL)}.
9597 @item ^-aL^/SKIP_MISSING=^@var{dir}
9598 @cindex @option{^-aL^/SKIP_MISSING^} (@command{gnatmake})
9599 Consider @var{dir} as being an externally provided Ada library.
9600 Instructs @command{gnatmake} to skip compilation units whose @file{.ALI}
9601 files have been located in directory @var{dir}. This allows you to have
9602 missing bodies for the units in @var{dir} and to ignore out of date bodies
9603 for the same units. You still need to specify
9604 the location of the specs for these units by using the switches
9605 @option{^-aI^/SOURCE_SEARCH=^@var{dir}}
9606 or @option{^-I^/SEARCH=^@var{dir}}.
9607 Note: this switch is provided for compatibility with previous versions
9608 of @command{gnatmake}. The easier method of causing standard libraries
9609 to be excluded from consideration is to write-protect the corresponding
9612 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
9613 @cindex @option{^-aO^/OBJECT_SEARCH^} (@command{gnatmake})
9614 When searching for library and object files, look in directory
9615 @var{dir}. The order in which library files are searched is described in
9616 @ref{Search Paths for gnatbind}.
9618 @item ^-A^/CONDITIONAL_SOURCE_SEARCH=^@var{dir}
9619 @cindex Search paths, for @command{gnatmake}
9620 @cindex @option{^-A^/CONDITIONAL_SOURCE_SEARCH^} (@command{gnatmake})
9621 Equivalent to @option{^-aL^/SKIP_MISSING=^@var{dir}
9622 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9624 @item ^-I^/SEARCH=^@var{dir}
9625 @cindex @option{^-I^/SEARCH^} (@command{gnatmake})
9626 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}
9627 ^-aI^/SOURCE_SEARCH=^@var{dir}}.
9629 @item ^-I-^/NOCURRENT_DIRECTORY^
9630 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@command{gnatmake})
9631 @cindex Source files, suppressing search
9632 Do not look for source files in the directory containing the source
9633 file named in the command line.
9634 Do not look for ALI or object files in the directory
9635 where @command{gnatmake} was invoked.
9637 @item ^-L^/LIBRARY_SEARCH=^@var{dir}
9638 @cindex @option{^-L^/LIBRARY_SEARCH^} (@command{gnatmake})
9639 @cindex Linker libraries
9640 Add directory @var{dir} to the list of directories in which the linker
9641 will search for libraries. This is equivalent to
9642 @option{-largs ^-L^/LIBRARY_SEARCH=^}@var{dir}.
9644 Furthermore, under Windows, the sources pointed to by the libraries path
9645 set in the registry are not searched for.
9649 @cindex @option{-nostdinc} (@command{gnatmake})
9650 Do not look for source files in the system default directory.
9653 @cindex @option{-nostdlib} (@command{gnatmake})
9654 Do not look for library files in the system default directory.
9656 @item --RTS=@var{rts-path}
9657 @cindex @option{--RTS} (@command{gnatmake})
9658 Specifies the default location of the runtime library. GNAT looks for the
9660 in the following directories, and stops as soon as a valid runtime is found
9661 (@file{adainclude} or @file{ada_source_path}, and @file{adalib} or
9662 @file{ada_object_path} present):
9665 @item <current directory>/$rts_path
9667 @item <default-search-dir>/$rts_path
9669 @item <default-search-dir>/rts-$rts_path
9673 The selected path is handled like a normal RTS path.
9677 @node Mode Switches for gnatmake
9678 @section Mode Switches for @command{gnatmake}
9681 The mode switches (referred to as @code{mode_switches}) allow the
9682 inclusion of switches that are to be passed to the compiler itself, the
9683 binder or the linker. The effect of a mode switch is to cause all
9684 subsequent switches up to the end of the switch list, or up to the next
9685 mode switch, to be interpreted as switches to be passed on to the
9686 designated component of GNAT.
9690 @item -cargs @var{switches}
9691 @cindex @option{-cargs} (@command{gnatmake})
9692 Compiler switches. Here @var{switches} is a list of switches
9693 that are valid switches for @command{gcc}. They will be passed on to
9694 all compile steps performed by @command{gnatmake}.
9696 @item -bargs @var{switches}
9697 @cindex @option{-bargs} (@command{gnatmake})
9698 Binder switches. Here @var{switches} is a list of switches
9699 that are valid switches for @code{gnatbind}. They will be passed on to
9700 all bind steps performed by @command{gnatmake}.
9702 @item -largs @var{switches}
9703 @cindex @option{-largs} (@command{gnatmake})
9704 Linker switches. Here @var{switches} is a list of switches
9705 that are valid switches for @command{gnatlink}. They will be passed on to
9706 all link steps performed by @command{gnatmake}.
9708 @item -margs @var{switches}
9709 @cindex @option{-margs} (@command{gnatmake})
9710 Make switches. The switches are directly interpreted by @command{gnatmake},
9711 regardless of any previous occurrence of @option{-cargs}, @option{-bargs}
9715 @node Notes on the Command Line
9716 @section Notes on the Command Line
9719 This section contains some additional useful notes on the operation
9720 of the @command{gnatmake} command.
9724 @cindex Recompilation, by @command{gnatmake}
9725 If @command{gnatmake} finds no ALI files, it recompiles the main program
9726 and all other units required by the main program.
9727 This means that @command{gnatmake}
9728 can be used for the initial compile, as well as during subsequent steps of
9729 the development cycle.
9732 If you enter @code{gnatmake @var{file}.adb}, where @file{@var{file}.adb}
9733 is a subunit or body of a generic unit, @command{gnatmake} recompiles
9734 @file{@var{file}.adb} (because it finds no ALI) and stops, issuing a
9738 In @command{gnatmake} the switch @option{^-I^/SEARCH^}
9739 is used to specify both source and
9740 library file paths. Use @option{^-aI^/SOURCE_SEARCH^}
9741 instead if you just want to specify
9742 source paths only and @option{^-aO^/OBJECT_SEARCH^}
9743 if you want to specify library paths
9747 @command{gnatmake} will ignore any files whose ALI file is write-protected.
9748 This may conveniently be used to exclude standard libraries from
9749 consideration and in particular it means that the use of the
9750 @option{^-f^/FORCE_COMPILE^} switch will not recompile these files
9751 unless @option{^-a^/ALL_FILES^} is also specified.
9754 @command{gnatmake} has been designed to make the use of Ada libraries
9755 particularly convenient. Assume you have an Ada library organized
9756 as follows: @i{^obj-dir^[OBJ_DIR]^} contains the objects and ALI files for
9757 of your Ada compilation units,
9758 whereas @i{^include-dir^[INCLUDE_DIR]^} contains the
9759 specs of these units, but no bodies. Then to compile a unit
9760 stored in @code{main.adb}, which uses this Ada library you would just type
9764 $ gnatmake -aI@var{include-dir} -aL@var{obj-dir} main
9767 $ gnatmake /SOURCE_SEARCH=@i{[INCLUDE_DIR]}
9768 /SKIP_MISSING=@i{[OBJ_DIR]} main
9773 Using @command{gnatmake} along with the
9774 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}
9775 switch provides a mechanism for avoiding unnecessary recompilations. Using
9777 you can update the comments/format of your
9778 source files without having to recompile everything. Note, however, that
9779 adding or deleting lines in a source files may render its debugging
9780 info obsolete. If the file in question is a spec, the impact is rather
9781 limited, as that debugging info will only be useful during the
9782 elaboration phase of your program. For bodies the impact can be more
9783 significant. In all events, your debugger will warn you if a source file
9784 is more recent than the corresponding object, and alert you to the fact
9785 that the debugging information may be out of date.
9788 @node How gnatmake Works
9789 @section How @command{gnatmake} Works
9792 Generally @command{gnatmake} automatically performs all necessary
9793 recompilations and you don't need to worry about how it works. However,
9794 it may be useful to have some basic understanding of the @command{gnatmake}
9795 approach and in particular to understand how it uses the results of
9796 previous compilations without incorrectly depending on them.
9798 First a definition: an object file is considered @dfn{up to date} if the
9799 corresponding ALI file exists and if all the source files listed in the
9800 dependency section of this ALI file have time stamps matching those in
9801 the ALI file. This means that neither the source file itself nor any
9802 files that it depends on have been modified, and hence there is no need
9803 to recompile this file.
9805 @command{gnatmake} works by first checking if the specified main unit is up
9806 to date. If so, no compilations are required for the main unit. If not,
9807 @command{gnatmake} compiles the main program to build a new ALI file that
9808 reflects the latest sources. Then the ALI file of the main unit is
9809 examined to find all the source files on which the main program depends,
9810 and @command{gnatmake} recursively applies the above procedure on all these
9813 This process ensures that @command{gnatmake} only trusts the dependencies
9814 in an existing ALI file if they are known to be correct. Otherwise it
9815 always recompiles to determine a new, guaranteed accurate set of
9816 dependencies. As a result the program is compiled ``upside down'' from what may
9817 be more familiar as the required order of compilation in some other Ada
9818 systems. In particular, clients are compiled before the units on which
9819 they depend. The ability of GNAT to compile in any order is critical in
9820 allowing an order of compilation to be chosen that guarantees that
9821 @command{gnatmake} will recompute a correct set of new dependencies if
9824 When invoking @command{gnatmake} with several @var{file_names}, if a unit is
9825 imported by several of the executables, it will be recompiled at most once.
9827 Note: when using non-standard naming conventions
9828 (@pxref{Using Other File Names}), changing through a configuration pragmas
9829 file the version of a source and invoking @command{gnatmake} to recompile may
9830 have no effect, if the previous version of the source is still accessible
9831 by @command{gnatmake}. It may be necessary to use the switch
9832 ^-f^/FORCE_COMPILE^.
9834 @node Examples of gnatmake Usage
9835 @section Examples of @command{gnatmake} Usage
9838 @item gnatmake hello.adb
9839 Compile all files necessary to bind and link the main program
9840 @file{hello.adb} (containing unit @code{Hello}) and bind and link the
9841 resulting object files to generate an executable file @file{^hello^HELLO.EXE^}.
9843 @item gnatmake main1 main2 main3
9844 Compile all files necessary to bind and link the main programs
9845 @file{main1.adb} (containing unit @code{Main1}), @file{main2.adb}
9846 (containing unit @code{Main2}) and @file{main3.adb}
9847 (containing unit @code{Main3}) and bind and link the resulting object files
9848 to generate three executable files @file{^main1^MAIN1.EXE^},
9849 @file{^main2^MAIN2.EXE^}
9850 and @file{^main3^MAIN3.EXE^}.
9853 @item gnatmake -q Main_Unit -cargs -O2 -bargs -l
9857 @item gnatmake Main_Unit /QUIET
9858 /COMPILER_QUALIFIERS /OPTIMIZE=ALL
9859 /BINDER_QUALIFIERS /ORDER_OF_ELABORATION
9861 Compile all files necessary to bind and link the main program unit
9862 @code{Main_Unit} (from file @file{main_unit.adb}). All compilations will
9863 be done with optimization level 2 and the order of elaboration will be
9864 listed by the binder. @command{gnatmake} will operate in quiet mode, not
9865 displaying commands it is executing.
9868 @c *************************
9869 @node Improving Performance
9870 @chapter Improving Performance
9871 @cindex Improving performance
9874 This chapter presents several topics related to program performance.
9875 It first describes some of the tradeoffs that need to be considered
9876 and some of the techniques for making your program run faster.
9877 It then documents the @command{gnatelim} tool and unused subprogram/data
9878 elimination feature, which can reduce the size of program executables.
9880 Note: to invoke @command{gnatelim} with a project file, use the @code{gnat}
9881 driver (see @ref{The GNAT Driver and Project Files}).
9885 * Performance Considerations::
9886 * Text_IO Suggestions::
9887 * Reducing Size of Ada Executables with gnatelim::
9888 * Reducing Size of Executables with unused subprogram/data elimination::
9892 @c *****************************
9893 @node Performance Considerations
9894 @section Performance Considerations
9897 The GNAT system provides a number of options that allow a trade-off
9902 performance of the generated code
9905 speed of compilation
9908 minimization of dependences and recompilation
9911 the degree of run-time checking.
9915 The defaults (if no options are selected) aim at improving the speed
9916 of compilation and minimizing dependences, at the expense of performance
9917 of the generated code:
9924 no inlining of subprogram calls
9927 all run-time checks enabled except overflow and elaboration checks
9931 These options are suitable for most program development purposes. This
9932 chapter describes how you can modify these choices, and also provides
9933 some guidelines on debugging optimized code.
9936 * Controlling Run-Time Checks::
9937 * Use of Restrictions::
9938 * Optimization Levels::
9939 * Debugging Optimized Code::
9940 * Inlining of Subprograms::
9941 * Other Optimization Switches::
9942 * Optimization and Strict Aliasing::
9945 * Coverage Analysis::
9949 @node Controlling Run-Time Checks
9950 @subsection Controlling Run-Time Checks
9953 By default, GNAT generates all run-time checks, except integer overflow
9954 checks, stack overflow checks, and checks for access before elaboration on
9955 subprogram calls. The latter are not required in default mode, because all
9956 necessary checking is done at compile time.
9957 @cindex @option{-gnatp} (@command{gcc})
9958 @cindex @option{-gnato} (@command{gcc})
9959 Two gnat switches, @option{-gnatp} and @option{-gnato} allow this default to
9960 be modified. @xref{Run-Time Checks}.
9962 Our experience is that the default is suitable for most development
9965 We treat integer overflow specially because these
9966 are quite expensive and in our experience are not as important as other
9967 run-time checks in the development process. Note that division by zero
9968 is not considered an overflow check, and divide by zero checks are
9969 generated where required by default.
9971 Elaboration checks are off by default, and also not needed by default, since
9972 GNAT uses a static elaboration analysis approach that avoids the need for
9973 run-time checking. This manual contains a full chapter discussing the issue
9974 of elaboration checks, and if the default is not satisfactory for your use,
9975 you should read this chapter.
9977 For validity checks, the minimal checks required by the Ada Reference
9978 Manual (for case statements and assignments to array elements) are on
9979 by default. These can be suppressed by use of the @option{-gnatVn} switch.
9980 Note that in Ada 83, there were no validity checks, so if the Ada 83 mode
9981 is acceptable (or when comparing GNAT performance with an Ada 83 compiler),
9982 it may be reasonable to routinely use @option{-gnatVn}. Validity checks
9983 are also suppressed entirely if @option{-gnatp} is used.
9985 @cindex Overflow checks
9986 @cindex Checks, overflow
9989 @cindex pragma Suppress
9990 @cindex pragma Unsuppress
9991 Note that the setting of the switches controls the default setting of
9992 the checks. They may be modified using either @code{pragma Suppress} (to
9993 remove checks) or @code{pragma Unsuppress} (to add back suppressed
9994 checks) in the program source.
9996 @node Use of Restrictions
9997 @subsection Use of Restrictions
10000 The use of pragma Restrictions allows you to control which features are
10001 permitted in your program. Apart from the obvious point that if you avoid
10002 relatively expensive features like finalization (enforceable by the use
10003 of pragma Restrictions (No_Finalization), the use of this pragma does not
10004 affect the generated code in most cases.
10006 One notable exception to this rule is that the possibility of task abort
10007 results in some distributed overhead, particularly if finalization or
10008 exception handlers are used. The reason is that certain sections of code
10009 have to be marked as non-abortable.
10011 If you use neither the @code{abort} statement, nor asynchronous transfer
10012 of control (@code{select @dots{} then abort}), then this distributed overhead
10013 is removed, which may have a general positive effect in improving
10014 overall performance. Especially code involving frequent use of tasking
10015 constructs and controlled types will show much improved performance.
10016 The relevant restrictions pragmas are
10018 @smallexample @c ada
10019 pragma Restrictions (No_Abort_Statements);
10020 pragma Restrictions (Max_Asynchronous_Select_Nesting => 0);
10024 It is recommended that these restriction pragmas be used if possible. Note
10025 that this also means that you can write code without worrying about the
10026 possibility of an immediate abort at any point.
10028 @node Optimization Levels
10029 @subsection Optimization Levels
10030 @cindex @option{^-O^/OPTIMIZE^} (@command{gcc})
10033 Without any optimization ^option,^qualifier,^
10034 the compiler's goal is to reduce the cost of
10035 compilation and to make debugging produce the expected results.
10036 Statements are independent: if you stop the program with a breakpoint between
10037 statements, you can then assign a new value to any variable or change
10038 the program counter to any other statement in the subprogram and get exactly
10039 the results you would expect from the source code.
10041 Turning on optimization makes the compiler attempt to improve the
10042 performance and/or code size at the expense of compilation time and
10043 possibly the ability to debug the program.
10045 If you use multiple
10046 ^-O options, with or without level numbers,^/OPTIMIZE qualifiers,^
10047 the last such option is the one that is effective.
10050 The default is optimization off. This results in the fastest compile
10051 times, but GNAT makes absolutely no attempt to optimize, and the
10052 generated programs are considerably larger and slower than when
10053 optimization is enabled. You can use the
10055 @option{-O} switch (the permitted forms are @option{-O0}, @option{-O1}
10056 @option{-O2}, @option{-O3}, and @option{-Os})
10059 @code{OPTIMIZE} qualifier
10061 to @command{gcc} to control the optimization level:
10064 @item ^-O0^/OPTIMIZE=NONE^
10065 No optimization (the default);
10066 generates unoptimized code but has
10067 the fastest compilation time.
10069 Note that many other compilers do fairly extensive optimization
10070 even if ``no optimization'' is specified. With gcc, it is
10071 very unusual to use ^-O0^/OPTIMIZE=NONE^ for production if
10072 execution time is of any concern, since ^-O0^/OPTIMIZE=NONE^
10073 really does mean no optimization at all. This difference between
10074 gcc and other compilers should be kept in mind when doing
10075 performance comparisons.
10077 @item ^-O1^/OPTIMIZE=SOME^
10078 Moderate optimization;
10079 optimizes reasonably well but does not
10080 degrade compilation time significantly.
10082 @item ^-O2^/OPTIMIZE=ALL^
10084 @itemx /OPTIMIZE=DEVELOPMENT
10087 generates highly optimized code and has
10088 the slowest compilation time.
10090 @item ^-O3^/OPTIMIZE=INLINING^
10091 Full optimization as in @option{-O2},
10092 and also attempts automatic inlining of small
10093 subprograms within a unit (@pxref{Inlining of Subprograms}).
10095 @item ^-Os^/OPTIMIZE=SPACE^
10096 Optimize space usage of resulting program.
10100 Higher optimization levels perform more global transformations on the
10101 program and apply more expensive analysis algorithms in order to generate
10102 faster and more compact code. The price in compilation time, and the
10103 resulting improvement in execution time,
10104 both depend on the particular application and the hardware environment.
10105 You should experiment to find the best level for your application.
10107 Since the precise set of optimizations done at each level will vary from
10108 release to release (and sometime from target to target), it is best to think
10109 of the optimization settings in general terms.
10110 @xref{Optimize Options,, Options That Control Optimization, gcc, Using
10111 the GNU Compiler Collection (GCC)}, for details about
10112 ^the @option{-O} settings and a number of @option{-f} options that^how to^
10113 individually enable or disable specific optimizations.
10115 Unlike some other compilation systems, ^@command{gcc}^GNAT^ has
10116 been tested extensively at all optimization levels. There are some bugs
10117 which appear only with optimization turned on, but there have also been
10118 bugs which show up only in @emph{unoptimized} code. Selecting a lower
10119 level of optimization does not improve the reliability of the code
10120 generator, which in practice is highly reliable at all optimization
10123 Note regarding the use of @option{-O3}: The use of this optimization level
10124 is generally discouraged with GNAT, since it often results in larger
10125 executables which run more slowly. See further discussion of this point
10126 in @ref{Inlining of Subprograms}.
10128 @node Debugging Optimized Code
10129 @subsection Debugging Optimized Code
10130 @cindex Debugging optimized code
10131 @cindex Optimization and debugging
10134 Although it is possible to do a reasonable amount of debugging at
10136 nonzero optimization levels,
10137 the higher the level the more likely that
10140 @option{/OPTIMIZE} settings other than @code{NONE},
10141 such settings will make it more likely that
10143 source-level constructs will have been eliminated by optimization.
10144 For example, if a loop is strength-reduced, the loop
10145 control variable may be completely eliminated and thus cannot be
10146 displayed in the debugger.
10147 This can only happen at @option{-O2} or @option{-O3}.
10148 Explicit temporary variables that you code might be eliminated at
10149 ^level^setting^ @option{-O1} or higher.
10151 The use of the @option{^-g^/DEBUG^} switch,
10152 @cindex @option{^-g^/DEBUG^} (@command{gcc})
10153 which is needed for source-level debugging,
10154 affects the size of the program executable on disk,
10155 and indeed the debugging information can be quite large.
10156 However, it has no effect on the generated code (and thus does not
10157 degrade performance)
10159 Since the compiler generates debugging tables for a compilation unit before
10160 it performs optimizations, the optimizing transformations may invalidate some
10161 of the debugging data. You therefore need to anticipate certain
10162 anomalous situations that may arise while debugging optimized code.
10163 These are the most common cases:
10167 @i{The ``hopping Program Counter'':} Repeated @code{step} or @code{next}
10169 the PC bouncing back and forth in the code. This may result from any of
10170 the following optimizations:
10174 @i{Common subexpression elimination:} using a single instance of code for a
10175 quantity that the source computes several times. As a result you
10176 may not be able to stop on what looks like a statement.
10179 @i{Invariant code motion:} moving an expression that does not change within a
10180 loop, to the beginning of the loop.
10183 @i{Instruction scheduling:} moving instructions so as to
10184 overlap loads and stores (typically) with other code, or in
10185 general to move computations of values closer to their uses. Often
10186 this causes you to pass an assignment statement without the assignment
10187 happening and then later bounce back to the statement when the
10188 value is actually needed. Placing a breakpoint on a line of code
10189 and then stepping over it may, therefore, not always cause all the
10190 expected side-effects.
10194 @i{The ``big leap'':} More commonly known as @emph{cross-jumping}, in which
10195 two identical pieces of code are merged and the program counter suddenly
10196 jumps to a statement that is not supposed to be executed, simply because
10197 it (and the code following) translates to the same thing as the code
10198 that @emph{was} supposed to be executed. This effect is typically seen in
10199 sequences that end in a jump, such as a @code{goto}, a @code{return}, or
10200 a @code{break} in a C @code{^switch^switch^} statement.
10203 @i{The ``roving variable'':} The symptom is an unexpected value in a variable.
10204 There are various reasons for this effect:
10208 In a subprogram prologue, a parameter may not yet have been moved to its
10212 A variable may be dead, and its register re-used. This is
10213 probably the most common cause.
10216 As mentioned above, the assignment of a value to a variable may
10220 A variable may be eliminated entirely by value propagation or
10221 other means. In this case, GCC may incorrectly generate debugging
10222 information for the variable
10226 In general, when an unexpected value appears for a local variable or parameter
10227 you should first ascertain if that value was actually computed by
10228 your program, as opposed to being incorrectly reported by the debugger.
10230 array elements in an object designated by an access value
10231 are generally less of a problem, once you have ascertained that the access
10233 Typically, this means checking variables in the preceding code and in the
10234 calling subprogram to verify that the value observed is explainable from other
10235 values (one must apply the procedure recursively to those
10236 other values); or re-running the code and stopping a little earlier
10237 (perhaps before the call) and stepping to better see how the variable obtained
10238 the value in question; or continuing to step @emph{from} the point of the
10239 strange value to see if code motion had simply moved the variable's
10244 In light of such anomalies, a recommended technique is to use @option{-O0}
10245 early in the software development cycle, when extensive debugging capabilities
10246 are most needed, and then move to @option{-O1} and later @option{-O2} as
10247 the debugger becomes less critical.
10248 Whether to use the @option{^-g^/DEBUG^} switch in the release version is
10249 a release management issue.
10251 Note that if you use @option{-g} you can then use the @command{strip} program
10252 on the resulting executable,
10253 which removes both debugging information and global symbols.
10256 @node Inlining of Subprograms
10257 @subsection Inlining of Subprograms
10260 A call to a subprogram in the current unit is inlined if all the
10261 following conditions are met:
10265 The optimization level is at least @option{-O1}.
10268 The called subprogram is suitable for inlining: It must be small enough
10269 and not contain something that @command{gcc} cannot support in inlined
10273 @cindex pragma Inline
10275 Either @code{pragma Inline} applies to the subprogram, or it is local
10276 to the unit and called once from within it, or it is small and automatic
10277 inlining (optimization level @option{-O3}) is specified.
10281 Calls to subprograms in @code{with}'ed units are normally not inlined.
10282 To achieve actual inlining (that is, replacement of the call by the code
10283 in the body of the subprogram), the following conditions must all be true.
10287 The optimization level is at least @option{-O1}.
10290 The called subprogram is suitable for inlining: It must be small enough
10291 and not contain something that @command{gcc} cannot support in inlined
10295 The call appears in a body (not in a package spec).
10298 There is a @code{pragma Inline} for the subprogram.
10301 @cindex @option{-gnatn} (@command{gcc})
10302 The @option{^-gnatn^/INLINE^} switch
10303 is used in the @command{gcc} command line
10306 Even if all these conditions are met, it may not be possible for
10307 the compiler to inline the call, due to the length of the body,
10308 or features in the body that make it impossible for the compiler
10309 to do the inlining.
10311 Note that specifying the @option{-gnatn} switch causes additional
10312 compilation dependencies. Consider the following:
10314 @smallexample @c ada
10334 With the default behavior (no @option{-gnatn} switch specified), the
10335 compilation of the @code{Main} procedure depends only on its own source,
10336 @file{main.adb}, and the spec of the package in file @file{r.ads}. This
10337 means that editing the body of @code{R} does not require recompiling
10340 On the other hand, the call @code{R.Q} is not inlined under these
10341 circumstances. If the @option{-gnatn} switch is present when @code{Main}
10342 is compiled, the call will be inlined if the body of @code{Q} is small
10343 enough, but now @code{Main} depends on the body of @code{R} in
10344 @file{r.adb} as well as on the spec. This means that if this body is edited,
10345 the main program must be recompiled. Note that this extra dependency
10346 occurs whether or not the call is in fact inlined by @command{gcc}.
10348 The use of front end inlining with @option{-gnatN} generates similar
10349 additional dependencies.
10351 @cindex @option{^-fno-inline^/INLINE=SUPPRESS^} (@command{gcc})
10352 Note: The @option{^-fno-inline^/INLINE=SUPPRESS^} switch
10353 can be used to prevent
10354 all inlining. This switch overrides all other conditions and ensures
10355 that no inlining occurs. The extra dependences resulting from
10356 @option{-gnatn} will still be active, even if
10357 this switch is used to suppress the resulting inlining actions.
10359 @cindex @option{-fno-inline-functions} (@command{gcc})
10360 Note: The @option{-fno-inline-functions} switch can be used to prevent
10361 automatic inlining of small subprograms if @option{-O3} is used.
10363 @cindex @option{-fno-inline-functions-called-once} (@command{gcc})
10364 Note: The @option{-fno-inline-functions-called-once} switch
10365 can be used to prevent inlining of subprograms local to the unit
10366 and called once from within it if @option{-O1} is used.
10368 Note regarding the use of @option{-O3}: There is no difference in inlining
10369 behavior between @option{-O2} and @option{-O3} for subprograms with an explicit
10370 pragma @code{Inline} assuming the use of @option{-gnatn}
10371 or @option{-gnatN} (the switches that activate inlining). If you have used
10372 pragma @code{Inline} in appropriate cases, then it is usually much better
10373 to use @option{-O2} and @option{-gnatn} and avoid the use of @option{-O3} which
10374 in this case only has the effect of inlining subprograms you did not
10375 think should be inlined. We often find that the use of @option{-O3} slows
10376 down code by performing excessive inlining, leading to increased instruction
10377 cache pressure from the increased code size. So the bottom line here is
10378 that you should not automatically assume that @option{-O3} is better than
10379 @option{-O2}, and indeed you should use @option{-O3} only if tests show that
10380 it actually improves performance.
10382 @node Other Optimization Switches
10383 @subsection Other Optimization Switches
10384 @cindex Optimization Switches
10386 Since @code{GNAT} uses the @command{gcc} back end, all the specialized
10387 @command{gcc} optimization switches are potentially usable. These switches
10388 have not been extensively tested with GNAT but can generally be expected
10389 to work. Examples of switches in this category are
10390 @option{-funroll-loops} and
10391 the various target-specific @option{-m} options (in particular, it has been
10392 observed that @option{-march=pentium4} can significantly improve performance
10393 on appropriate machines). For full details of these switches, see
10394 @ref{Submodel Options,, Hardware Models and Configurations, gcc, Using
10395 the GNU Compiler Collection (GCC)}.
10397 @node Optimization and Strict Aliasing
10398 @subsection Optimization and Strict Aliasing
10400 @cindex Strict Aliasing
10401 @cindex No_Strict_Aliasing
10404 The strong typing capabilities of Ada allow an optimizer to generate
10405 efficient code in situations where other languages would be forced to
10406 make worst case assumptions preventing such optimizations. Consider
10407 the following example:
10409 @smallexample @c ada
10412 type Int1 is new Integer;
10413 type Int2 is new Integer;
10414 type Int1A is access Int1;
10415 type Int2A is access Int2;
10422 for J in Data'Range loop
10423 if Data (J) = Int1V.all then
10424 Int2V.all := Int2V.all + 1;
10433 In this example, since the variable @code{Int1V} can only access objects
10434 of type @code{Int1}, and @code{Int2V} can only access objects of type
10435 @code{Int2}, there is no possibility that the assignment to
10436 @code{Int2V.all} affects the value of @code{Int1V.all}. This means that
10437 the compiler optimizer can "know" that the value @code{Int1V.all} is constant
10438 for all iterations of the loop and avoid the extra memory reference
10439 required to dereference it each time through the loop.
10441 This kind of optimization, called strict aliasing analysis, is
10442 triggered by specifying an optimization level of @option{-O2} or
10443 higher or @option{-Os} and allows @code{GNAT} to generate more efficient code
10444 when access values are involved.
10446 However, although this optimization is always correct in terms of
10447 the formal semantics of the Ada Reference Manual, difficulties can
10448 arise if features like @code{Unchecked_Conversion} are used to break
10449 the typing system. Consider the following complete program example:
10451 @smallexample @c ada
10454 type int1 is new integer;
10455 type int2 is new integer;
10456 type a1 is access int1;
10457 type a2 is access int2;
10462 function to_a2 (Input : a1) return a2;
10465 with Unchecked_Conversion;
10467 function to_a2 (Input : a1) return a2 is
10469 new Unchecked_Conversion (a1, a2);
10471 return to_a2u (Input);
10477 with Text_IO; use Text_IO;
10479 v1 : a1 := new int1;
10480 v2 : a2 := to_a2 (v1);
10484 put_line (int1'image (v1.all));
10490 This program prints out 0 in @option{-O0} or @option{-O1}
10491 mode, but it prints out 1 in @option{-O2} mode. That's
10492 because in strict aliasing mode, the compiler can and
10493 does assume that the assignment to @code{v2.all} could not
10494 affect the value of @code{v1.all}, since different types
10497 This behavior is not a case of non-conformance with the standard, since
10498 the Ada RM specifies that an unchecked conversion where the resulting
10499 bit pattern is not a correct value of the target type can result in an
10500 abnormal value and attempting to reference an abnormal value makes the
10501 execution of a program erroneous. That's the case here since the result
10502 does not point to an object of type @code{int2}. This means that the
10503 effect is entirely unpredictable.
10505 However, although that explanation may satisfy a language
10506 lawyer, in practice an applications programmer expects an
10507 unchecked conversion involving pointers to create true
10508 aliases and the behavior of printing 1 seems plain wrong.
10509 In this case, the strict aliasing optimization is unwelcome.
10511 Indeed the compiler recognizes this possibility, and the
10512 unchecked conversion generates a warning:
10515 p2.adb:5:07: warning: possible aliasing problem with type "a2"
10516 p2.adb:5:07: warning: use -fno-strict-aliasing switch for references
10517 p2.adb:5:07: warning: or use "pragma No_Strict_Aliasing (a2);"
10521 Unfortunately the problem is recognized when compiling the body of
10522 package @code{p2}, but the actual "bad" code is generated while
10523 compiling the body of @code{m} and this latter compilation does not see
10524 the suspicious @code{Unchecked_Conversion}.
10526 As implied by the warning message, there are approaches you can use to
10527 avoid the unwanted strict aliasing optimization in a case like this.
10529 One possibility is to simply avoid the use of @option{-O2}, but
10530 that is a bit drastic, since it throws away a number of useful
10531 optimizations that do not involve strict aliasing assumptions.
10533 A less drastic approach is to compile the program using the
10534 option @option{-fno-strict-aliasing}. Actually it is only the
10535 unit containing the dereferencing of the suspicious pointer
10536 that needs to be compiled. So in this case, if we compile
10537 unit @code{m} with this switch, then we get the expected
10538 value of zero printed. Analyzing which units might need
10539 the switch can be painful, so a more reasonable approach
10540 is to compile the entire program with options @option{-O2}
10541 and @option{-fno-strict-aliasing}. If the performance is
10542 satisfactory with this combination of options, then the
10543 advantage is that the entire issue of possible "wrong"
10544 optimization due to strict aliasing is avoided.
10546 To avoid the use of compiler switches, the configuration
10547 pragma @code{No_Strict_Aliasing} with no parameters may be
10548 used to specify that for all access types, the strict
10549 aliasing optimization should be suppressed.
10551 However, these approaches are still overkill, in that they causes
10552 all manipulations of all access values to be deoptimized. A more
10553 refined approach is to concentrate attention on the specific
10554 access type identified as problematic.
10556 First, if a careful analysis of uses of the pointer shows
10557 that there are no possible problematic references, then
10558 the warning can be suppressed by bracketing the
10559 instantiation of @code{Unchecked_Conversion} to turn
10562 @smallexample @c ada
10563 pragma Warnings (Off);
10565 new Unchecked_Conversion (a1, a2);
10566 pragma Warnings (On);
10570 Of course that approach is not appropriate for this particular
10571 example, since indeed there is a problematic reference. In this
10572 case we can take one of two other approaches.
10574 The first possibility is to move the instantiation of unchecked
10575 conversion to the unit in which the type is declared. In
10576 this example, we would move the instantiation of
10577 @code{Unchecked_Conversion} from the body of package
10578 @code{p2} to the spec of package @code{p1}. Now the
10579 warning disappears. That's because any use of the
10580 access type knows there is a suspicious unchecked
10581 conversion, and the strict aliasing optimization
10582 is automatically suppressed for the type.
10584 If it is not practical to move the unchecked conversion to the same unit
10585 in which the destination access type is declared (perhaps because the
10586 source type is not visible in that unit), you may use pragma
10587 @code{No_Strict_Aliasing} for the type. This pragma must occur in the
10588 same declarative sequence as the declaration of the access type:
10590 @smallexample @c ada
10591 type a2 is access int2;
10592 pragma No_Strict_Aliasing (a2);
10596 Here again, the compiler now knows that the strict aliasing optimization
10597 should be suppressed for any reference to type @code{a2} and the
10598 expected behavior is obtained.
10600 Finally, note that although the compiler can generate warnings for
10601 simple cases of unchecked conversions, there are tricker and more
10602 indirect ways of creating type incorrect aliases which the compiler
10603 cannot detect. Examples are the use of address overlays and unchecked
10604 conversions involving composite types containing access types as
10605 components. In such cases, no warnings are generated, but there can
10606 still be aliasing problems. One safe coding practice is to forbid the
10607 use of address clauses for type overlaying, and to allow unchecked
10608 conversion only for primitive types. This is not really a significant
10609 restriction since any possible desired effect can be achieved by
10610 unchecked conversion of access values.
10612 The aliasing analysis done in strict aliasing mode can certainly
10613 have significant benefits. We have seen cases of large scale
10614 application code where the time is increased by up to 5% by turning
10615 this optimization off. If you have code that includes significant
10616 usage of unchecked conversion, you might want to just stick with
10617 @option{-O1} and avoid the entire issue. If you get adequate
10618 performance at this level of optimization level, that's probably
10619 the safest approach. If tests show that you really need higher
10620 levels of optimization, then you can experiment with @option{-O2}
10621 and @option{-O2 -fno-strict-aliasing} to see how much effect this
10622 has on size and speed of the code. If you really need to use
10623 @option{-O2} with strict aliasing in effect, then you should
10624 review any uses of unchecked conversion of access types,
10625 particularly if you are getting the warnings described above.
10628 @node Coverage Analysis
10629 @subsection Coverage Analysis
10632 GNAT supports the HP Performance Coverage Analyzer (PCA), which allows
10633 the user to determine the distribution of execution time across a program,
10634 @pxref{Profiling} for details of usage.
10638 @node Text_IO Suggestions
10639 @section @code{Text_IO} Suggestions
10640 @cindex @code{Text_IO} and performance
10643 The @code{Ada.Text_IO} package has fairly high overheads due in part to
10644 the requirement of maintaining page and line counts. If performance
10645 is critical, a recommendation is to use @code{Stream_IO} instead of
10646 @code{Text_IO} for volume output, since this package has less overhead.
10648 If @code{Text_IO} must be used, note that by default output to the standard
10649 output and standard error files is unbuffered (this provides better
10650 behavior when output statements are used for debugging, or if the
10651 progress of a program is observed by tracking the output, e.g. by
10652 using the Unix @command{tail -f} command to watch redirected output.
10654 If you are generating large volumes of output with @code{Text_IO} and
10655 performance is an important factor, use a designated file instead
10656 of the standard output file, or change the standard output file to
10657 be buffered using @code{Interfaces.C_Streams.setvbuf}.
10661 @node Reducing Size of Ada Executables with gnatelim
10662 @section Reducing Size of Ada Executables with @code{gnatelim}
10666 This section describes @command{gnatelim}, a tool which detects unused
10667 subprograms and helps the compiler to create a smaller executable for your
10672 * Running gnatelim::
10673 * Correcting the List of Eliminate Pragmas::
10674 * Making Your Executables Smaller::
10675 * Summary of the gnatelim Usage Cycle::
10678 @node About gnatelim
10679 @subsection About @code{gnatelim}
10682 When a program shares a set of Ada
10683 packages with other programs, it may happen that this program uses
10684 only a fraction of the subprograms defined in these packages. The code
10685 created for these unused subprograms increases the size of the executable.
10687 @code{gnatelim} tracks unused subprograms in an Ada program and
10688 outputs a list of GNAT-specific pragmas @code{Eliminate} marking all the
10689 subprograms that are declared but never called. By placing the list of
10690 @code{Eliminate} pragmas in the GNAT configuration file @file{gnat.adc} and
10691 recompiling your program, you may decrease the size of its executable,
10692 because the compiler will not generate the code for 'eliminated' subprograms.
10693 @xref{Pragma Eliminate,,, gnat_rm, GNAT Reference Manual}, for more
10694 information about this pragma.
10696 @code{gnatelim} needs as its input data the name of the main subprogram
10697 and a bind file for a main subprogram.
10699 To create a bind file for @code{gnatelim}, run @code{gnatbind} for
10700 the main subprogram. @code{gnatelim} can work with both Ada and C
10701 bind files; when both are present, it uses the Ada bind file.
10702 The following commands will build the program and create the bind file:
10705 $ gnatmake ^-c Main_Prog^/ACTIONS=COMPILE MAIN_PROG^
10706 $ gnatbind main_prog
10709 Note that @code{gnatelim} needs neither object nor ALI files.
10711 @node Running gnatelim
10712 @subsection Running @code{gnatelim}
10715 @code{gnatelim} has the following command-line interface:
10718 $ gnatelim @ovar{options} name
10722 @code{name} should be a name of a source file that contains the main subprogram
10723 of a program (partition).
10725 @code{gnatelim} has the following switches:
10730 @cindex @option{^-q^/QUIET^} (@command{gnatelim})
10731 Quiet mode: by default @code{gnatelim} outputs to the standard error
10732 stream the number of program units left to be processed. This option turns
10735 @item ^-v^/VERBOSE^
10736 @cindex @option{^-v^/VERBOSE^} (@command{gnatelim})
10737 Verbose mode: @code{gnatelim} version information is printed as Ada
10738 comments to the standard output stream. Also, in addition to the number of
10739 program units left @code{gnatelim} will output the name of the current unit
10743 @cindex @option{^-a^/ALL^} (@command{gnatelim})
10744 Also look for subprograms from the GNAT run time that can be eliminated. Note
10745 that when @file{gnat.adc} is produced using this switch, the entire program
10746 must be recompiled with switch @option{^-a^/ALL_FILES^} to @command{gnatmake}.
10748 @item ^-I^/INCLUDE_DIRS=^@var{dir}
10749 @cindex @option{^-I^/INCLUDE_DIRS^} (@command{gnatelim})
10750 When looking for source files also look in directory @var{dir}. Specifying
10751 @option{^-I-^/INCLUDE_DIRS=-^} instructs @code{gnatelim} not to look for
10752 sources in the current directory.
10754 @item ^-b^/BIND_FILE=^@var{bind_file}
10755 @cindex @option{^-b^/BIND_FILE^} (@command{gnatelim})
10756 Specifies @var{bind_file} as the bind file to process. If not set, the name
10757 of the bind file is computed from the full expanded Ada name
10758 of a main subprogram.
10760 @item ^-C^/CONFIG_FILE=^@var{config_file}
10761 @cindex @option{^-C^/CONFIG_FILE^} (@command{gnatelim})
10762 Specifies a file @var{config_file} that contains configuration pragmas. The
10763 file must be specified with full path.
10765 @item ^--GCC^/COMPILER^=@var{compiler_name}
10766 @cindex @option{^-GCC^/COMPILER^} (@command{gnatelim})
10767 Instructs @code{gnatelim} to use specific @command{gcc} compiler instead of one
10768 available on the path.
10770 @item ^--GNATMAKE^/GNATMAKE^=@var{gnatmake_name}
10771 @cindex @option{^--GNATMAKE^/GNATMAKE^} (@command{gnatelim})
10772 Instructs @code{gnatelim} to use specific @command{gnatmake} instead of one
10773 available on the path.
10777 @code{gnatelim} sends its output to the standard output stream, and all the
10778 tracing and debug information is sent to the standard error stream.
10779 In order to produce a proper GNAT configuration file
10780 @file{gnat.adc}, redirection must be used:
10784 $ PIPE GNAT ELIM MAIN_PROG.ADB > GNAT.ADC
10787 $ gnatelim main_prog.adb > gnat.adc
10796 $ gnatelim main_prog.adb >> gnat.adc
10800 in order to append the @code{gnatelim} output to the existing contents of
10804 @node Correcting the List of Eliminate Pragmas
10805 @subsection Correcting the List of Eliminate Pragmas
10808 In some rare cases @code{gnatelim} may try to eliminate
10809 subprograms that are actually called in the program. In this case, the
10810 compiler will generate an error message of the form:
10813 file.adb:106:07: cannot call eliminated subprogram "My_Prog"
10817 You will need to manually remove the wrong @code{Eliminate} pragmas from
10818 the @file{gnat.adc} file. You should recompile your program
10819 from scratch after that, because you need a consistent @file{gnat.adc} file
10820 during the entire compilation.
10822 @node Making Your Executables Smaller
10823 @subsection Making Your Executables Smaller
10826 In order to get a smaller executable for your program you now have to
10827 recompile the program completely with the new @file{gnat.adc} file
10828 created by @code{gnatelim} in your current directory:
10831 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10835 (Use the @option{^-f^/FORCE_COMPILE^} option for @command{gnatmake} to
10836 recompile everything
10837 with the set of pragmas @code{Eliminate} that you have obtained with
10838 @command{gnatelim}).
10840 Be aware that the set of @code{Eliminate} pragmas is specific to each
10841 program. It is not recommended to merge sets of @code{Eliminate}
10842 pragmas created for different programs in one @file{gnat.adc} file.
10844 @node Summary of the gnatelim Usage Cycle
10845 @subsection Summary of the gnatelim Usage Cycle
10848 Here is a quick summary of the steps to be taken in order to reduce
10849 the size of your executables with @code{gnatelim}. You may use
10850 other GNAT options to control the optimization level,
10851 to produce the debugging information, to set search path, etc.
10855 Produce a bind file
10858 $ gnatmake ^-c main_prog^/ACTIONS=COMPILE MAIN_PROG^
10859 $ gnatbind main_prog
10863 Generate a list of @code{Eliminate} pragmas
10866 $ PIPE GNAT ELIM MAIN_PROG > GNAT.ADC
10869 $ gnatelim main_prog >@r{[}>@r{]} gnat.adc
10874 Recompile the application
10877 $ gnatmake ^-f main_prog^/FORCE_COMPILE MAIN_PROG^
10882 @node Reducing Size of Executables with unused subprogram/data elimination
10883 @section Reducing Size of Executables with Unused Subprogram/Data Elimination
10884 @findex unused subprogram/data elimination
10887 This section describes how you can eliminate unused subprograms and data from
10888 your executable just by setting options at compilation time.
10891 * About unused subprogram/data elimination::
10892 * Compilation options::
10893 * Example of unused subprogram/data elimination::
10896 @node About unused subprogram/data elimination
10897 @subsection About unused subprogram/data elimination
10900 By default, an executable contains all code and data of its composing objects
10901 (directly linked or coming from statically linked libraries), even data or code
10902 never used by this executable.
10904 This feature will allow you to eliminate such unused code from your
10905 executable, making it smaller (in disk and in memory).
10907 This functionality is available on all Linux platforms except for the IA-64
10908 architecture and on all cross platforms using the ELF binary file format.
10909 In both cases GNU binutils version 2.16 or later are required to enable it.
10911 @node Compilation options
10912 @subsection Compilation options
10915 The operation of eliminating the unused code and data from the final executable
10916 is directly performed by the linker.
10918 In order to do this, it has to work with objects compiled with the
10920 @option{-ffunction-sections} @option{-fdata-sections}.
10921 @cindex @option{-ffunction-sections} (@command{gcc})
10922 @cindex @option{-fdata-sections} (@command{gcc})
10923 These options are usable with C and Ada files.
10924 They will place respectively each
10925 function or data in a separate section in the resulting object file.
10927 Once the objects and static libraries are created with these options, the
10928 linker can perform the dead code elimination. You can do this by setting
10929 the @option{-Wl,--gc-sections} option to gcc command or in the
10930 @option{-largs} section of @command{gnatmake}. This will perform a
10931 garbage collection of code and data never referenced.
10933 If the linker performs a partial link (@option{-r} ld linker option), then you
10934 will need to provide one or several entry point using the
10935 @option{-e} / @option{--entry} ld option.
10937 Note that objects compiled without the @option{-ffunction-sections} and
10938 @option{-fdata-sections} options can still be linked with the executable.
10939 However, no dead code elimination will be performed on those objects (they will
10942 The GNAT static library is now compiled with -ffunction-sections and
10943 -fdata-sections on some platforms. This allows you to eliminate the unused code
10944 and data of the GNAT library from your executable.
10946 @node Example of unused subprogram/data elimination
10947 @subsection Example of unused subprogram/data elimination
10950 Here is a simple example:
10952 @smallexample @c ada
10961 Used_Data : Integer;
10962 Unused_Data : Integer;
10964 procedure Used (Data : Integer);
10965 procedure Unused (Data : Integer);
10968 package body Aux is
10969 procedure Used (Data : Integer) is
10974 procedure Unused (Data : Integer) is
10976 Unused_Data := Data;
10982 @code{Unused} and @code{Unused_Data} are never referenced in this code
10983 excerpt, and hence they may be safely removed from the final executable.
10988 $ nm test | grep used
10989 020015f0 T aux__unused
10990 02005d88 B aux__unused_data
10991 020015cc T aux__used
10992 02005d84 B aux__used_data
10994 $ gnatmake test -cargs -fdata-sections -ffunction-sections \
10995 -largs -Wl,--gc-sections
10997 $ nm test | grep used
10998 02005350 T aux__used
10999 0201ffe0 B aux__used_data
11003 It can be observed that the procedure @code{Unused} and the object
11004 @code{Unused_Data} are removed by the linker when using the
11005 appropriate options.
11007 @c ********************************
11008 @node Renaming Files Using gnatchop
11009 @chapter Renaming Files Using @code{gnatchop}
11013 This chapter discusses how to handle files with multiple units by using
11014 the @code{gnatchop} utility. This utility is also useful in renaming
11015 files to meet the standard GNAT default file naming conventions.
11018 * Handling Files with Multiple Units::
11019 * Operating gnatchop in Compilation Mode::
11020 * Command Line for gnatchop::
11021 * Switches for gnatchop::
11022 * Examples of gnatchop Usage::
11025 @node Handling Files with Multiple Units
11026 @section Handling Files with Multiple Units
11029 The basic compilation model of GNAT requires that a file submitted to the
11030 compiler have only one unit and there be a strict correspondence
11031 between the file name and the unit name.
11033 The @code{gnatchop} utility allows both of these rules to be relaxed,
11034 allowing GNAT to process files which contain multiple compilation units
11035 and files with arbitrary file names. @code{gnatchop}
11036 reads the specified file and generates one or more output files,
11037 containing one unit per file. The unit and the file name correspond,
11038 as required by GNAT.
11040 If you want to permanently restructure a set of ``foreign'' files so that
11041 they match the GNAT rules, and do the remaining development using the
11042 GNAT structure, you can simply use @command{gnatchop} once, generate the
11043 new set of files and work with them from that point on.
11045 Alternatively, if you want to keep your files in the ``foreign'' format,
11046 perhaps to maintain compatibility with some other Ada compilation
11047 system, you can set up a procedure where you use @command{gnatchop} each
11048 time you compile, regarding the source files that it writes as temporary
11049 files that you throw away.
11051 Note that if your file containing multiple units starts with a byte order
11052 mark (BOM) specifying UTF-8 encoding, then the files generated by gnatchop
11053 will each start with a copy of this BOM, meaning that they can be compiled
11054 automatically in UTF-8 mode without needing to specify an explicit encoding.
11056 @node Operating gnatchop in Compilation Mode
11057 @section Operating gnatchop in Compilation Mode
11060 The basic function of @code{gnatchop} is to take a file with multiple units
11061 and split it into separate files. The boundary between files is reasonably
11062 clear, except for the issue of comments and pragmas. In default mode, the
11063 rule is that any pragmas between units belong to the previous unit, except
11064 that configuration pragmas always belong to the following unit. Any comments
11065 belong to the following unit. These rules
11066 almost always result in the right choice of
11067 the split point without needing to mark it explicitly and most users will
11068 find this default to be what they want. In this default mode it is incorrect to
11069 submit a file containing only configuration pragmas, or one that ends in
11070 configuration pragmas, to @code{gnatchop}.
11072 However, using a special option to activate ``compilation mode'',
11074 can perform another function, which is to provide exactly the semantics
11075 required by the RM for handling of configuration pragmas in a compilation.
11076 In the absence of configuration pragmas (at the main file level), this
11077 option has no effect, but it causes such configuration pragmas to be handled
11078 in a quite different manner.
11080 First, in compilation mode, if @code{gnatchop} is given a file that consists of
11081 only configuration pragmas, then this file is appended to the
11082 @file{gnat.adc} file in the current directory. This behavior provides
11083 the required behavior described in the RM for the actions to be taken
11084 on submitting such a file to the compiler, namely that these pragmas
11085 should apply to all subsequent compilations in the same compilation
11086 environment. Using GNAT, the current directory, possibly containing a
11087 @file{gnat.adc} file is the representation
11088 of a compilation environment. For more information on the
11089 @file{gnat.adc} file, see @ref{Handling of Configuration Pragmas}.
11091 Second, in compilation mode, if @code{gnatchop}
11092 is given a file that starts with
11093 configuration pragmas, and contains one or more units, then these
11094 configuration pragmas are prepended to each of the chopped files. This
11095 behavior provides the required behavior described in the RM for the
11096 actions to be taken on compiling such a file, namely that the pragmas
11097 apply to all units in the compilation, but not to subsequently compiled
11100 Finally, if configuration pragmas appear between units, they are appended
11101 to the previous unit. This results in the previous unit being illegal,
11102 since the compiler does not accept configuration pragmas that follow
11103 a unit. This provides the required RM behavior that forbids configuration
11104 pragmas other than those preceding the first compilation unit of a
11107 For most purposes, @code{gnatchop} will be used in default mode. The
11108 compilation mode described above is used only if you need exactly
11109 accurate behavior with respect to compilations, and you have files
11110 that contain multiple units and configuration pragmas. In this
11111 circumstance the use of @code{gnatchop} with the compilation mode
11112 switch provides the required behavior, and is for example the mode
11113 in which GNAT processes the ACVC tests.
11115 @node Command Line for gnatchop
11116 @section Command Line for @code{gnatchop}
11119 The @code{gnatchop} command has the form:
11122 $ gnatchop switches @var{file name} @r{[}@var{file name} @dots{}@r{]}
11127 The only required argument is the file name of the file to be chopped.
11128 There are no restrictions on the form of this file name. The file itself
11129 contains one or more Ada units, in normal GNAT format, concatenated
11130 together. As shown, more than one file may be presented to be chopped.
11132 When run in default mode, @code{gnatchop} generates one output file in
11133 the current directory for each unit in each of the files.
11135 @var{directory}, if specified, gives the name of the directory to which
11136 the output files will be written. If it is not specified, all files are
11137 written to the current directory.
11139 For example, given a
11140 file called @file{hellofiles} containing
11142 @smallexample @c ada
11147 with Text_IO; use Text_IO;
11150 Put_Line ("Hello");
11160 $ gnatchop ^hellofiles^HELLOFILES.^
11164 generates two files in the current directory, one called
11165 @file{hello.ads} containing the single line that is the procedure spec,
11166 and the other called @file{hello.adb} containing the remaining text. The
11167 original file is not affected. The generated files can be compiled in
11171 When gnatchop is invoked on a file that is empty or that contains only empty
11172 lines and/or comments, gnatchop will not fail, but will not produce any
11175 For example, given a
11176 file called @file{toto.txt} containing
11178 @smallexample @c ada
11190 $ gnatchop ^toto.txt^TOT.TXT^
11194 will not produce any new file and will result in the following warnings:
11197 toto.txt:1:01: warning: empty file, contains no compilation units
11198 no compilation units found
11199 no source files written
11202 @node Switches for gnatchop
11203 @section Switches for @code{gnatchop}
11206 @command{gnatchop} recognizes the following switches:
11212 @cindex @option{--version} @command{gnatchop}
11213 Display Copyright and version, then exit disregarding all other options.
11216 @cindex @option{--help} @command{gnatchop}
11217 If @option{--version} was not used, display usage, then exit disregarding
11220 @item ^-c^/COMPILATION^
11221 @cindex @option{^-c^/COMPILATION^} (@code{gnatchop})
11222 Causes @code{gnatchop} to operate in compilation mode, in which
11223 configuration pragmas are handled according to strict RM rules. See
11224 previous section for a full description of this mode.
11227 @item -gnat@var{xxx}
11228 This passes the given @option{-gnat@var{xxx}} switch to @code{gnat} which is
11229 used to parse the given file. Not all @var{xxx} options make sense,
11230 but for example, the use of @option{-gnati2} allows @code{gnatchop} to
11231 process a source file that uses Latin-2 coding for identifiers.
11235 Causes @code{gnatchop} to generate a brief help summary to the standard
11236 output file showing usage information.
11238 @item ^-k@var{mm}^/FILE_NAME_MAX_LENGTH=@var{mm}^
11239 @cindex @option{^-k^/FILE_NAME_MAX_LENGTH^} (@code{gnatchop})
11240 Limit generated file names to the specified number @code{mm}
11242 This is useful if the
11243 resulting set of files is required to be interoperable with systems
11244 which limit the length of file names.
11246 If no value is given, or
11247 if no @code{/FILE_NAME_MAX_LENGTH} qualifier is given,
11248 a default of 39, suitable for OpenVMS Alpha
11249 Systems, is assumed
11252 No space is allowed between the @option{-k} and the numeric value. The numeric
11253 value may be omitted in which case a default of @option{-k8},
11255 with DOS-like file systems, is used. If no @option{-k} switch
11257 there is no limit on the length of file names.
11260 @item ^-p^/PRESERVE^
11261 @cindex @option{^-p^/PRESERVE^} (@code{gnatchop})
11262 Causes the file ^modification^creation^ time stamp of the input file to be
11263 preserved and used for the time stamp of the output file(s). This may be
11264 useful for preserving coherency of time stamps in an environment where
11265 @code{gnatchop} is used as part of a standard build process.
11268 @cindex @option{^-q^/QUIET^} (@code{gnatchop})
11269 Causes output of informational messages indicating the set of generated
11270 files to be suppressed. Warnings and error messages are unaffected.
11272 @item ^-r^/REFERENCE^
11273 @cindex @option{^-r^/REFERENCE^} (@code{gnatchop})
11274 @findex Source_Reference
11275 Generate @code{Source_Reference} pragmas. Use this switch if the output
11276 files are regarded as temporary and development is to be done in terms
11277 of the original unchopped file. This switch causes
11278 @code{Source_Reference} pragmas to be inserted into each of the
11279 generated files to refers back to the original file name and line number.
11280 The result is that all error messages refer back to the original
11282 In addition, the debugging information placed into the object file (when
11283 the @option{^-g^/DEBUG^} switch of @command{gcc} or @command{gnatmake} is
11285 also refers back to this original file so that tools like profilers and
11286 debuggers will give information in terms of the original unchopped file.
11288 If the original file to be chopped itself contains
11289 a @code{Source_Reference}
11290 pragma referencing a third file, then gnatchop respects
11291 this pragma, and the generated @code{Source_Reference} pragmas
11292 in the chopped file refer to the original file, with appropriate
11293 line numbers. This is particularly useful when @code{gnatchop}
11294 is used in conjunction with @code{gnatprep} to compile files that
11295 contain preprocessing statements and multiple units.
11297 @item ^-v^/VERBOSE^
11298 @cindex @option{^-v^/VERBOSE^} (@code{gnatchop})
11299 Causes @code{gnatchop} to operate in verbose mode. The version
11300 number and copyright notice are output, as well as exact copies of
11301 the gnat1 commands spawned to obtain the chop control information.
11303 @item ^-w^/OVERWRITE^
11304 @cindex @option{^-w^/OVERWRITE^} (@code{gnatchop})
11305 Overwrite existing file names. Normally @code{gnatchop} regards it as a
11306 fatal error if there is already a file with the same name as a
11307 file it would otherwise output, in other words if the files to be
11308 chopped contain duplicated units. This switch bypasses this
11309 check, and causes all but the last instance of such duplicated
11310 units to be skipped.
11313 @item --GCC=@var{xxxx}
11314 @cindex @option{--GCC=} (@code{gnatchop})
11315 Specify the path of the GNAT parser to be used. When this switch is used,
11316 no attempt is made to add the prefix to the GNAT parser executable.
11320 @node Examples of gnatchop Usage
11321 @section Examples of @code{gnatchop} Usage
11325 @item gnatchop /OVERWRITE HELLO_S.ADA [PRERELEASE.FILES]
11328 @item gnatchop -w hello_s.ada prerelease/files
11331 Chops the source file @file{hello_s.ada}. The output files will be
11332 placed in the directory @file{^prerelease/files^[PRERELEASE.FILES]^},
11334 files with matching names in that directory (no files in the current
11335 directory are modified).
11337 @item gnatchop ^archive^ARCHIVE.^
11338 Chops the source file @file{^archive^ARCHIVE.^}
11339 into the current directory. One
11340 useful application of @code{gnatchop} is in sending sets of sources
11341 around, for example in email messages. The required sources are simply
11342 concatenated (for example, using a ^Unix @code{cat}^VMS @code{APPEND/NEW}^
11344 @command{gnatchop} is used at the other end to reconstitute the original
11347 @item gnatchop file1 file2 file3 direc
11348 Chops all units in files @file{file1}, @file{file2}, @file{file3}, placing
11349 the resulting files in the directory @file{direc}. Note that if any units
11350 occur more than once anywhere within this set of files, an error message
11351 is generated, and no files are written. To override this check, use the
11352 @option{^-w^/OVERWRITE^} switch,
11353 in which case the last occurrence in the last file will
11354 be the one that is output, and earlier duplicate occurrences for a given
11355 unit will be skipped.
11358 @node Configuration Pragmas
11359 @chapter Configuration Pragmas
11360 @cindex Configuration pragmas
11361 @cindex Pragmas, configuration
11364 Configuration pragmas include those pragmas described as
11365 such in the Ada Reference Manual, as well as
11366 implementation-dependent pragmas that are configuration pragmas.
11367 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
11368 for details on these additional GNAT-specific configuration pragmas.
11369 Most notably, the pragma @code{Source_File_Name}, which allows
11370 specifying non-default names for source files, is a configuration
11371 pragma. The following is a complete list of configuration pragmas
11372 recognized by GNAT:
11380 Assume_No_Invalid_Values
11385 Compile_Time_Warning
11387 Component_Alignment
11388 Convention_Identifier
11396 External_Name_Casing
11399 Float_Representation
11412 Priority_Specific_Dispatching
11415 Propagate_Exceptions
11418 Restricted_Run_Time
11420 Restrictions_Warnings
11423 Source_File_Name_Project
11426 Suppress_Exception_Locations
11427 Task_Dispatching_Policy
11433 Wide_Character_Encoding
11438 * Handling of Configuration Pragmas::
11439 * The Configuration Pragmas Files::
11442 @node Handling of Configuration Pragmas
11443 @section Handling of Configuration Pragmas
11445 Configuration pragmas may either appear at the start of a compilation
11446 unit, in which case they apply only to that unit, or they may apply to
11447 all compilations performed in a given compilation environment.
11449 GNAT also provides the @code{gnatchop} utility to provide an automatic
11450 way to handle configuration pragmas following the semantics for
11451 compilations (that is, files with multiple units), described in the RM.
11452 See @ref{Operating gnatchop in Compilation Mode} for details.
11453 However, for most purposes, it will be more convenient to edit the
11454 @file{gnat.adc} file that contains configuration pragmas directly,
11455 as described in the following section.
11457 @node The Configuration Pragmas Files
11458 @section The Configuration Pragmas Files
11459 @cindex @file{gnat.adc}
11462 In GNAT a compilation environment is defined by the current
11463 directory at the time that a compile command is given. This current
11464 directory is searched for a file whose name is @file{gnat.adc}. If
11465 this file is present, it is expected to contain one or more
11466 configuration pragmas that will be applied to the current compilation.
11467 However, if the switch @option{-gnatA} is used, @file{gnat.adc} is not
11470 Configuration pragmas may be entered into the @file{gnat.adc} file
11471 either by running @code{gnatchop} on a source file that consists only of
11472 configuration pragmas, or more conveniently by
11473 direct editing of the @file{gnat.adc} file, which is a standard format
11476 In addition to @file{gnat.adc}, additional files containing configuration
11477 pragmas may be applied to the current compilation using the switch
11478 @option{-gnatec}@var{path}. @var{path} must designate an existing file that
11479 contains only configuration pragmas. These configuration pragmas are
11480 in addition to those found in @file{gnat.adc} (provided @file{gnat.adc}
11481 is present and switch @option{-gnatA} is not used).
11483 It is allowed to specify several switches @option{-gnatec}, all of which
11484 will be taken into account.
11486 If you are using project file, a separate mechanism is provided using
11487 project attributes, see @ref{Specifying Configuration Pragmas} for more
11491 Of special interest to GNAT OpenVMS Alpha is the following
11492 configuration pragma:
11494 @smallexample @c ada
11496 pragma Extend_System (Aux_DEC);
11501 In the presence of this pragma, GNAT adds to the definition of the
11502 predefined package SYSTEM all the additional types and subprograms that are
11503 defined in HP Ada. See @ref{Compatibility with HP Ada} for details.
11506 @node Handling Arbitrary File Naming Conventions Using gnatname
11507 @chapter Handling Arbitrary File Naming Conventions Using @code{gnatname}
11508 @cindex Arbitrary File Naming Conventions
11511 * Arbitrary File Naming Conventions::
11512 * Running gnatname::
11513 * Switches for gnatname::
11514 * Examples of gnatname Usage::
11517 @node Arbitrary File Naming Conventions
11518 @section Arbitrary File Naming Conventions
11521 The GNAT compiler must be able to know the source file name of a compilation
11522 unit. When using the standard GNAT default file naming conventions
11523 (@code{.ads} for specs, @code{.adb} for bodies), the GNAT compiler
11524 does not need additional information.
11527 When the source file names do not follow the standard GNAT default file naming
11528 conventions, the GNAT compiler must be given additional information through
11529 a configuration pragmas file (@pxref{Configuration Pragmas})
11531 When the non-standard file naming conventions are well-defined,
11532 a small number of pragmas @code{Source_File_Name} specifying a naming pattern
11533 (@pxref{Alternative File Naming Schemes}) may be sufficient. However,
11534 if the file naming conventions are irregular or arbitrary, a number
11535 of pragma @code{Source_File_Name} for individual compilation units
11537 To help maintain the correspondence between compilation unit names and
11538 source file names within the compiler,
11539 GNAT provides a tool @code{gnatname} to generate the required pragmas for a
11542 @node Running gnatname
11543 @section Running @code{gnatname}
11546 The usual form of the @code{gnatname} command is
11549 $ gnatname @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}
11550 @r{[}--and @ovar{switches} @var{naming_pattern} @ovar{naming_patterns}@r{]}
11554 All of the arguments are optional. If invoked without any argument,
11555 @code{gnatname} will display its usage.
11558 When used with at least one naming pattern, @code{gnatname} will attempt to
11559 find all the compilation units in files that follow at least one of the
11560 naming patterns. To find these compilation units,
11561 @code{gnatname} will use the GNAT compiler in syntax-check-only mode on all
11565 One or several Naming Patterns may be given as arguments to @code{gnatname}.
11566 Each Naming Pattern is enclosed between double quotes.
11567 A Naming Pattern is a regular expression similar to the wildcard patterns
11568 used in file names by the Unix shells or the DOS prompt.
11571 @code{gnatname} may be called with several sections of directories/patterns.
11572 Sections are separated by switch @code{--and}. In each section, there must be
11573 at least one pattern. If no directory is specified in a section, the current
11574 directory (or the project directory is @code{-P} is used) is implied.
11575 The options other that the directory switches and the patterns apply globally
11576 even if they are in different sections.
11579 Examples of Naming Patterns are
11588 For a more complete description of the syntax of Naming Patterns,
11589 see the second kind of regular expressions described in @file{g-regexp.ads}
11590 (the ``Glob'' regular expressions).
11593 When invoked with no switch @code{-P}, @code{gnatname} will create a
11594 configuration pragmas file @file{gnat.adc} in the current working directory,
11595 with pragmas @code{Source_File_Name} for each file that contains a valid Ada
11598 @node Switches for gnatname
11599 @section Switches for @code{gnatname}
11602 Switches for @code{gnatname} must precede any specified Naming Pattern.
11605 You may specify any of the following switches to @code{gnatname}:
11611 @cindex @option{--version} @command{gnatname}
11612 Display Copyright and version, then exit disregarding all other options.
11615 @cindex @option{--help} @command{gnatname}
11616 If @option{--version} was not used, display usage, then exit disregarding
11620 Start another section of directories/patterns.
11622 @item ^-c^/CONFIG_FILE=^@file{file}
11623 @cindex @option{^-c^/CONFIG_FILE^} (@code{gnatname})
11624 Create a configuration pragmas file @file{file} (instead of the default
11627 There may be zero, one or more space between @option{-c} and
11630 @file{file} may include directory information. @file{file} must be
11631 writable. There may be only one switch @option{^-c^/CONFIG_FILE^}.
11632 When a switch @option{^-c^/CONFIG_FILE^} is
11633 specified, no switch @option{^-P^/PROJECT_FILE^} may be specified (see below).
11635 @item ^-d^/SOURCE_DIRS=^@file{dir}
11636 @cindex @option{^-d^/SOURCE_DIRS^} (@code{gnatname})
11637 Look for source files in directory @file{dir}. There may be zero, one or more
11638 spaces between @option{^-d^/SOURCE_DIRS=^} and @file{dir}.
11639 When a switch @option{^-d^/SOURCE_DIRS^}
11640 is specified, the current working directory will not be searched for source
11641 files, unless it is explicitly specified with a @option{^-d^/SOURCE_DIRS^}
11642 or @option{^-D^/DIR_FILES^} switch.
11643 Several switches @option{^-d^/SOURCE_DIRS^} may be specified.
11644 If @file{dir} is a relative path, it is relative to the directory of
11645 the configuration pragmas file specified with switch
11646 @option{^-c^/CONFIG_FILE^},
11647 or to the directory of the project file specified with switch
11648 @option{^-P^/PROJECT_FILE^} or,
11649 if neither switch @option{^-c^/CONFIG_FILE^}
11650 nor switch @option{^-P^/PROJECT_FILE^} are specified, it is relative to the
11651 current working directory. The directory
11652 specified with switch @option{^-d^/SOURCE_DIRS^} must exist and be readable.
11654 @item ^-D^/DIRS_FILE=^@file{file}
11655 @cindex @option{^-D^/DIRS_FILE^} (@code{gnatname})
11656 Look for source files in all directories listed in text file @file{file}.
11657 There may be zero, one or more spaces between @option{^-D^/DIRS_FILE=^}
11659 @file{file} must be an existing, readable text file.
11660 Each nonempty line in @file{file} must be a directory.
11661 Specifying switch @option{^-D^/DIRS_FILE^} is equivalent to specifying as many
11662 switches @option{^-d^/SOURCE_DIRS^} as there are nonempty lines in
11665 @item ^-f^/FOREIGN_PATTERN=^@file{pattern}
11666 @cindex @option{^-f^/FOREIGN_PATTERN^} (@code{gnatname})
11667 Foreign patterns. Using this switch, it is possible to add sources of languages
11668 other than Ada to the list of sources of a project file.
11669 It is only useful if a ^-P^/PROJECT_FILE^ switch is used.
11672 gnatname ^-Pprj -f"*.c"^/PROJECT_FILE=PRJ /FOREIGN_PATTERN=*.C^ "*.ada"
11675 will look for Ada units in all files with the @file{.ada} extension,
11676 and will add to the list of file for project @file{prj.gpr} the C files
11677 with extension @file{.^c^C^}.
11680 @cindex @option{^-h^/HELP^} (@code{gnatname})
11681 Output usage (help) information. The output is written to @file{stdout}.
11683 @item ^-P^/PROJECT_FILE=^@file{proj}
11684 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatname})
11685 Create or update project file @file{proj}. There may be zero, one or more space
11686 between @option{-P} and @file{proj}. @file{proj} may include directory
11687 information. @file{proj} must be writable.
11688 There may be only one switch @option{^-P^/PROJECT_FILE^}.
11689 When a switch @option{^-P^/PROJECT_FILE^} is specified,
11690 no switch @option{^-c^/CONFIG_FILE^} may be specified.
11692 @item ^-v^/VERBOSE^
11693 @cindex @option{^-v^/VERBOSE^} (@code{gnatname})
11694 Verbose mode. Output detailed explanation of behavior to @file{stdout}.
11695 This includes name of the file written, the name of the directories to search
11696 and, for each file in those directories whose name matches at least one of
11697 the Naming Patterns, an indication of whether the file contains a unit,
11698 and if so the name of the unit.
11700 @item ^-v -v^/VERBOSE /VERBOSE^
11701 @cindex @option{^-v -v^/VERBOSE /VERBOSE^} (@code{gnatname})
11702 Very Verbose mode. In addition to the output produced in verbose mode,
11703 for each file in the searched directories whose name matches none of
11704 the Naming Patterns, an indication is given that there is no match.
11706 @item ^-x^/EXCLUDED_PATTERN=^@file{pattern}
11707 @cindex @option{^-x^/EXCLUDED_PATTERN^} (@code{gnatname})
11708 Excluded patterns. Using this switch, it is possible to exclude some files
11709 that would match the name patterns. For example,
11711 gnatname ^-x "*_nt.ada"^/EXCLUDED_PATTERN=*_nt.ada^ "*.ada"
11714 will look for Ada units in all files with the @file{.ada} extension,
11715 except those whose names end with @file{_nt.ada}.
11719 @node Examples of gnatname Usage
11720 @section Examples of @code{gnatname} Usage
11724 $ gnatname /CONFIG_FILE=[HOME.ME]NAMES.ADC /SOURCE_DIRS=SOURCES "[a-z]*.ada*"
11730 $ gnatname -c /home/me/names.adc -d sources "[a-z]*.ada*"
11735 In this example, the directory @file{^/home/me^[HOME.ME]^} must already exist
11736 and be writable. In addition, the directory
11737 @file{^/home/me/sources^[HOME.ME.SOURCES]^} (specified by
11738 @option{^-d sources^/SOURCE_DIRS=SOURCES^}) must exist and be readable.
11741 Note the optional spaces after @option{-c} and @option{-d}.
11746 $ gnatname -P/home/me/proj -x "*_nt_body.ada"
11747 -dsources -dsources/plus -Dcommon_dirs.txt "body_*" "spec_*"
11750 $ gnatname /PROJECT_FILE=[HOME.ME]PROJ
11751 /EXCLUDED_PATTERN=*_nt_body.ada
11752 /SOURCE_DIRS=(SOURCES,[SOURCES.PLUS])
11753 /DIRS_FILE=COMMON_DIRS.TXT "body_*" "spec_*"
11757 Note that several switches @option{^-d^/SOURCE_DIRS^} may be used,
11758 even in conjunction with one or several switches
11759 @option{^-D^/DIRS_FILE^}. Several Naming Patterns and one excluded pattern
11760 are used in this example.
11762 @c *****************************************
11763 @c * G N A T P r o j e c t M a n a g e r *
11764 @c *****************************************
11765 @node GNAT Project Manager
11766 @chapter GNAT Project Manager
11770 * Examples of Project Files::
11771 * Project File Syntax::
11772 * Objects and Sources in Project Files::
11773 * Importing Projects::
11774 * Project Extension::
11775 * Project Hierarchy Extension::
11776 * External References in Project Files::
11777 * Packages in Project Files::
11778 * Variables from Imported Projects::
11780 * Library Projects::
11781 * Stand-alone Library Projects::
11782 * Switches Related to Project Files::
11783 * Tools Supporting Project Files::
11784 * An Extended Example::
11785 * Project File Complete Syntax::
11788 @c ****************
11789 @c * Introduction *
11790 @c ****************
11793 @section Introduction
11796 This chapter describes GNAT's @emph{Project Manager}, a facility that allows
11797 you to manage complex builds involving a number of source files, directories,
11798 and compilation options for different system configurations. In particular,
11799 project files allow you to specify:
11802 The directory or set of directories containing the source files, and/or the
11803 names of the specific source files themselves
11805 The directory in which the compiler's output
11806 (@file{ALI} files, object files, tree files) is to be placed
11808 The directory in which the executable programs is to be placed
11810 ^Switch^Switch^ settings for any of the project-enabled tools
11811 (@command{gnatmake}, compiler, binder, linker, @code{gnatls}, @code{gnatxref},
11812 @code{gnatfind}); you can apply these settings either globally or to individual
11815 The source files containing the main subprogram(s) to be built
11817 The source programming language(s) (currently Ada and/or C)
11819 Source file naming conventions; you can specify these either globally or for
11820 individual compilation units
11827 @node Project Files
11828 @subsection Project Files
11831 Project files are written in a syntax close to that of Ada, using familiar
11832 notions such as packages, context clauses, declarations, default values,
11833 assignments, and inheritance. Finally, project files can be built
11834 hierarchically from other project files, simplifying complex system
11835 integration and project reuse.
11837 A @dfn{project} is a specific set of values for various compilation properties.
11838 The settings for a given project are described by means of
11839 a @dfn{project file}, which is a text file written in an Ada-like syntax.
11840 Property values in project files are either strings or lists of strings.
11841 Properties that are not explicitly set receive default values. A project
11842 file may interrogate the values of @dfn{external variables} (user-defined
11843 command-line switches or environment variables), and it may specify property
11844 settings conditionally, based on the value of such variables.
11846 In simple cases, a project's source files depend only on other source files
11847 in the same project, or on the predefined libraries. (@emph{Dependence} is
11849 the Ada technical sense; as in one Ada unit @code{with}ing another.) However,
11850 the Project Manager also allows more sophisticated arrangements,
11851 where the source files in one project depend on source files in other
11855 One project can @emph{import} other projects containing needed source files.
11857 You can organize GNAT projects in a hierarchy: a @emph{child} project
11858 can extend a @emph{parent} project, inheriting the parent's source files and
11859 optionally overriding any of them with alternative versions
11863 More generally, the Project Manager lets you structure large development
11864 efforts into hierarchical subsystems, where build decisions are delegated
11865 to the subsystem level, and thus different compilation environments
11866 (^switch^switch^ settings) used for different subsystems.
11868 The Project Manager is invoked through the
11869 @option{^-P^/PROJECT_FILE=^@emph{projectfile}}
11870 switch to @command{gnatmake} or to the @command{^gnat^GNAT^} front driver.
11872 There may be zero, one or more spaces between @option{-P} and
11873 @option{@emph{projectfile}}.
11875 If you want to define (on the command line) an external variable that is
11876 queried by the project file, you must use the
11877 @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
11878 The Project Manager parses and interprets the project file, and drives the
11879 invoked tool based on the project settings.
11881 The Project Manager supports a wide range of development strategies,
11882 for systems of all sizes. Here are some typical practices that are
11886 Using a common set of source files, but generating object files in different
11887 directories via different ^switch^switch^ settings
11889 Using a mostly-shared set of source files, but with different versions of
11894 The destination of an executable can be controlled inside a project file
11895 using the @option{^-o^-o^}
11897 In the absence of such a ^switch^switch^ either inside
11898 the project file or on the command line, any executable files generated by
11899 @command{gnatmake} are placed in the directory @code{Exec_Dir} specified
11900 in the project file. If no @code{Exec_Dir} is specified, they will be placed
11901 in the object directory of the project.
11903 You can use project files to achieve some of the effects of a source
11904 versioning system (for example, defining separate projects for
11905 the different sets of sources that comprise different releases) but the
11906 Project Manager is independent of any source configuration management tools
11907 that might be used by the developers.
11909 The next section introduces the main features of GNAT's project facility
11910 through a sequence of examples; subsequent sections will present the syntax
11911 and semantics in more detail. A more formal description of the project
11912 facility appears in @ref{Project File Reference,,, gnat_rm, GNAT
11915 @c *****************************
11916 @c * Examples of Project Files *
11917 @c *****************************
11919 @node Examples of Project Files
11920 @section Examples of Project Files
11922 This section illustrates some of the typical uses of project files and
11923 explains their basic structure and behavior.
11926 * Common Sources with Different ^Switches^Switches^ and Directories::
11927 * Using External Variables::
11928 * Importing Other Projects::
11929 * Extending a Project::
11932 @node Common Sources with Different ^Switches^Switches^ and Directories
11933 @subsection Common Sources with Different ^Switches^Switches^ and Directories
11937 * Specifying the Object Directory::
11938 * Specifying the Exec Directory::
11939 * Project File Packages::
11940 * Specifying ^Switch^Switch^ Settings::
11941 * Main Subprograms::
11942 * Executable File Names::
11943 * Source File Naming Conventions::
11944 * Source Language(s)::
11948 Suppose that the Ada source files @file{pack.ads}, @file{pack.adb}, and
11949 @file{proc.adb} are in the @file{/common} directory. The file
11950 @file{proc.adb} contains an Ada main subprogram @code{Proc} that @code{with}s
11951 package @code{Pack}. We want to compile these source files under two sets
11952 of ^switches^switches^:
11955 When debugging, we want to pass the @option{-g} switch to @command{gnatmake},
11956 and the @option{^-gnata^-gnata^},
11957 @option{^-gnato^-gnato^},
11958 and @option{^-gnatE^-gnatE^} switches to the
11959 compiler; the compiler's output is to appear in @file{/common/debug}
11961 When preparing a release version, we want to pass the @option{^-O2^O2^} switch
11962 to the compiler; the compiler's output is to appear in @file{/common/release}
11966 The GNAT project files shown below, respectively @file{debug.gpr} and
11967 @file{release.gpr} in the @file{/common} directory, achieve these effects.
11980 ^/common/debug^[COMMON.DEBUG]^
11985 ^/common/release^[COMMON.RELEASE]^
11990 Here are the corresponding project files:
11992 @smallexample @c projectfile
11995 for Object_Dir use "debug";
11996 for Main use ("proc");
11999 for ^Default_Switches^Default_Switches^ ("Ada")
12001 for Executable ("proc.adb") use "proc1";
12006 package Compiler is
12007 for ^Default_Switches^Default_Switches^ ("Ada")
12008 use ("-fstack-check",
12011 "^-gnatE^-gnatE^");
12017 @smallexample @c projectfile
12020 for Object_Dir use "release";
12021 for Exec_Dir use ".";
12022 for Main use ("proc");
12024 package Compiler is
12025 for ^Default_Switches^Default_Switches^ ("Ada")
12033 The name of the project defined by @file{debug.gpr} is @code{"Debug"} (case
12034 insensitive), and analogously the project defined by @file{release.gpr} is
12035 @code{"Release"}. For consistency the file should have the same name as the
12036 project, and the project file's extension should be @code{"gpr"}. These
12037 conventions are not required, but a warning is issued if they are not followed.
12039 If the current directory is @file{^/temp^[TEMP]^}, then the command
12041 gnatmake ^-P/common/debug.gpr^/PROJECT_FILE=[COMMON]DEBUG^
12045 generates object and ALI files in @file{^/common/debug^[COMMON.DEBUG]^},
12046 as well as the @code{^proc1^PROC1.EXE^} executable,
12047 using the ^switch^switch^ settings defined in the project file.
12049 Likewise, the command
12051 gnatmake ^-P/common/release.gpr^/PROJECT_FILE=[COMMON]RELEASE^
12055 generates object and ALI files in @file{^/common/release^[COMMON.RELEASE]^},
12056 and the @code{^proc^PROC.EXE^}
12057 executable in @file{^/common^[COMMON]^},
12058 using the ^switch^switch^ settings from the project file.
12061 @unnumberedsubsubsec Source Files
12064 If a project file does not explicitly specify a set of source directories or
12065 a set of source files, then by default the project's source files are the
12066 Ada source files in the project file directory. Thus @file{pack.ads},
12067 @file{pack.adb}, and @file{proc.adb} are the source files for both projects.
12069 @node Specifying the Object Directory
12070 @unnumberedsubsubsec Specifying the Object Directory
12073 Several project properties are modeled by Ada-style @emph{attributes};
12074 a property is defined by supplying the equivalent of an Ada attribute
12075 definition clause in the project file.
12076 A project's object directory is another such a property; the corresponding
12077 attribute is @code{Object_Dir}, and its value is also a string expression,
12078 specified either as absolute or relative. In the later case,
12079 it is relative to the project file directory. Thus the compiler's
12080 output is directed to @file{^/common/debug^[COMMON.DEBUG]^}
12081 (for the @code{Debug} project)
12082 and to @file{^/common/release^[COMMON.RELEASE]^}
12083 (for the @code{Release} project).
12084 If @code{Object_Dir} is not specified, then the default is the project file
12087 @node Specifying the Exec Directory
12088 @unnumberedsubsubsec Specifying the Exec Directory
12091 A project's exec directory is another property; the corresponding
12092 attribute is @code{Exec_Dir}, and its value is also a string expression,
12093 either specified as relative or absolute. If @code{Exec_Dir} is not specified,
12094 then the default is the object directory (which may also be the project file
12095 directory if attribute @code{Object_Dir} is not specified). Thus the executable
12096 is placed in @file{^/common/debug^[COMMON.DEBUG]^}
12097 for the @code{Debug} project (attribute @code{Exec_Dir} not specified)
12098 and in @file{^/common^[COMMON]^} for the @code{Release} project.
12100 @node Project File Packages
12101 @unnumberedsubsubsec Project File Packages
12104 A GNAT tool that is integrated with the Project Manager is modeled by a
12105 corresponding package in the project file. In the example above,
12106 The @code{Debug} project defines the packages @code{Builder}
12107 (for @command{gnatmake}) and @code{Compiler};
12108 the @code{Release} project defines only the @code{Compiler} package.
12110 The Ada-like package syntax is not to be taken literally. Although packages in
12111 project files bear a surface resemblance to packages in Ada source code, the
12112 notation is simply a way to convey a grouping of properties for a named
12113 entity. Indeed, the package names permitted in project files are restricted
12114 to a predefined set, corresponding to the project-aware tools, and the contents
12115 of packages are limited to a small set of constructs.
12116 The packages in the example above contain attribute definitions.
12118 @node Specifying ^Switch^Switch^ Settings
12119 @unnumberedsubsubsec Specifying ^Switch^Switch^ Settings
12122 ^Switch^Switch^ settings for a project-aware tool can be specified through
12123 attributes in the package that corresponds to the tool.
12124 The example above illustrates one of the relevant attributes,
12125 @code{^Default_Switches^Default_Switches^}, which is defined in packages
12126 in both project files.
12127 Unlike simple attributes like @code{Source_Dirs},
12128 @code{^Default_Switches^Default_Switches^} is
12129 known as an @emph{associative array}. When you define this attribute, you must
12130 supply an ``index'' (a literal string), and the effect of the attribute
12131 definition is to set the value of the array at the specified index.
12132 For the @code{^Default_Switches^Default_Switches^} attribute,
12133 the index is a programming language (in our case, Ada),
12134 and the value specified (after @code{use}) must be a list
12135 of string expressions.
12137 The attributes permitted in project files are restricted to a predefined set.
12138 Some may appear at project level, others in packages.
12139 For any attribute that is an associative array, the index must always be a
12140 literal string, but the restrictions on this string (e.g., a file name or a
12141 language name) depend on the individual attribute.
12142 Also depending on the attribute, its specified value will need to be either a
12143 string or a string list.
12145 In the @code{Debug} project, we set the switches for two tools,
12146 @command{gnatmake} and the compiler, and thus we include the two corresponding
12147 packages; each package defines the @code{^Default_Switches^Default_Switches^}
12148 attribute with index @code{"Ada"}.
12149 Note that the package corresponding to
12150 @command{gnatmake} is named @code{Builder}. The @code{Release} project is
12151 similar, but only includes the @code{Compiler} package.
12153 In project @code{Debug} above, the ^switches^switches^ starting with
12154 @option{-gnat} that are specified in package @code{Compiler}
12155 could have been placed in package @code{Builder}, since @command{gnatmake}
12156 transmits all such ^switches^switches^ to the compiler.
12158 @node Main Subprograms
12159 @unnumberedsubsubsec Main Subprograms
12162 One of the specifiable properties of a project is a list of files that contain
12163 main subprograms. This property is captured in the @code{Main} attribute,
12164 whose value is a list of strings. If a project defines the @code{Main}
12165 attribute, it is not necessary to identify the main subprogram(s) when
12166 invoking @command{gnatmake} (@pxref{gnatmake and Project Files}).
12168 @node Executable File Names
12169 @unnumberedsubsubsec Executable File Names
12172 By default, the executable file name corresponding to a main source is
12173 deduced from the main source file name. Through the attributes
12174 @code{Executable} and @code{Executable_Suffix} of package @code{Builder},
12175 it is possible to change this default.
12176 In project @code{Debug} above, the executable file name
12177 for main source @file{^proc.adb^PROC.ADB^} is
12178 @file{^proc1^PROC1.EXE^}.
12179 Attribute @code{Executable_Suffix}, when specified, may change the suffix
12180 of the executable files, when no attribute @code{Executable} applies:
12181 its value replace the platform-specific executable suffix.
12182 Attributes @code{Executable} and @code{Executable_Suffix} are the only ways to
12183 specify a non-default executable file name when several mains are built at once
12184 in a single @command{gnatmake} command.
12186 @node Source File Naming Conventions
12187 @unnumberedsubsubsec Source File Naming Conventions
12190 Since the project files above do not specify any source file naming
12191 conventions, the GNAT defaults are used. The mechanism for defining source
12192 file naming conventions -- a package named @code{Naming} --
12193 is described below (@pxref{Naming Schemes}).
12195 @node Source Language(s)
12196 @unnumberedsubsubsec Source Language(s)
12199 Since the project files do not specify a @code{Languages} attribute, by
12200 default the GNAT tools assume that the language of the project file is Ada.
12201 More generally, a project can comprise source files
12202 in Ada, C, and/or other languages.
12204 @node Using External Variables
12205 @subsection Using External Variables
12208 Instead of supplying different project files for debug and release, we can
12209 define a single project file that queries an external variable (set either
12210 on the command line or via an ^environment variable^logical name^) in order to
12211 conditionally define the appropriate settings. Again, assume that the
12212 source files @file{pack.ads}, @file{pack.adb}, and @file{proc.adb} are
12213 located in directory @file{^/common^[COMMON]^}. The following project file,
12214 @file{build.gpr}, queries the external variable named @code{STYLE} and
12215 defines an object directory and ^switch^switch^ settings based on whether
12216 the value is @code{"deb"} (debug) or @code{"rel"} (release), and where
12217 the default is @code{"deb"}.
12219 @smallexample @c projectfile
12222 for Main use ("proc");
12224 type Style_Type is ("deb", "rel");
12225 Style : Style_Type := external ("STYLE", "deb");
12229 for Object_Dir use "debug";
12232 for Object_Dir use "release";
12233 for Exec_Dir use ".";
12242 for ^Default_Switches^Default_Switches^ ("Ada")
12244 for Executable ("proc") use "proc1";
12253 package Compiler is
12257 for ^Default_Switches^Default_Switches^ ("Ada")
12258 use ("^-gnata^-gnata^",
12260 "^-gnatE^-gnatE^");
12263 for ^Default_Switches^Default_Switches^ ("Ada")
12274 @code{Style_Type} is an example of a @emph{string type}, which is the project
12275 file analog of an Ada enumeration type but whose components are string literals
12276 rather than identifiers. @code{Style} is declared as a variable of this type.
12278 The form @code{external("STYLE", "deb")} is known as an
12279 @emph{external reference}; its first argument is the name of an
12280 @emph{external variable}, and the second argument is a default value to be
12281 used if the external variable doesn't exist. You can define an external
12282 variable on the command line via the @option{^-X^/EXTERNAL_REFERENCE^} switch,
12283 or you can use ^an environment variable^a logical name^
12284 as an external variable.
12286 Each @code{case} construct is expanded by the Project Manager based on the
12287 value of @code{Style}. Thus the command
12290 gnatmake -P/common/build.gpr -XSTYLE=deb
12296 gnatmake /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=deb
12301 is equivalent to the @command{gnatmake} invocation using the project file
12302 @file{debug.gpr} in the earlier example. So is the command
12304 gnatmake ^-P/common/build.gpr^/PROJECT_FILE=[COMMON]BUILD.GPR^
12308 since @code{"deb"} is the default for @code{STYLE}.
12314 gnatmake -P/common/build.gpr -XSTYLE=rel
12320 GNAT MAKE /PROJECT_FILE=[COMMON]BUILD.GPR /EXTERNAL_REFERENCE=STYLE=rel
12325 is equivalent to the @command{gnatmake} invocation using the project file
12326 @file{release.gpr} in the earlier example.
12328 @node Importing Other Projects
12329 @subsection Importing Other Projects
12330 @cindex @code{ADA_PROJECT_PATH}
12331 @cindex @code{GPR_PROJECT_PATH}
12334 A compilation unit in a source file in one project may depend on compilation
12335 units in source files in other projects. To compile this unit under
12336 control of a project file, the
12337 dependent project must @emph{import} the projects containing the needed source
12339 This effect is obtained using syntax similar to an Ada @code{with} clause,
12340 but where @code{with}ed entities are strings that denote project files.
12342 As an example, suppose that the two projects @code{GUI_Proj} and
12343 @code{Comm_Proj} are defined in the project files @file{gui_proj.gpr} and
12344 @file{comm_proj.gpr} in directories @file{^/gui^[GUI]^}
12345 and @file{^/comm^[COMM]^}, respectively.
12346 Suppose that the source files for @code{GUI_Proj} are
12347 @file{gui.ads} and @file{gui.adb}, and that the source files for
12348 @code{Comm_Proj} are @file{comm.ads} and @file{comm.adb}, where each set of
12349 files is located in its respective project file directory. Schematically:
12368 We want to develop an application in directory @file{^/app^[APP]^} that
12369 @code{with} the packages @code{GUI} and @code{Comm}, using the properties of
12370 the corresponding project files (e.g.@: the ^switch^switch^ settings
12371 and object directory).
12372 Skeletal code for a main procedure might be something like the following:
12374 @smallexample @c ada
12377 procedure App_Main is
12386 Here is a project file, @file{app_proj.gpr}, that achieves the desired
12389 @smallexample @c projectfile
12391 with "/gui/gui_proj", "/comm/comm_proj";
12392 project App_Proj is
12393 for Main use ("app_main");
12399 Building an executable is achieved through the command:
12401 gnatmake ^-P/app/app_proj^/PROJECT_FILE=[APP]APP_PROJ^
12404 which will generate the @code{^app_main^APP_MAIN.EXE^} executable
12405 in the directory where @file{app_proj.gpr} resides.
12407 If an imported project file uses the standard extension (@code{^gpr^GPR^}) then
12408 (as illustrated above) the @code{with} clause can omit the extension.
12410 Our example specified an absolute path for each imported project file.
12411 Alternatively, the directory name of an imported object can be omitted
12415 The imported project file is in the same directory as the importing project
12418 You have defined one or two ^environment variables^logical names^
12419 that includes the directory containing
12420 the needed project file. The syntax of @code{GPR_PROJECT_PATH} and
12421 @code{ADA_PROJECT_PATH} is the same as
12422 the syntax of @code{ADA_INCLUDE_PATH} and @code{ADA_OBJECTS_PATH}: a list of
12423 directory names separated by colons (semicolons on Windows).
12427 Thus, if we define @code{ADA_PROJECT_PATH} or @code{GPR_PROJECT_PATH}
12428 to include @file{^/gui^[GUI]^} and
12429 @file{^/comm^[COMM]^}, then our project file @file{app_proj.gpr} can be written
12432 @smallexample @c projectfile
12434 with "gui_proj", "comm_proj";
12435 project App_Proj is
12436 for Main use ("app_main");
12442 Importing other projects can create ambiguities.
12443 For example, the same unit might be present in different imported projects, or
12444 it might be present in both the importing project and in an imported project.
12445 Both of these conditions are errors. Note that in the current version of
12446 the Project Manager, it is illegal to have an ambiguous unit even if the
12447 unit is never referenced by the importing project. This restriction may be
12448 relaxed in a future release.
12450 @node Extending a Project
12451 @subsection Extending a Project
12454 In large software systems it is common to have multiple
12455 implementations of a common interface; in Ada terms, multiple versions of a
12456 package body for the same spec. For example, one implementation
12457 might be safe for use in tasking programs, while another might only be used
12458 in sequential applications. This can be modeled in GNAT using the concept
12459 of @emph{project extension}. If one project (the ``child'') @emph{extends}
12460 another project (the ``parent'') then by default all source files of the
12461 parent project are inherited by the child, but the child project can
12462 override any of the parent's source files with new versions, and can also
12463 add new files. This facility is the project analog of a type extension in
12464 Object-Oriented Programming. Project hierarchies are permitted (a child
12465 project may be the parent of yet another project), and a project that
12466 inherits one project can also import other projects.
12468 As an example, suppose that directory @file{^/seq^[SEQ]^} contains the project
12469 file @file{seq_proj.gpr} as well as the source files @file{pack.ads},
12470 @file{pack.adb}, and @file{proc.adb}:
12483 Note that the project file can simply be empty (that is, no attribute or
12484 package is defined):
12486 @smallexample @c projectfile
12488 project Seq_Proj is
12494 implying that its source files are all the Ada source files in the project
12497 Suppose we want to supply an alternate version of @file{pack.adb}, in
12498 directory @file{^/tasking^[TASKING]^}, but use the existing versions of
12499 @file{pack.ads} and @file{proc.adb}. We can define a project
12500 @code{Tasking_Proj} that inherits @code{Seq_Proj}:
12504 ^/tasking^[TASKING]^
12510 project Tasking_Proj extends "/seq/seq_proj" is
12516 The version of @file{pack.adb} used in a build depends on which project file
12519 Note that we could have obtained the desired behavior using project import
12520 rather than project inheritance; a @code{base} project would contain the
12521 sources for @file{pack.ads} and @file{proc.adb}, a sequential project would
12522 import @code{base} and add @file{pack.adb}, and likewise a tasking project
12523 would import @code{base} and add a different version of @file{pack.adb}. The
12524 choice depends on whether other sources in the original project need to be
12525 overridden. If they do, then project extension is necessary, otherwise,
12526 importing is sufficient.
12529 In a project file that extends another project file, it is possible to
12530 indicate that an inherited source is not part of the sources of the extending
12531 project. This is necessary sometimes when a package spec has been overloaded
12532 and no longer requires a body: in this case, it is necessary to indicate that
12533 the inherited body is not part of the sources of the project, otherwise there
12534 will be a compilation error when compiling the spec.
12536 For that purpose, the attribute @code{Excluded_Source_Files} is used.
12537 Its value is a string list: a list of file names. It is also possible to use
12538 attribute @code{Excluded_Source_List_File}. Its value is a single string:
12539 the file name of a text file containing a list of file names, one per line.
12541 @smallexample @c @projectfile
12542 project B extends "a" is
12543 for Source_Files use ("pkg.ads");
12544 -- New spec of Pkg does not need a completion
12545 for Excluded_Source_Files use ("pkg.adb");
12549 Attribute @code{Excluded_Source_Files} may also be used to check if a source
12550 is still needed: if it is possible to build using @command{gnatmake} when such
12551 a source is put in attribute @code{Excluded_Source_Files} of a project P, then
12552 it is possible to remove the source completely from a system that includes
12555 @c ***********************
12556 @c * Project File Syntax *
12557 @c ***********************
12559 @node Project File Syntax
12560 @section Project File Syntax
12564 * Qualified Projects::
12570 * Associative Array Attributes::
12571 * case Constructions::
12575 This section describes the structure of project files.
12577 A project may be an @emph{independent project}, entirely defined by a single
12578 project file. Any Ada source file in an independent project depends only
12579 on the predefined library and other Ada source files in the same project.
12582 A project may also @dfn{depend on} other projects, in either or both of
12583 the following ways:
12585 @item It may import any number of projects
12586 @item It may extend at most one other project
12590 The dependence relation is a directed acyclic graph (the subgraph reflecting
12591 the ``extends'' relation is a tree).
12593 A project's @dfn{immediate sources} are the source files directly defined by
12594 that project, either implicitly by residing in the project file's directory,
12595 or explicitly through any of the source-related attributes described below.
12596 More generally, a project @var{proj}'s @dfn{sources} are the immediate sources
12597 of @var{proj} together with the immediate sources (unless overridden) of any
12598 project on which @var{proj} depends (either directly or indirectly).
12601 @subsection Basic Syntax
12604 As seen in the earlier examples, project files have an Ada-like syntax.
12605 The minimal project file is:
12606 @smallexample @c projectfile
12615 The identifier @code{Empty} is the name of the project.
12616 This project name must be present after the reserved
12617 word @code{end} at the end of the project file, followed by a semi-colon.
12619 Any name in a project file, such as the project name or a variable name,
12620 has the same syntax as an Ada identifier.
12622 The reserved words of project files are the Ada 95 reserved words plus
12623 @code{extends}, @code{external}, and @code{project}. Note that the only Ada
12624 reserved words currently used in project file syntax are:
12660 Comments in project files have the same syntax as in Ada, two consecutive
12661 hyphens through the end of the line.
12663 @node Qualified Projects
12664 @subsection Qualified Projects
12667 Before the reserved @code{project}, there may be one or two "qualifiers", that
12668 is identifiers or other reserved words, to qualify the project.
12670 The current list of qualifiers is:
12674 @code{abstract}: qualify a project with no sources. A qualified abstract
12675 project must either have no declaration of attributes @code{Source_Dirs},
12676 @code{Source_Files}, @code{Languages} or @code{Source_List_File}, or one of
12677 @code{Source_Dirs}, @code{Source_Files}, or @code{Languages} must be declared
12678 as empty. If it extends another project, the project it extends must also be a
12679 qualified abstract project.
12682 @code{standard}: a standard project is a non library project with sources.
12685 @code{aggregate}: for future extension
12688 @code{aggregate library}: for future extension
12691 @code{library}: a library project must declare both attributes
12692 @code{Library_Name} and @code{Library_Dir}.
12695 @code{configuration}: a configuration project cannot be in a project tree.
12699 @subsection Packages
12702 A project file may contain @emph{packages}. The name of a package must be one
12703 of the identifiers from the following list. A package
12704 with a given name may only appear once in a project file. Package names are
12705 case insensitive. The following package names are legal:
12721 @code{Cross_Reference}
12727 @code{Pretty_Printer}
12737 @code{Language_Processing}
12741 In its simplest form, a package may be empty:
12743 @smallexample @c projectfile
12753 A package may contain @emph{attribute declarations},
12754 @emph{variable declarations} and @emph{case constructions}, as will be
12757 When there is ambiguity between a project name and a package name,
12758 the name always designates the project. To avoid possible confusion, it is
12759 always a good idea to avoid naming a project with one of the
12760 names allowed for packages or any name that starts with @code{gnat}.
12763 @subsection Expressions
12766 An @emph{expression} is either a @emph{string expression} or a
12767 @emph{string list expression}.
12769 A @emph{string expression} is either a @emph{simple string expression} or a
12770 @emph{compound string expression}.
12772 A @emph{simple string expression} is one of the following:
12774 @item A literal string; e.g.@: @code{"comm/my_proj.gpr"}
12775 @item A string-valued variable reference (@pxref{Variables})
12776 @item A string-valued attribute reference (@pxref{Attributes})
12777 @item An external reference (@pxref{External References in Project Files})
12781 A @emph{compound string expression} is a concatenation of string expressions,
12782 using the operator @code{"&"}
12784 Path & "/" & File_Name & ".ads"
12788 A @emph{string list expression} is either a
12789 @emph{simple string list expression} or a
12790 @emph{compound string list expression}.
12792 A @emph{simple string list expression} is one of the following:
12794 @item A parenthesized list of zero or more string expressions,
12795 separated by commas
12797 File_Names := (File_Name, "gnat.adc", File_Name & ".orig");
12800 @item A string list-valued variable reference
12801 @item A string list-valued attribute reference
12805 A @emph{compound string list expression} is the concatenation (using
12806 @code{"&"}) of a simple string list expression and an expression. Note that
12807 each term in a compound string list expression, except the first, may be
12808 either a string expression or a string list expression.
12810 @smallexample @c projectfile
12812 File_Name_List := () & File_Name; -- One string in this list
12813 Extended_File_Name_List := File_Name_List & (File_Name & ".orig");
12815 Big_List := File_Name_List & Extended_File_Name_List;
12816 -- Concatenation of two string lists: three strings
12817 Illegal_List := "gnat.adc" & Extended_File_Name_List;
12818 -- Illegal: must start with a string list
12823 @subsection String Types
12826 A @emph{string type declaration} introduces a discrete set of string literals.
12827 If a string variable is declared to have this type, its value
12828 is restricted to the given set of literals.
12830 Here is an example of a string type declaration:
12832 @smallexample @c projectfile
12833 type OS is ("NT", "nt", "Unix", "GNU/Linux", "other OS");
12837 Variables of a string type are called @emph{typed variables}; all other
12838 variables are called @emph{untyped variables}. Typed variables are
12839 particularly useful in @code{case} constructions, to support conditional
12840 attribute declarations.
12841 (@pxref{case Constructions}).
12843 The string literals in the list are case sensitive and must all be different.
12844 They may include any graphic characters allowed in Ada, including spaces.
12846 A string type may only be declared at the project level, not inside a package.
12848 A string type may be referenced by its name if it has been declared in the same
12849 project file, or by an expanded name whose prefix is the name of the project
12850 in which it is declared.
12853 @subsection Variables
12856 A variable may be declared at the project file level, or within a package.
12857 Here are some examples of variable declarations:
12859 @smallexample @c projectfile
12861 This_OS : OS := external ("OS"); -- a typed variable declaration
12862 That_OS := "GNU/Linux"; -- an untyped variable declaration
12867 The syntax of a @emph{typed variable declaration} is identical to the Ada
12868 syntax for an object declaration. By contrast, the syntax of an untyped
12869 variable declaration is identical to an Ada assignment statement. In fact,
12870 variable declarations in project files have some of the characteristics of
12871 an assignment, in that successive declarations for the same variable are
12872 allowed. Untyped variable declarations do establish the expected kind of the
12873 variable (string or string list), and successive declarations for it must
12874 respect the initial kind.
12877 A string variable declaration (typed or untyped) declares a variable
12878 whose value is a string. This variable may be used as a string expression.
12879 @smallexample @c projectfile
12880 File_Name := "readme.txt";
12881 Saved_File_Name := File_Name & ".saved";
12885 A string list variable declaration declares a variable whose value is a list
12886 of strings. The list may contain any number (zero or more) of strings.
12888 @smallexample @c projectfile
12890 List_With_One_Element := ("^-gnaty^-gnaty^");
12891 List_With_Two_Elements := List_With_One_Element & "^-gnatg^-gnatg^";
12892 Long_List := ("main.ada", "pack1_.ada", "pack1.ada", "pack2_.ada"
12893 "pack2.ada", "util_.ada", "util.ada");
12897 The same typed variable may not be declared more than once at project level,
12898 and it may not be declared more than once in any package; it is in effect
12901 The same untyped variable may be declared several times. Declarations are
12902 elaborated in the order in which they appear, so the new value replaces
12903 the old one, and any subsequent reference to the variable uses the new value.
12904 However, as noted above, if a variable has been declared as a string, all
12906 declarations must give it a string value. Similarly, if a variable has
12907 been declared as a string list, all subsequent declarations
12908 must give it a string list value.
12910 A @emph{variable reference} may take several forms:
12913 @item The simple variable name, for a variable in the current package (if any)
12914 or in the current project
12915 @item An expanded name, whose prefix is a context name.
12919 A @emph{context} may be one of the following:
12922 @item The name of an existing package in the current project
12923 @item The name of an imported project of the current project
12924 @item The name of an ancestor project (i.e., a project extended by the current
12925 project, either directly or indirectly)
12926 @item An expanded name whose prefix is an imported/parent project name, and
12927 whose selector is a package name in that project.
12931 A variable reference may be used in an expression.
12934 @subsection Attributes
12937 A project (and its packages) may have @emph{attributes} that define
12938 the project's properties. Some attributes have values that are strings;
12939 others have values that are string lists.
12941 There are two categories of attributes: @emph{simple attributes}
12942 and @emph{associative arrays} (@pxref{Associative Array Attributes}).
12944 Legal project attribute names, and attribute names for each legal package are
12945 listed below. Attributes names are case-insensitive.
12947 The following attributes are defined on projects (all are simple attributes):
12949 @multitable @columnfractions .4 .3
12950 @item @emph{Attribute Name}
12952 @item @code{Source_Files}
12954 @item @code{Source_Dirs}
12956 @item @code{Source_List_File}
12958 @item @code{Object_Dir}
12960 @item @code{Exec_Dir}
12962 @item @code{Excluded_Source_Dirs}
12964 @item @code{Excluded_Source_Files}
12966 @item @code{Excluded_Source_List_File}
12968 @item @code{Languages}
12972 @item @code{Library_Dir}
12974 @item @code{Library_Name}
12976 @item @code{Library_Kind}
12978 @item @code{Library_Version}
12980 @item @code{Library_Interface}
12982 @item @code{Library_Auto_Init}
12984 @item @code{Library_Options}
12986 @item @code{Library_Src_Dir}
12988 @item @code{Library_ALI_Dir}
12990 @item @code{Library_GCC}
12992 @item @code{Library_Symbol_File}
12994 @item @code{Library_Symbol_Policy}
12996 @item @code{Library_Reference_Symbol_File}
12998 @item @code{Externally_Built}
13003 The following attributes are defined for package @code{Naming}
13004 (@pxref{Naming Schemes}):
13006 @multitable @columnfractions .4 .2 .2 .2
13007 @item Attribute Name @tab Category @tab Index @tab Value
13008 @item @code{Spec_Suffix}
13009 @tab associative array
13012 @item @code{Body_Suffix}
13013 @tab associative array
13016 @item @code{Separate_Suffix}
13017 @tab simple attribute
13020 @item @code{Casing}
13021 @tab simple attribute
13024 @item @code{Dot_Replacement}
13025 @tab simple attribute
13029 @tab associative array
13033 @tab associative array
13036 @item @code{Specification_Exceptions}
13037 @tab associative array
13040 @item @code{Implementation_Exceptions}
13041 @tab associative array
13047 The following attributes are defined for packages @code{Builder},
13048 @code{Compiler}, @code{Binder},
13049 @code{Linker}, @code{Cross_Reference}, and @code{Finder}
13050 (@pxref{^Switches^Switches^ and Project Files}).
13052 @multitable @columnfractions .4 .2 .2 .2
13053 @item Attribute Name @tab Category @tab Index @tab Value
13054 @item @code{^Default_Switches^Default_Switches^}
13055 @tab associative array
13058 @item @code{^Switches^Switches^}
13059 @tab associative array
13065 In addition, package @code{Compiler} has a single string attribute
13066 @code{Local_Configuration_Pragmas} and package @code{Builder} has a single
13067 string attribute @code{Global_Configuration_Pragmas}.
13070 Each simple attribute has a default value: the empty string (for string-valued
13071 attributes) and the empty list (for string list-valued attributes).
13073 An attribute declaration defines a new value for an attribute.
13075 Examples of simple attribute declarations:
13077 @smallexample @c projectfile
13078 for Object_Dir use "objects";
13079 for Source_Dirs use ("units", "test/drivers");
13083 The syntax of a @dfn{simple attribute declaration} is similar to that of an
13084 attribute definition clause in Ada.
13086 Attributes references may be appear in expressions.
13087 The general form for such a reference is @code{<entity>'<attribute>}:
13088 Associative array attributes are functions. Associative
13089 array attribute references must have an argument that is a string literal.
13093 @smallexample @c projectfile
13095 Naming'Dot_Replacement
13096 Imported_Project'Source_Dirs
13097 Imported_Project.Naming'Casing
13098 Builder'^Default_Switches^Default_Switches^("Ada")
13102 The prefix of an attribute may be:
13104 @item @code{project} for an attribute of the current project
13105 @item The name of an existing package of the current project
13106 @item The name of an imported project
13107 @item The name of a parent project that is extended by the current project
13108 @item An expanded name whose prefix is imported/parent project name,
13109 and whose selector is a package name
13114 @smallexample @c projectfile
13117 for Source_Dirs use project'Source_Dirs & "units";
13118 for Source_Dirs use project'Source_Dirs & "test/drivers"
13124 In the first attribute declaration, initially the attribute @code{Source_Dirs}
13125 has the default value: an empty string list. After this declaration,
13126 @code{Source_Dirs} is a string list of one element: @code{"units"}.
13127 After the second attribute declaration @code{Source_Dirs} is a string list of
13128 two elements: @code{"units"} and @code{"test/drivers"}.
13130 Note: this example is for illustration only. In practice,
13131 the project file would contain only one attribute declaration:
13133 @smallexample @c projectfile
13134 for Source_Dirs use ("units", "test/drivers");
13137 @node Associative Array Attributes
13138 @subsection Associative Array Attributes
13141 Some attributes are defined as @emph{associative arrays}. An associative
13142 array may be regarded as a function that takes a string as a parameter
13143 and delivers a string or string list value as its result.
13145 Here are some examples of single associative array attribute associations:
13147 @smallexample @c projectfile
13148 for Body ("main") use "Main.ada";
13149 for ^Switches^Switches^ ("main.ada")
13151 "^-gnatv^-gnatv^");
13152 for ^Switches^Switches^ ("main.ada")
13153 use Builder'^Switches^Switches^ ("main.ada")
13158 Like untyped variables and simple attributes, associative array attributes
13159 may be declared several times. Each declaration supplies a new value for the
13160 attribute, and replaces the previous setting.
13163 An associative array attribute may be declared as a full associative array
13164 declaration, with the value of the same attribute in an imported or extended
13167 @smallexample @c projectfile
13169 for Default_Switches use Default.Builder'Default_Switches;
13174 In this example, @code{Default} must be either a project imported by the
13175 current project, or the project that the current project extends. If the
13176 attribute is in a package (in this case, in package @code{Builder}), the same
13177 package needs to be specified.
13180 A full associative array declaration replaces any other declaration for the
13181 attribute, including other full associative array declaration. Single
13182 associative array associations may be declare after a full associative
13183 declaration, modifying the value for a single association of the attribute.
13185 @node case Constructions
13186 @subsection @code{case} Constructions
13189 A @code{case} construction is used in a project file to effect conditional
13191 Here is a typical example:
13193 @smallexample @c projectfile
13196 type OS_Type is ("GNU/Linux", "Unix", "NT", "VMS");
13198 OS : OS_Type := external ("OS", "GNU/Linux");
13202 package Compiler is
13204 when "GNU/Linux" | "Unix" =>
13205 for ^Default_Switches^Default_Switches^ ("Ada")
13206 use ("^-gnath^-gnath^");
13208 for ^Default_Switches^Default_Switches^ ("Ada")
13209 use ("^-gnatP^-gnatP^");
13218 The syntax of a @code{case} construction is based on the Ada case statement
13219 (although there is no @code{null} construction for empty alternatives).
13221 The case expression must be a typed string variable.
13222 Each alternative comprises the reserved word @code{when}, either a list of
13223 literal strings separated by the @code{"|"} character or the reserved word
13224 @code{others}, and the @code{"=>"} token.
13225 Each literal string must belong to the string type that is the type of the
13227 An @code{others} alternative, if present, must occur last.
13229 After each @code{=>}, there are zero or more constructions. The only
13230 constructions allowed in a case construction are other case constructions,
13231 attribute declarations and variable declarations. String type declarations and
13232 package declarations are not allowed. Variable declarations are restricted to
13233 variables that have already been declared before the case construction.
13235 The value of the case variable is often given by an external reference
13236 (@pxref{External References in Project Files}).
13238 @c ****************************************
13239 @c * Objects and Sources in Project Files *
13240 @c ****************************************
13242 @node Objects and Sources in Project Files
13243 @section Objects and Sources in Project Files
13246 * Object Directory::
13248 * Source Directories::
13249 * Source File Names::
13253 Each project has exactly one object directory and one or more source
13254 directories. The source directories must contain at least one source file,
13255 unless the project file explicitly specifies that no source files are present
13256 (@pxref{Source File Names}).
13258 @node Object Directory
13259 @subsection Object Directory
13262 The object directory for a project is the directory containing the compiler's
13263 output (such as @file{ALI} files and object files) for the project's immediate
13266 The object directory is given by the value of the attribute @code{Object_Dir}
13267 in the project file.
13269 @smallexample @c projectfile
13270 for Object_Dir use "objects";
13274 The attribute @code{Object_Dir} has a string value, the path name of the object
13275 directory. The path name may be absolute or relative to the directory of the
13276 project file. This directory must already exist, and be readable and writable.
13278 By default, when the attribute @code{Object_Dir} is not given an explicit value
13279 or when its value is the empty string, the object directory is the same as the
13280 directory containing the project file.
13282 @node Exec Directory
13283 @subsection Exec Directory
13286 The exec directory for a project is the directory containing the executables
13287 for the project's main subprograms.
13289 The exec directory is given by the value of the attribute @code{Exec_Dir}
13290 in the project file.
13292 @smallexample @c projectfile
13293 for Exec_Dir use "executables";
13297 The attribute @code{Exec_Dir} has a string value, the path name of the exec
13298 directory. The path name may be absolute or relative to the directory of the
13299 project file. This directory must already exist, and be writable.
13301 By default, when the attribute @code{Exec_Dir} is not given an explicit value
13302 or when its value is the empty string, the exec directory is the same as the
13303 object directory of the project file.
13305 @node Source Directories
13306 @subsection Source Directories
13309 The source directories of a project are specified by the project file
13310 attribute @code{Source_Dirs}.
13312 This attribute's value is a string list. If the attribute is not given an
13313 explicit value, then there is only one source directory, the one where the
13314 project file resides.
13316 A @code{Source_Dirs} attribute that is explicitly defined to be the empty list,
13319 @smallexample @c projectfile
13320 for Source_Dirs use ();
13324 indicates that the project contains no source files.
13326 Otherwise, each string in the string list designates one or more
13327 source directories.
13329 @smallexample @c projectfile
13330 for Source_Dirs use ("sources", "test/drivers");
13334 If a string in the list ends with @code{"/**"}, then the directory whose path
13335 name precedes the two asterisks, as well as all its subdirectories
13336 (recursively), are source directories.
13338 @smallexample @c projectfile
13339 for Source_Dirs use ("/system/sources/**");
13343 Here the directory @code{/system/sources} and all of its subdirectories
13344 (recursively) are source directories.
13346 To specify that the source directories are the directory of the project file
13347 and all of its subdirectories, you can declare @code{Source_Dirs} as follows:
13348 @smallexample @c projectfile
13349 for Source_Dirs use ("./**");
13353 Each of the source directories must exist and be readable.
13355 @node Source File Names
13356 @subsection Source File Names
13359 In a project that contains source files, their names may be specified by the
13360 attributes @code{Source_Files} (a string list) or @code{Source_List_File}
13361 (a string). Source file names never include any directory information.
13363 If the attribute @code{Source_Files} is given an explicit value, then each
13364 element of the list is a source file name.
13366 @smallexample @c projectfile
13367 for Source_Files use ("main.adb");
13368 for Source_Files use ("main.adb", "pack1.ads", "pack2.adb");
13372 If the attribute @code{Source_Files} is not given an explicit value,
13373 but the attribute @code{Source_List_File} is given a string value,
13374 then the source file names are contained in the text file whose path name
13375 (absolute or relative to the directory of the project file) is the
13376 value of the attribute @code{Source_List_File}.
13378 Each line in the file that is not empty or is not a comment
13379 contains a source file name.
13381 @smallexample @c projectfile
13382 for Source_List_File use "source_list.txt";
13386 By default, if neither the attribute @code{Source_Files} nor the attribute
13387 @code{Source_List_File} is given an explicit value, then each file in the
13388 source directories that conforms to the project's naming scheme
13389 (@pxref{Naming Schemes}) is an immediate source of the project.
13391 A warning is issued if both attributes @code{Source_Files} and
13392 @code{Source_List_File} are given explicit values. In this case, the attribute
13393 @code{Source_Files} prevails.
13395 Each source file name must be the name of one existing source file
13396 in one of the source directories.
13398 A @code{Source_Files} attribute whose value is an empty list
13399 indicates that there are no source files in the project.
13401 If the order of the source directories is known statically, that is if
13402 @code{"/**"} is not used in the string list @code{Source_Dirs}, then there may
13403 be several files with the same source file name. In this case, only the file
13404 in the first directory is considered as an immediate source of the project
13405 file. If the order of the source directories is not known statically, it is
13406 an error to have several files with the same source file name.
13408 Projects can be specified to have no Ada source
13409 files: the value of @code{Source_Dirs} or @code{Source_Files} may be an empty
13410 list, or the @code{"Ada"} may be absent from @code{Languages}:
13412 @smallexample @c projectfile
13413 for Source_Dirs use ();
13414 for Source_Files use ();
13415 for Languages use ("C", "C++");
13419 Otherwise, a project must contain at least one immediate source.
13421 Projects with no source files are useful as template packages
13422 (@pxref{Packages in Project Files}) for other projects; in particular to
13423 define a package @code{Naming} (@pxref{Naming Schemes}).
13425 @c ****************************
13426 @c * Importing Projects *
13427 @c ****************************
13429 @node Importing Projects
13430 @section Importing Projects
13431 @cindex @code{ADA_PROJECT_PATH}
13432 @cindex @code{GPR_PROJECT_PATH}
13435 An immediate source of a project P may depend on source files that
13436 are neither immediate sources of P nor in the predefined library.
13437 To get this effect, P must @emph{import} the projects that contain the needed
13440 @smallexample @c projectfile
13442 with "project1", "utilities.gpr";
13443 with "/namings/apex.gpr";
13450 As can be seen in this example, the syntax for importing projects is similar
13451 to the syntax for importing compilation units in Ada. However, project files
13452 use literal strings instead of names, and the @code{with} clause identifies
13453 project files rather than packages.
13455 Each literal string is the file name or path name (absolute or relative) of a
13456 project file. If a string corresponds to a file name, with no path or a
13457 relative path, then its location is determined by the @emph{project path}. The
13458 latter can be queried using @code{gnatls -v}. It contains:
13462 In first position, the directory containing the current project file.
13464 In last position, the default project directory. This default project directory
13465 is part of the GNAT installation and is the standard place to install project
13466 files giving access to standard support libraries.
13468 @ref{Installing a library}
13472 In between, all the directories referenced in the
13473 ^environment variables^logical names^ @env{GPR_PROJECT_PATH}
13474 and @env{ADA_PROJECT_PATH} if they exist, and in that order.
13478 If a relative pathname is used, as in
13480 @smallexample @c projectfile
13485 then the full path for the project is constructed by concatenating this
13486 relative path to those in the project path, in order, until a matching file is
13487 found. Any symbolic link will be fully resolved in the directory of the
13488 importing project file before the imported project file is examined.
13490 If the @code{with}'ed project file name does not have an extension,
13491 the default is @file{^.gpr^.GPR^}. If a file with this extension is not found,
13492 then the file name as specified in the @code{with} clause (no extension) will
13493 be used. In the above example, if a file @code{project1.gpr} is found, then it
13494 will be used; otherwise, if a file @code{^project1^PROJECT1^} exists
13495 then it will be used; if neither file exists, this is an error.
13497 A warning is issued if the name of the project file does not match the
13498 name of the project; this check is case insensitive.
13500 Any source file that is an immediate source of the imported project can be
13501 used by the immediate sources of the importing project, transitively. Thus
13502 if @code{A} imports @code{B}, and @code{B} imports @code{C}, the immediate
13503 sources of @code{A} may depend on the immediate sources of @code{C}, even if
13504 @code{A} does not import @code{C} explicitly. However, this is not recommended,
13505 because if and when @code{B} ceases to import @code{C}, some sources in
13506 @code{A} will no longer compile.
13508 A side effect of this capability is that normally cyclic dependencies are not
13509 permitted: if @code{A} imports @code{B} (directly or indirectly) then @code{B}
13510 is not allowed to import @code{A}. However, there are cases when cyclic
13511 dependencies would be beneficial. For these cases, another form of import
13512 between projects exists, the @code{limited with}: a project @code{A} that
13513 imports a project @code{B} with a straight @code{with} may also be imported,
13514 directly or indirectly, by @code{B} on the condition that imports from @code{B}
13515 to @code{A} include at least one @code{limited with}.
13517 @smallexample @c 0projectfile
13523 limited with "../a/a.gpr";
13531 limited with "../a/a.gpr";
13537 In the above legal example, there are two project cycles:
13540 @item A -> C -> D -> A
13544 In each of these cycle there is one @code{limited with}: import of @code{A}
13545 from @code{B} and import of @code{A} from @code{D}.
13547 The difference between straight @code{with} and @code{limited with} is that
13548 the name of a project imported with a @code{limited with} cannot be used in the
13549 project that imports it. In particular, its packages cannot be renamed and
13550 its variables cannot be referred to.
13552 An exception to the above rules for @code{limited with} is that for the main
13553 project specified to @command{gnatmake} or to the @command{GNAT} driver a
13554 @code{limited with} is equivalent to a straight @code{with}. For example,
13555 in the example above, projects @code{B} and @code{D} could not be main
13556 projects for @command{gnatmake} or to the @command{GNAT} driver, because they
13557 each have a @code{limited with} that is the only one in a cycle of importing
13560 @c *********************
13561 @c * Project Extension *
13562 @c *********************
13564 @node Project Extension
13565 @section Project Extension
13568 During development of a large system, it is sometimes necessary to use
13569 modified versions of some of the source files, without changing the original
13570 sources. This can be achieved through the @emph{project extension} facility.
13572 @smallexample @c projectfile
13573 project Modified_Utilities extends "/baseline/utilities.gpr" is @dots{}
13577 A project extension declaration introduces an extending project
13578 (the @emph{child}) and a project being extended (the @emph{parent}).
13580 By default, a child project inherits all the sources of its parent.
13581 However, inherited sources can be overridden: a unit in a parent is hidden
13582 by a unit of the same name in the child.
13584 Inherited sources are considered to be sources (but not immediate sources)
13585 of the child project; see @ref{Project File Syntax}.
13587 An inherited source file retains any switches specified in the parent project.
13589 For example if the project @code{Utilities} contains the spec and the
13590 body of an Ada package @code{Util_IO}, then the project
13591 @code{Modified_Utilities} can contain a new body for package @code{Util_IO}.
13592 The original body of @code{Util_IO} will not be considered in program builds.
13593 However, the package spec will still be found in the project
13596 A child project can have only one parent, except when it is qualified as
13597 abstract. But it may import any number of other projects.
13599 A project is not allowed to import directly or indirectly at the same time a
13600 child project and any of its ancestors.
13602 @c *******************************
13603 @c * Project Hierarchy Extension *
13604 @c *******************************
13606 @node Project Hierarchy Extension
13607 @section Project Hierarchy Extension
13610 When extending a large system spanning multiple projects, it is often
13611 inconvenient to extend every project in the hierarchy that is impacted by a
13612 small change introduced. In such cases, it is possible to create a virtual
13613 extension of entire hierarchy using @code{extends all} relationship.
13615 When the project is extended using @code{extends all} inheritance, all projects
13616 that are imported by it, both directly and indirectly, are considered virtually
13617 extended. That is, the Project Manager creates "virtual projects"
13618 that extend every project in the hierarchy; all these virtual projects have
13619 no sources of their own and have as object directory the object directory of
13620 the root of "extending all" project.
13622 It is possible to explicitly extend one or more projects in the hierarchy
13623 in order to modify the sources. These extending projects must be imported by
13624 the "extending all" project, which will replace the corresponding virtual
13625 projects with the explicit ones.
13627 When building such a project hierarchy extension, the Project Manager will
13628 ensure that both modified sources and sources in virtual extending projects
13629 that depend on them, are recompiled.
13631 By means of example, consider the following hierarchy of projects.
13635 project A, containing package P1
13637 project B importing A and containing package P2 which depends on P1
13639 project C importing B and containing package P3 which depends on P2
13643 We want to modify packages P1 and P3.
13645 This project hierarchy will need to be extended as follows:
13649 Create project A1 that extends A, placing modified P1 there:
13651 @smallexample @c 0projectfile
13652 project A1 extends "(@dots{})/A" is
13657 Create project C1 that "extends all" C and imports A1, placing modified
13660 @smallexample @c 0projectfile
13661 with "(@dots{})/A1";
13662 project C1 extends all "(@dots{})/C" is
13667 When you build project C1, your entire modified project space will be
13668 recompiled, including the virtual project B1 that has been impacted by the
13669 "extending all" inheritance of project C.
13671 Note that if a Library Project in the hierarchy is virtually extended,
13672 the virtual project that extends the Library Project is not a Library Project.
13674 @c ****************************************
13675 @c * External References in Project Files *
13676 @c ****************************************
13678 @node External References in Project Files
13679 @section External References in Project Files
13682 A project file may contain references to external variables; such references
13683 are called @emph{external references}.
13685 An external variable is either defined as part of the environment (an
13686 environment variable in Unix, for example) or else specified on the command
13687 line via the @option{^-X^/EXTERNAL_REFERENCE=^@emph{vbl}=@emph{value}} switch.
13688 If both, then the command line value is used.
13690 The value of an external reference is obtained by means of the built-in
13691 function @code{external}, which returns a string value.
13692 This function has two forms:
13694 @item @code{external (external_variable_name)}
13695 @item @code{external (external_variable_name, default_value)}
13699 Each parameter must be a string literal. For example:
13701 @smallexample @c projectfile
13703 external ("OS", "GNU/Linux")
13707 In the form with one parameter, the function returns the value of
13708 the external variable given as parameter. If this name is not present in the
13709 environment, the function returns an empty string.
13711 In the form with two string parameters, the second argument is
13712 the value returned when the variable given as the first argument is not
13713 present in the environment. In the example above, if @code{"OS"} is not
13714 the name of ^an environment variable^a logical name^ and is not passed on
13715 the command line, then the returned value is @code{"GNU/Linux"}.
13717 An external reference may be part of a string expression or of a string
13718 list expression, and can therefore appear in a variable declaration or
13719 an attribute declaration.
13721 @smallexample @c projectfile
13723 type Mode_Type is ("Debug", "Release");
13724 Mode : Mode_Type := external ("MODE");
13731 @c *****************************
13732 @c * Packages in Project Files *
13733 @c *****************************
13735 @node Packages in Project Files
13736 @section Packages in Project Files
13739 A @emph{package} defines the settings for project-aware tools within a
13741 For each such tool one can declare a package; the names for these
13742 packages are preset (@pxref{Packages}).
13743 A package may contain variable declarations, attribute declarations, and case
13746 @smallexample @c projectfile
13749 package Builder is -- used by gnatmake
13750 for ^Default_Switches^Default_Switches^ ("Ada")
13759 The syntax of package declarations mimics that of package in Ada.
13761 Most of the packages have an attribute
13762 @code{^Default_Switches^Default_Switches^}.
13763 This attribute is an associative array, and its value is a string list.
13764 The index of the associative array is the name of a programming language (case
13765 insensitive). This attribute indicates the ^switch^switch^
13766 or ^switches^switches^ to be used
13767 with the corresponding tool.
13769 Some packages also have another attribute, @code{^Switches^Switches^},
13770 an associative array whose value is a string list.
13771 The index is the name of a source file.
13772 This attribute indicates the ^switch^switch^
13773 or ^switches^switches^ to be used by the corresponding
13774 tool when dealing with this specific file.
13776 Further information on these ^switch^switch^-related attributes is found in
13777 @ref{^Switches^Switches^ and Project Files}.
13779 A package may be declared as a @emph{renaming} of another package; e.g., from
13780 the project file for an imported project.
13782 @smallexample @c projectfile
13784 with "/global/apex.gpr";
13786 package Naming renames Apex.Naming;
13793 Packages that are renamed in other project files often come from project files
13794 that have no sources: they are just used as templates. Any modification in the
13795 template will be reflected automatically in all the project files that rename
13796 a package from the template.
13798 In addition to the tool-oriented packages, you can also declare a package
13799 named @code{Naming} to establish specialized source file naming conventions
13800 (@pxref{Naming Schemes}).
13802 @c ************************************
13803 @c * Variables from Imported Projects *
13804 @c ************************************
13806 @node Variables from Imported Projects
13807 @section Variables from Imported Projects
13810 An attribute or variable defined in an imported or parent project can
13811 be used in expressions in the importing / extending project.
13812 Such an attribute or variable is denoted by an expanded name whose prefix
13813 is either the name of the project or the expanded name of a package within
13816 @smallexample @c projectfile
13819 project Main extends "base" is
13820 Var1 := Imported.Var;
13821 Var2 := Base.Var & ".new";
13826 for ^Default_Switches^Default_Switches^ ("Ada")
13827 use Imported.Builder'Ada_^Switches^Switches^ &
13828 "^-gnatg^-gnatg^" &
13834 package Compiler is
13835 for ^Default_Switches^Default_Switches^ ("Ada")
13836 use Base.Compiler'Ada_^Switches^Switches^;
13847 The value of @code{Var1} is a copy of the variable @code{Var} defined
13848 in the project file @file{"imported.gpr"}
13850 the value of @code{Var2} is a copy of the value of variable @code{Var}
13851 defined in the project file @file{base.gpr}, concatenated with @code{".new"}
13853 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13854 @code{Builder} is a string list that includes in its value a copy of the value
13855 of @code{Ada_^Switches^Switches^} defined in the @code{Builder} package
13856 in project file @file{imported.gpr} plus two new elements:
13857 @option{"^-gnatg^-gnatg^"}
13858 and @option{"^-v^-v^"};
13860 attribute @code{^Default_Switches^Default_Switches^ ("Ada")} in package
13861 @code{Compiler} is a copy of the variable @code{Ada_^Switches^Switches^}
13862 defined in the @code{Compiler} package in project file @file{base.gpr},
13863 the project being extended.
13866 @c ******************
13867 @c * Naming Schemes *
13868 @c ******************
13870 @node Naming Schemes
13871 @section Naming Schemes
13874 Sometimes an Ada software system is ported from a foreign compilation
13875 environment to GNAT, and the file names do not use the default GNAT
13876 conventions. Instead of changing all the file names (which for a variety
13877 of reasons might not be possible), you can define the relevant file
13878 naming scheme in the @code{Naming} package in your project file.
13881 Note that the use of pragmas described in
13882 @ref{Alternative File Naming Schemes} by mean of a configuration
13883 pragmas file is not supported when using project files. You must use
13884 the features described in this paragraph. You can however use specify
13885 other configuration pragmas (@pxref{Specifying Configuration Pragmas}).
13888 For example, the following
13889 package models the Apex file naming rules:
13891 @smallexample @c projectfile
13894 for Casing use "lowercase";
13895 for Dot_Replacement use ".";
13896 for Spec_Suffix ("Ada") use ".1.ada";
13897 for Body_Suffix ("Ada") use ".2.ada";
13904 For example, the following package models the HP Ada file naming rules:
13906 @smallexample @c projectfile
13909 for Casing use "lowercase";
13910 for Dot_Replacement use "__";
13911 for Spec_Suffix ("Ada") use "_.^ada^ada^";
13912 for Body_Suffix ("Ada") use ".^ada^ada^";
13918 (Note that @code{Casing} is @code{"lowercase"} because GNAT gets the file
13919 names in lower case)
13923 You can define the following attributes in package @code{Naming}:
13927 @item @code{Casing}
13928 This must be a string with one of the three values @code{"lowercase"},
13929 @code{"uppercase"} or @code{"mixedcase"}; these strings are case insensitive.
13932 If @code{Casing} is not specified, then the default is @code{"lowercase"}.
13934 @item @code{Dot_Replacement}
13935 This must be a string whose value satisfies the following conditions:
13938 @item It must not be empty
13939 @item It cannot start or end with an alphanumeric character
13940 @item It cannot be a single underscore
13941 @item It cannot start with an underscore followed by an alphanumeric
13942 @item It cannot contain a dot @code{'.'} except if the entire string
13947 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
13949 @item @code{Spec_Suffix}
13950 This is an associative array (indexed by the programming language name, case
13951 insensitive) whose value is a string that must satisfy the following
13955 @item It must not be empty
13956 @item It must include at least one dot
13959 If @code{Spec_Suffix ("Ada")} is not specified, then the default is
13960 @code{"^.ads^.ADS^"}.
13962 @item @code{Body_Suffix}
13963 This is an associative array (indexed by the programming language name, case
13964 insensitive) whose value is a string that must satisfy the following
13968 @item It must not be empty
13969 @item It must include at least one dot
13970 @item It cannot be the same as @code{Spec_Suffix ("Ada")}
13973 If @code{Body_Suffix ("Ada")} and @code{Spec_Suffix ("Ada")} end with the
13974 same string, then a file name that ends with the longest of these two suffixes
13975 will be a body if the longest suffix is @code{Body_Suffix ("Ada")} or a spec
13976 if the longest suffix is @code{Spec_Suffix ("Ada")}.
13978 If the suffix does not start with a '.', a file with a name exactly equal
13979 to the suffix will also be part of the project (for instance if you define
13980 the suffix as @code{Makefile}, a file called @file{Makefile} will be part
13981 of the project. This is not interesting in general when using projects to
13982 compile. However, it might become useful when a project is also used to
13983 find the list of source files in an editor, like the GNAT Programming System
13986 If @code{Body_Suffix ("Ada")} is not specified, then the default is
13987 @code{"^.adb^.ADB^"}.
13989 @item @code{Separate_Suffix}
13990 This must be a string whose value satisfies the same conditions as
13991 @code{Body_Suffix}. The same "longest suffix" rules apply.
13994 If @code{Separate_Suffix ("Ada")} is not specified, then it defaults to same
13995 value as @code{Body_Suffix ("Ada")}.
13999 You can use the associative array attribute @code{Spec} to define
14000 the source file name for an individual Ada compilation unit's spec. The array
14001 index must be a string literal that identifies the Ada unit (case insensitive).
14002 The value of this attribute must be a string that identifies the file that
14003 contains this unit's spec (case sensitive or insensitive depending on the
14006 @smallexample @c projectfile
14007 for Spec ("MyPack.MyChild") use "mypack.mychild.spec";
14010 When the source file contains several units, you can indicate at what
14011 position the unit occurs in the file, with the following. The first unit
14012 in the file has index 1
14014 @smallexample @c projectfile
14015 for Body ("top") use "foo.a" at 1;
14016 for Body ("foo") use "foo.a" at 2;
14021 You can use the associative array attribute @code{Body} to
14022 define the source file name for an individual Ada compilation unit's body
14023 (possibly a subunit). The array index must be a string literal that identifies
14024 the Ada unit (case insensitive). The value of this attribute must be a string
14025 that identifies the file that contains this unit's body or subunit (case
14026 sensitive or insensitive depending on the operating system).
14028 @smallexample @c projectfile
14029 for Body ("MyPack.MyChild") use "mypack.mychild.body";
14033 @c ********************
14034 @c * Library Projects *
14035 @c ********************
14037 @node Library Projects
14038 @section Library Projects
14041 @emph{Library projects} are projects whose object code is placed in a library.
14042 (Note that this facility is not yet supported on all platforms).
14044 @code{gnatmake} or @code{gprbuild} will collect all object files into a
14045 single archive, which might either be a shared or a static library. This
14046 library can later on be linked with multiple executables, potentially
14047 reducing their sizes.
14049 If your project file specifies languages other than Ada, but you are still
14050 using @code{gnatmake} to compile and link, the latter will not try to
14051 compile your sources other than Ada (you should use @code{gprbuild} if that
14052 is your intent). However, @code{gnatmake} will automatically link all object
14053 files found in the object directory, whether or not they were compiled from
14054 an Ada source file. This specific behavior only applies when multiple
14055 languages are specified.
14057 To create a library project, you need to define in its project file
14058 two project-level attributes: @code{Library_Name} and @code{Library_Dir}.
14059 Additionally, you may define other library-related attributes such as
14060 @code{Library_Kind}, @code{Library_Version}, @code{Library_Interface},
14061 @code{Library_Auto_Init}, @code{Library_Options} and @code{Library_GCC}.
14063 The @code{Library_Name} attribute has a string value. There is no restriction
14064 on the name of a library. It is the responsibility of the developer to
14065 choose a name that will be accepted by the platform. It is recommended to
14066 choose names that could be Ada identifiers; such names are almost guaranteed
14067 to be acceptable on all platforms.
14069 The @code{Library_Dir} attribute has a string value that designates the path
14070 (absolute or relative) of the directory where the library will reside.
14071 It must designate an existing directory, and this directory must be writable,
14072 different from the project's object directory and from any source directory
14073 in the project tree.
14075 If both @code{Library_Name} and @code{Library_Dir} are specified and
14076 are legal, then the project file defines a library project. The optional
14077 library-related attributes are checked only for such project files.
14079 The @code{Library_Kind} attribute has a string value that must be one of the
14080 following (case insensitive): @code{"static"}, @code{"dynamic"} or
14081 @code{"relocatable"} (which is a synonym for @code{"dynamic"}). If this
14082 attribute is not specified, the library is a static library, that is
14083 an archive of object files that can be potentially linked into a
14084 static executable. Otherwise, the library may be dynamic or
14085 relocatable, that is a library that is loaded only at the start of execution.
14087 If you need to build both a static and a dynamic library, you should use two
14088 different object directories, since in some cases some extra code needs to
14089 be generated for the latter. For such cases, it is recommended to either use
14090 two different project files, or a single one which uses external variables
14091 to indicate what kind of library should be build.
14093 The @code{Library_ALI_Dir} attribute may be specified to indicate the
14094 directory where the ALI files of the library will be copied. When it is
14095 not specified, the ALI files are copied to the directory specified in
14096 attribute @code{Library_Dir}. The directory specified by @code{Library_ALI_Dir}
14097 must be writable and different from the project's object directory and from
14098 any source directory in the project tree.
14100 The @code{Library_Version} attribute has a string value whose interpretation
14101 is platform dependent. It has no effect on VMS and Windows. On Unix, it is
14102 used only for dynamic/relocatable libraries as the internal name of the
14103 library (the @code{"soname"}). If the library file name (built from the
14104 @code{Library_Name}) is different from the @code{Library_Version}, then the
14105 library file will be a symbolic link to the actual file whose name will be
14106 @code{Library_Version}.
14110 @smallexample @c projectfile
14116 for Library_Dir use "lib_dir";
14117 for Library_Name use "dummy";
14118 for Library_Kind use "relocatable";
14119 for Library_Version use "libdummy.so." & Version;
14126 Directory @file{lib_dir} will contain the internal library file whose name
14127 will be @file{libdummy.so.1}, and @file{libdummy.so} will be a symbolic link to
14128 @file{libdummy.so.1}.
14130 When @command{gnatmake} detects that a project file
14131 is a library project file, it will check all immediate sources of the project
14132 and rebuild the library if any of the sources have been recompiled.
14134 Standard project files can import library project files. In such cases,
14135 the libraries will only be rebuilt if some of its sources are recompiled
14136 because they are in the closure of some other source in an importing project.
14137 Sources of the library project files that are not in such a closure will
14138 not be checked, unless the full library is checked, because one of its sources
14139 needs to be recompiled.
14141 For instance, assume the project file @code{A} imports the library project file
14142 @code{L}. The immediate sources of A are @file{a1.adb}, @file{a2.ads} and
14143 @file{a2.adb}. The immediate sources of L are @file{l1.ads}, @file{l1.adb},
14144 @file{l2.ads}, @file{l2.adb}.
14146 If @file{l1.adb} has been modified, then the library associated with @code{L}
14147 will be rebuilt when compiling all the immediate sources of @code{A} only
14148 if @file{a1.ads}, @file{a2.ads} or @file{a2.adb} includes a statement
14151 To be sure that all the sources in the library associated with @code{L} are
14152 up to date, and that all the sources of project @code{A} are also up to date,
14153 the following two commands needs to be used:
14160 When a library is built or rebuilt, an attempt is made first to delete all
14161 files in the library directory.
14162 All @file{ALI} files will also be copied from the object directory to the
14163 library directory. To build executables, @command{gnatmake} will use the
14164 library rather than the individual object files.
14167 It is also possible to create library project files for third-party libraries
14168 that are precompiled and cannot be compiled locally thanks to the
14169 @code{externally_built} attribute. (See @ref{Installing a library}).
14172 @c *******************************
14173 @c * Stand-alone Library Projects *
14174 @c *******************************
14176 @node Stand-alone Library Projects
14177 @section Stand-alone Library Projects
14180 A Stand-alone Library is a library that contains the necessary code to
14181 elaborate the Ada units that are included in the library. A Stand-alone
14182 Library is suitable to be used in an executable when the main is not
14183 in Ada. However, Stand-alone Libraries may also be used with an Ada main
14186 A Stand-alone Library Project is a Library Project where the library is
14187 a Stand-alone Library.
14189 To be a Stand-alone Library Project, in addition to the two attributes
14190 that make a project a Library Project (@code{Library_Name} and
14191 @code{Library_Dir}, see @ref{Library Projects}), the attribute
14192 @code{Library_Interface} must be defined.
14194 @smallexample @c projectfile
14196 for Library_Dir use "lib_dir";
14197 for Library_Name use "dummy";
14198 for Library_Interface use ("int1", "int1.child");
14202 Attribute @code{Library_Interface} has a nonempty string list value,
14203 each string in the list designating a unit contained in an immediate source
14204 of the project file.
14206 When a Stand-alone Library is built, first the binder is invoked to build
14207 a package whose name depends on the library name
14208 (^b~dummy.ads/b^B$DUMMY.ADS/B^ in the example above).
14209 This binder-generated package includes initialization and
14210 finalization procedures whose
14211 names depend on the library name (dummyinit and dummyfinal in the example
14212 above). The object corresponding to this package is included in the library.
14214 A dynamic or relocatable Stand-alone Library is automatically initialized
14215 if automatic initialization of Stand-alone Libraries is supported on the
14216 platform and if attribute @code{Library_Auto_Init} is not specified or
14217 is specified with the value "true". A static Stand-alone Library is never
14218 automatically initialized.
14220 Single string attribute @code{Library_Auto_Init} may be specified with only
14221 two possible values: "false" or "true" (case-insensitive). Specifying
14222 "false" for attribute @code{Library_Auto_Init} will prevent automatic
14223 initialization of dynamic or relocatable libraries.
14225 When a non-automatically initialized Stand-alone Library is used
14226 in an executable, its initialization procedure must be called before
14227 any service of the library is used.
14228 When the main subprogram is in Ada, it may mean that the initialization
14229 procedure has to be called during elaboration of another package.
14231 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
14232 (those that are listed in attribute @code{Library_Interface}) are copied to
14233 the Library Directory. As a consequence, only the Interface Units may be
14234 imported from Ada units outside of the library. If other units are imported,
14235 the binding phase will fail.
14237 When a Stand-Alone Library is bound, the switches that are specified in
14238 the attribute @code{Default_Switches ("Ada")} in package @code{Binder} are
14239 used in the call to @command{gnatbind}.
14241 The string list attribute @code{Library_Options} may be used to specified
14242 additional switches to the call to @command{gcc} to link the library.
14244 The attribute @code{Library_Src_Dir}, may be specified for a
14245 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
14246 single string value. Its value must be the path (absolute or relative to the
14247 project directory) of an existing directory. This directory cannot be the
14248 object directory or one of the source directories, but it can be the same as
14249 the library directory. The sources of the Interface
14250 Units of the library, necessary to an Ada client of the library, will be
14251 copied to the designated directory, called Interface Copy directory.
14252 These sources includes the specs of the Interface Units, but they may also
14253 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
14254 are used, or when there is a generic units in the spec. Before the sources
14255 are copied to the Interface Copy directory, an attempt is made to delete all
14256 files in the Interface Copy directory.
14258 @c *************************************
14259 @c * Switches Related to Project Files *
14260 @c *************************************
14261 @node Switches Related to Project Files
14262 @section Switches Related to Project Files
14265 The following switches are used by GNAT tools that support project files:
14269 @item ^-P^/PROJECT_FILE=^@var{project}
14270 @cindex @option{^-P^/PROJECT_FILE^} (any project-aware tool)
14271 Indicates the name of a project file. This project file will be parsed with
14272 the verbosity indicated by @option{^-vP^MESSAGE_PROJECT_FILES=^@emph{x}},
14273 if any, and using the external references indicated
14274 by @option{^-X^/EXTERNAL_REFERENCE^} switches, if any.
14276 There may zero, one or more spaces between @option{-P} and @var{project}.
14280 There must be only one @option{^-P^/PROJECT_FILE^} switch on the command line.
14283 Since the Project Manager parses the project file only after all the switches
14284 on the command line are checked, the order of the switches
14285 @option{^-P^/PROJECT_FILE^},
14286 @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}}
14287 or @option{^-X^/EXTERNAL_REFERENCE^} is not significant.
14289 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
14290 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (any project-aware tool)
14291 Indicates that external variable @var{name} has the value @var{value}.
14292 The Project Manager will use this value for occurrences of
14293 @code{external(name)} when parsing the project file.
14297 If @var{name} or @var{value} includes a space, then @var{name=value} should be
14298 put between quotes.
14306 Several @option{^-X^/EXTERNAL_REFERENCE^} switches can be used simultaneously.
14307 If several @option{^-X^/EXTERNAL_REFERENCE^} switches specify the same
14308 @var{name}, only the last one is used.
14311 An external variable specified with a @option{^-X^/EXTERNAL_REFERENCE^} switch
14312 takes precedence over the value of the same name in the environment.
14314 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
14315 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (any project-aware tool)
14316 Indicates the verbosity of the parsing of GNAT project files.
14319 @option{-vP0} means Default;
14320 @option{-vP1} means Medium;
14321 @option{-vP2} means High.
14325 There are three possible options for this qualifier: DEFAULT, MEDIUM and
14330 The default is ^Default^DEFAULT^: no output for syntactically correct
14333 If several @option{^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}} switches are present,
14334 only the last one is used.
14336 @item ^-aP^/ADD_PROJECT_SEARCH_DIR=^<dir>
14337 @cindex @option{^-aP^/ADD_PROJECT_SEARCH_DIR=^} (any project-aware tool)
14338 Add directory <dir> at the beginning of the project search path, in order,
14339 after the current working directory.
14343 @cindex @option{-eL} (any project-aware tool)
14344 Follow all symbolic links when processing project files.
14347 @item ^--subdirs^/SUBDIRS^=<subdir>
14348 @cindex @option{^--subdirs^/SUBDIRS^=} (gnatmake and gnatclean)
14349 This switch is recognized by gnatmake and gnatclean. It indicate that the real
14350 directories (except the source directories) are the subdirectories <subdir>
14351 of the directories specified in the project files. This applies in particular
14352 to object directories, library directories and exec directories. If the
14353 subdirectories do not exist, they are created automatically.
14357 @c **********************************
14358 @c * Tools Supporting Project Files *
14359 @c **********************************
14361 @node Tools Supporting Project Files
14362 @section Tools Supporting Project Files
14365 * gnatmake and Project Files::
14366 * The GNAT Driver and Project Files::
14369 @node gnatmake and Project Files
14370 @subsection gnatmake and Project Files
14373 This section covers several topics related to @command{gnatmake} and
14374 project files: defining ^switches^switches^ for @command{gnatmake}
14375 and for the tools that it invokes; specifying configuration pragmas;
14376 the use of the @code{Main} attribute; building and rebuilding library project
14380 * ^Switches^Switches^ and Project Files::
14381 * Specifying Configuration Pragmas::
14382 * Project Files and Main Subprograms::
14383 * Library Project Files::
14386 @node ^Switches^Switches^ and Project Files
14387 @subsubsection ^Switches^Switches^ and Project Files
14390 It is not currently possible to specify VMS style qualifiers in the project
14391 files; only Unix style ^switches^switches^ may be specified.
14395 For each of the packages @code{Builder}, @code{Compiler}, @code{Binder}, and
14396 @code{Linker}, you can specify a @code{^Default_Switches^Default_Switches^}
14397 attribute, a @code{^Switches^Switches^} attribute, or both;
14398 as their names imply, these ^switch^switch^-related
14399 attributes affect the ^switches^switches^ that are used for each of these GNAT
14401 @command{gnatmake} is invoked. As will be explained below, these
14402 component-specific ^switches^switches^ precede
14403 the ^switches^switches^ provided on the @command{gnatmake} command line.
14405 The @code{^Default_Switches^Default_Switches^} attribute is an associative
14406 array indexed by language name (case insensitive) whose value is a string list.
14409 @smallexample @c projectfile
14411 package Compiler is
14412 for ^Default_Switches^Default_Switches^ ("Ada")
14413 use ("^-gnaty^-gnaty^",
14420 The @code{^Switches^Switches^} attribute is also an associative array,
14421 indexed by a file name (which may or may not be case sensitive, depending
14422 on the operating system) whose value is a string list. For example:
14424 @smallexample @c projectfile
14427 for ^Switches^Switches^ ("main1.adb")
14429 for ^Switches^Switches^ ("main2.adb")
14436 For the @code{Builder} package, the file names must designate source files
14437 for main subprograms. For the @code{Binder} and @code{Linker} packages, the
14438 file names must designate @file{ALI} or source files for main subprograms.
14439 In each case just the file name without an explicit extension is acceptable.
14441 For each tool used in a program build (@command{gnatmake}, the compiler, the
14442 binder, and the linker), the corresponding package @dfn{contributes} a set of
14443 ^switches^switches^ for each file on which the tool is invoked, based on the
14444 ^switch^switch^-related attributes defined in the package.
14445 In particular, the ^switches^switches^
14446 that each of these packages contributes for a given file @var{f} comprise:
14450 the value of attribute @code{^Switches^Switches^ (@var{f})},
14451 if it is specified in the package for the given file,
14453 otherwise, the value of @code{^Default_Switches^Default_Switches^ ("Ada")},
14454 if it is specified in the package.
14458 If neither of these attributes is defined in the package, then the package does
14459 not contribute any ^switches^switches^ for the given file.
14461 When @command{gnatmake} is invoked on a file, the ^switches^switches^ comprise
14462 two sets, in the following order: those contributed for the file
14463 by the @code{Builder} package;
14464 and the switches passed on the command line.
14466 When @command{gnatmake} invokes a tool (compiler, binder, linker) on a file,
14467 the ^switches^switches^ passed to the tool comprise three sets,
14468 in the following order:
14472 the applicable ^switches^switches^ contributed for the file
14473 by the @code{Builder} package in the project file supplied on the command line;
14476 those contributed for the file by the package (in the relevant project file --
14477 see below) corresponding to the tool; and
14480 the applicable switches passed on the command line.
14484 The term @emph{applicable ^switches^switches^} reflects the fact that
14485 @command{gnatmake} ^switches^switches^ may or may not be passed to individual
14486 tools, depending on the individual ^switch^switch^.
14488 @command{gnatmake} may invoke the compiler on source files from different
14489 projects. The Project Manager will use the appropriate project file to
14490 determine the @code{Compiler} package for each source file being compiled.
14491 Likewise for the @code{Binder} and @code{Linker} packages.
14493 As an example, consider the following package in a project file:
14495 @smallexample @c projectfile
14498 package Compiler is
14499 for ^Default_Switches^Default_Switches^ ("Ada")
14501 for ^Switches^Switches^ ("a.adb")
14503 for ^Switches^Switches^ ("b.adb")
14505 "^-gnaty^-gnaty^");
14512 If @command{gnatmake} is invoked with this project file, and it needs to
14513 compile, say, the files @file{a.adb}, @file{b.adb}, and @file{c.adb}, then
14514 @file{a.adb} will be compiled with the ^switch^switch^
14515 @option{^-O1^-O1^},
14516 @file{b.adb} with ^switches^switches^
14518 and @option{^-gnaty^-gnaty^},
14519 and @file{c.adb} with @option{^-g^-g^}.
14521 The following example illustrates the ordering of the ^switches^switches^
14522 contributed by different packages:
14524 @smallexample @c projectfile
14528 for ^Switches^Switches^ ("main.adb")
14536 package Compiler is
14537 for ^Switches^Switches^ ("main.adb")
14545 If you issue the command:
14548 gnatmake ^-Pproj2^/PROJECT_FILE=PROJ2^ -O0 main
14552 then the compiler will be invoked on @file{main.adb} with the following
14553 sequence of ^switches^switches^
14556 ^-g -O1 -O2 -O0^-g -O1 -O2 -O0^
14559 with the last @option{^-O^-O^}
14560 ^switch^switch^ having precedence over the earlier ones;
14561 several other ^switches^switches^
14562 (such as @option{^-c^-c^}) are added implicitly.
14564 The ^switches^switches^
14566 and @option{^-O1^-O1^} are contributed by package
14567 @code{Builder}, @option{^-O2^-O2^} is contributed
14568 by the package @code{Compiler}
14569 and @option{^-O0^-O0^} comes from the command line.
14571 The @option{^-g^-g^}
14572 ^switch^switch^ will also be passed in the invocation of
14573 @command{Gnatlink.}
14575 A final example illustrates switch contributions from packages in different
14578 @smallexample @c projectfile
14581 for Source_Files use ("pack.ads", "pack.adb");
14582 package Compiler is
14583 for ^Default_Switches^Default_Switches^ ("Ada")
14584 use ("^-gnata^-gnata^");
14592 for Source_Files use ("foo_main.adb", "bar_main.adb");
14594 for ^Switches^Switches^ ("foo_main.adb")
14602 -- Ada source file:
14604 procedure Foo_Main is
14612 gnatmake ^-PProj4^/PROJECT_FILE=PROJ4^ foo_main.adb -cargs -gnato
14616 then the ^switches^switches^ passed to the compiler for @file{foo_main.adb} are
14617 @option{^-g^-g^} (contributed by the package @code{Proj4.Builder}) and
14618 @option{^-gnato^-gnato^} (passed on the command line).
14619 When the imported package @code{Pack} is compiled, the ^switches^switches^ used
14620 are @option{^-g^-g^} from @code{Proj4.Builder},
14621 @option{^-gnata^-gnata^} (contributed from package @code{Proj3.Compiler},
14622 and @option{^-gnato^-gnato^} from the command line.
14625 When using @command{gnatmake} with project files, some ^switches^switches^ or
14626 arguments may be expressed as relative paths. As the working directory where
14627 compilation occurs may change, these relative paths are converted to absolute
14628 paths. For the ^switches^switches^ found in a project file, the relative paths
14629 are relative to the project file directory, for the switches on the command
14630 line, they are relative to the directory where @command{gnatmake} is invoked.
14631 The ^switches^switches^ for which this occurs are:
14637 ^-aI^-aI^, as well as all arguments that are not switches (arguments to
14639 ^-o^-o^, object files specified in package @code{Linker} or after
14640 -largs on the command line). The exception to this rule is the ^switch^switch^
14641 ^--RTS=^--RTS=^ for which a relative path argument is never converted.
14643 @node Specifying Configuration Pragmas
14644 @subsubsection Specifying Configuration Pragmas
14646 When using @command{gnatmake} with project files, if there exists a file
14647 @file{gnat.adc} that contains configuration pragmas, this file will be
14650 Configuration pragmas can be defined by means of the following attributes in
14651 project files: @code{Global_Configuration_Pragmas} in package @code{Builder}
14652 and @code{Local_Configuration_Pragmas} in package @code{Compiler}.
14654 Both these attributes are single string attributes. Their values is the path
14655 name of a file containing configuration pragmas. If a path name is relative,
14656 then it is relative to the project directory of the project file where the
14657 attribute is defined.
14659 When compiling a source, the configuration pragmas used are, in order,
14660 those listed in the file designated by attribute
14661 @code{Global_Configuration_Pragmas} in package @code{Builder} of the main
14662 project file, if it is specified, and those listed in the file designated by
14663 attribute @code{Local_Configuration_Pragmas} in package @code{Compiler} of
14664 the project file of the source, if it exists.
14666 @node Project Files and Main Subprograms
14667 @subsubsection Project Files and Main Subprograms
14670 When using a project file, you can invoke @command{gnatmake}
14671 with one or several main subprograms, by specifying their source files on the
14675 gnatmake ^-P^/PROJECT_FILE=^prj main1 main2 main3
14679 Each of these needs to be a source file of the same project, except
14680 when the switch ^-u^/UNIQUE^ is used.
14683 When ^-u^/UNIQUE^ is not used, all the mains need to be sources of the
14684 same project, one of the project in the tree rooted at the project specified
14685 on the command line. The package @code{Builder} of this common project, the
14686 "main project" is the one that is considered by @command{gnatmake}.
14689 When ^-u^/UNIQUE^ is used, the specified source files may be in projects
14690 imported directly or indirectly by the project specified on the command line.
14691 Note that if such a source file is not part of the project specified on the
14692 command line, the ^switches^switches^ found in package @code{Builder} of the
14693 project specified on the command line, if any, that are transmitted
14694 to the compiler will still be used, not those found in the project file of
14698 When using a project file, you can also invoke @command{gnatmake} without
14699 explicitly specifying any main, and the effect depends on whether you have
14700 defined the @code{Main} attribute. This attribute has a string list value,
14701 where each element in the list is the name of a source file (the file
14702 extension is optional) that contains a unit that can be a main subprogram.
14704 If the @code{Main} attribute is defined in a project file as a non-empty
14705 string list and the switch @option{^-u^/UNIQUE^} is not used on the command
14706 line, then invoking @command{gnatmake} with this project file but without any
14707 main on the command line is equivalent to invoking @command{gnatmake} with all
14708 the file names in the @code{Main} attribute on the command line.
14711 @smallexample @c projectfile
14714 for Main use ("main1", "main2", "main3");
14720 With this project file, @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^"}
14722 @code{"gnatmake ^-Pprj^/PROJECT_FILE=PRJ^ main1 main2 main3"}.
14724 When the project attribute @code{Main} is not specified, or is specified
14725 as an empty string list, or when the switch @option{-u} is used on the command
14726 line, then invoking @command{gnatmake} with no main on the command line will
14727 result in all immediate sources of the project file being checked, and
14728 potentially recompiled. Depending on the presence of the switch @option{-u},
14729 sources from other project files on which the immediate sources of the main
14730 project file depend are also checked and potentially recompiled. In other
14731 words, the @option{-u} switch is applied to all of the immediate sources of the
14734 When no main is specified on the command line and attribute @code{Main} exists
14735 and includes several mains, or when several mains are specified on the
14736 command line, the default ^switches^switches^ in package @code{Builder} will
14737 be used for all mains, even if there are specific ^switches^switches^
14738 specified for one or several mains.
14740 But the ^switches^switches^ from package @code{Binder} or @code{Linker} will be
14741 the specific ^switches^switches^ for each main, if they are specified.
14743 @node Library Project Files
14744 @subsubsection Library Project Files
14747 When @command{gnatmake} is invoked with a main project file that is a library
14748 project file, it is not allowed to specify one or more mains on the command
14752 When a library project file is specified, switches ^-b^/ACTION=BIND^ and
14753 ^-l^/ACTION=LINK^ have special meanings.
14756 @item ^-b^/ACTION=BIND^ is only allowed for stand-alone libraries. It indicates
14757 to @command{gnatmake} that @command{gnatbind} should be invoked for the
14760 @item ^-l^/ACTION=LINK^ may be used for all library projects. It indicates
14761 to @command{gnatmake} that the binder generated file should be compiled
14762 (in the case of a stand-alone library) and that the library should be built.
14766 @node The GNAT Driver and Project Files
14767 @subsection The GNAT Driver and Project Files
14770 A number of GNAT tools, other than @command{^gnatmake^gnatmake^}
14771 can benefit from project files:
14772 @command{^gnatbind^gnatbind^},
14773 @command{^gnatcheck^gnatcheck^}),
14774 @command{^gnatclean^gnatclean^}),
14775 @command{^gnatelim^gnatelim^},
14776 @command{^gnatfind^gnatfind^},
14777 @command{^gnatlink^gnatlink^},
14778 @command{^gnatls^gnatls^},
14779 @command{^gnatmetric^gnatmetric^},
14780 @command{^gnatpp^gnatpp^},
14781 @command{^gnatstub^gnatstub^},
14782 and @command{^gnatxref^gnatxref^}. However, none of these tools can be invoked
14783 directly with a project file switch (@option{^-P^/PROJECT_FILE=^}).
14784 They must be invoked through the @command{gnat} driver.
14786 The @command{gnat} driver is a wrapper that accepts a number of commands and
14787 calls the corresponding tool. It was designed initially for VMS platforms (to
14788 convert VMS qualifiers to Unix-style switches), but it is now available on all
14791 On non-VMS platforms, the @command{gnat} driver accepts the following commands
14792 (case insensitive):
14796 BIND to invoke @command{^gnatbind^gnatbind^}
14798 CHOP to invoke @command{^gnatchop^gnatchop^}
14800 CLEAN to invoke @command{^gnatclean^gnatclean^}
14802 COMP or COMPILE to invoke the compiler
14804 ELIM to invoke @command{^gnatelim^gnatelim^}
14806 FIND to invoke @command{^gnatfind^gnatfind^}
14808 KR or KRUNCH to invoke @command{^gnatkr^gnatkr^}
14810 LINK to invoke @command{^gnatlink^gnatlink^}
14812 LS or LIST to invoke @command{^gnatls^gnatls^}
14814 MAKE to invoke @command{^gnatmake^gnatmake^}
14816 NAME to invoke @command{^gnatname^gnatname^}
14818 PREP or PREPROCESS to invoke @command{^gnatprep^gnatprep^}
14820 PP or PRETTY to invoke @command{^gnatpp^gnatpp^}
14822 METRIC to invoke @command{^gnatmetric^gnatmetric^}
14824 STUB to invoke @command{^gnatstub^gnatstub^}
14826 XREF to invoke @command{^gnatxref^gnatxref^}
14830 (note that the compiler is invoked using the command
14831 @command{^gnatmake -f -u -c^gnatmake -f -u -c^}).
14834 On non-VMS platforms, between @command{gnat} and the command, two
14835 special switches may be used:
14839 @command{-v} to display the invocation of the tool.
14841 @command{-dn} to prevent the @command{gnat} driver from removing
14842 the temporary files it has created. These temporary files are
14843 configuration files and temporary file list files.
14847 The command may be followed by switches and arguments for the invoked
14851 gnat bind -C main.ali
14857 Switches may also be put in text files, one switch per line, and the text
14858 files may be specified with their path name preceded by '@@'.
14861 gnat bind @@args.txt main.ali
14865 In addition, for commands BIND, COMP or COMPILE, FIND, ELIM, LS or LIST, LINK,
14866 METRIC, PP or PRETTY, STUB and XREF, the project file related switches
14867 (@option{^-P^/PROJECT_FILE^},
14868 @option{^-X^/EXTERNAL_REFERENCE^} and
14869 @option{^-vP^/MESSAGES_PROJECT_FILE=^x}) may be used in addition to
14870 the switches of the invoking tool.
14873 When GNAT PP or GNAT PRETTY is used with a project file, but with no source
14874 specified on the command line, it invokes @command{^gnatpp^gnatpp^} with all
14875 the immediate sources of the specified project file.
14878 When GNAT METRIC is used with a project file, but with no source
14879 specified on the command line, it invokes @command{^gnatmetric^gnatmetric^}
14880 with all the immediate sources of the specified project file and with
14881 @option{^-d^/DIRECTORY^} with the parameter pointing to the object directory
14885 In addition, when GNAT PP, GNAT PRETTY or GNAT METRIC is used with
14886 a project file, no source is specified on the command line and
14887 switch ^-U^/ALL_PROJECTS^ is specified on the command line, then
14888 the underlying tool (^gnatpp^gnatpp^ or
14889 ^gnatmetric^gnatmetric^) is invoked for all sources of all projects,
14890 not only for the immediate sources of the main project.
14892 (-U stands for Universal or Union of the project files of the project tree)
14896 For each of the following commands, there is optionally a corresponding
14897 package in the main project.
14901 package @code{Binder} for command BIND (invoking @code{^gnatbind^gnatbind^})
14904 package @code{Check} for command CHECK (invoking
14905 @code{^gnatcheck^gnatcheck^})
14908 package @code{Compiler} for command COMP or COMPILE (invoking the compiler)
14911 package @code{Cross_Reference} for command XREF (invoking
14912 @code{^gnatxref^gnatxref^})
14915 package @code{Eliminate} for command ELIM (invoking
14916 @code{^gnatelim^gnatelim^})
14919 package @code{Finder} for command FIND (invoking @code{^gnatfind^gnatfind^})
14922 package @code{Gnatls} for command LS or LIST (invoking @code{^gnatls^gnatls^})
14925 package @code{Gnatstub} for command STUB
14926 (invoking @code{^gnatstub^gnatstub^})
14929 package @code{Linker} for command LINK (invoking @code{^gnatlink^gnatlink^})
14932 package @code{Check} for command CHECK
14933 (invoking @code{^gnatcheck^gnatcheck^})
14936 package @code{Metrics} for command METRIC
14937 (invoking @code{^gnatmetric^gnatmetric^})
14940 package @code{Pretty_Printer} for command PP or PRETTY
14941 (invoking @code{^gnatpp^gnatpp^})
14946 Package @code{Gnatls} has a unique attribute @code{^Switches^Switches^},
14947 a simple variable with a string list value. It contains ^switches^switches^
14948 for the invocation of @code{^gnatls^gnatls^}.
14950 @smallexample @c projectfile
14954 for ^Switches^Switches^
14963 All other packages have two attribute @code{^Switches^Switches^} and
14964 @code{^Default_Switches^Default_Switches^}.
14967 @code{^Switches^Switches^} is an associative array attribute, indexed by the
14968 source file name, that has a string list value: the ^switches^switches^ to be
14969 used when the tool corresponding to the package is invoked for the specific
14973 @code{^Default_Switches^Default_Switches^} is an associative array attribute,
14974 indexed by the programming language that has a string list value.
14975 @code{^Default_Switches^Default_Switches^ ("Ada")} contains the
14976 ^switches^switches^ for the invocation of the tool corresponding
14977 to the package, except if a specific @code{^Switches^Switches^} attribute
14978 is specified for the source file.
14980 @smallexample @c projectfile
14984 for Source_Dirs use ("./**");
14987 for ^Switches^Switches^ use
14994 package Compiler is
14995 for ^Default_Switches^Default_Switches^ ("Ada")
14996 use ("^-gnatv^-gnatv^",
14997 "^-gnatwa^-gnatwa^");
15003 for ^Default_Switches^Default_Switches^ ("Ada")
15011 for ^Default_Switches^Default_Switches^ ("Ada")
15013 for ^Switches^Switches^ ("main.adb")
15022 for ^Default_Switches^Default_Switches^ ("Ada")
15029 package Cross_Reference is
15030 for ^Default_Switches^Default_Switches^ ("Ada")
15035 end Cross_Reference;
15041 With the above project file, commands such as
15044 ^gnat comp -Pproj main^GNAT COMP /PROJECT_FILE=PROJ MAIN^
15045 ^gnat ls -Pproj main^GNAT LIST /PROJECT_FILE=PROJ MAIN^
15046 ^gnat xref -Pproj main^GNAT XREF /PROJECT_FILE=PROJ MAIN^
15047 ^gnat bind -Pproj main.ali^GNAT BIND /PROJECT_FILE=PROJ MAIN.ALI^
15048 ^gnat link -Pproj main.ali^GNAT LINK /PROJECT_FILE=PROJ MAIN.ALI^
15052 will set up the environment properly and invoke the tool with the switches
15053 found in the package corresponding to the tool:
15054 @code{^Default_Switches^Default_Switches^ ("Ada")} for all tools,
15055 except @code{^Switches^Switches^ ("main.adb")}
15056 for @code{^gnatlink^gnatlink^}.
15057 It is also possible to invoke some of the tools,
15058 @code{^gnatcheck^gnatcheck^}),
15059 @code{^gnatmetric^gnatmetric^}),
15060 and @code{^gnatpp^gnatpp^})
15061 on a set of project units thanks to the combination of the switches
15062 @option{-P}, @option{-U} and possibly the main unit when one is interested
15063 in its closure. For instance,
15067 will compute the metrics for all the immediate units of project
15070 gnat metric -Pproj -U
15072 will compute the metrics for all the units of the closure of projects
15073 rooted at @code{proj}.
15075 gnat metric -Pproj -U main_unit
15077 will compute the metrics for the closure of units rooted at
15078 @code{main_unit}. This last possibility relies implicitly
15079 on @command{gnatbind}'s option @option{-R}.
15081 @c **********************
15082 @node An Extended Example
15083 @section An Extended Example
15086 Suppose that we have two programs, @var{prog1} and @var{prog2},
15087 whose sources are in corresponding directories. We would like
15088 to build them with a single @command{gnatmake} command, and we want to place
15089 their object files into @file{build} subdirectories of the source directories.
15090 Furthermore, we want to have to have two separate subdirectories
15091 in @file{build} -- @file{release} and @file{debug} -- which will contain
15092 the object files compiled with different set of compilation flags.
15094 In other words, we have the following structure:
15111 Here are the project files that we must place in a directory @file{main}
15112 to maintain this structure:
15116 @item We create a @code{Common} project with a package @code{Compiler} that
15117 specifies the compilation ^switches^switches^:
15122 @b{project} Common @b{is}
15124 @b{for} Source_Dirs @b{use} (); -- No source files
15128 @b{type} Build_Type @b{is} ("release", "debug");
15129 Build : Build_Type := External ("BUILD", "debug");
15132 @b{package} Compiler @b{is}
15133 @b{case} Build @b{is}
15134 @b{when} "release" =>
15135 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15136 @b{use} ("^-O2^-O2^");
15137 @b{when} "debug" =>
15138 @b{for} ^Default_Switches^Default_Switches^ ("Ada")
15139 @b{use} ("^-g^-g^");
15147 @item We create separate projects for the two programs:
15154 @b{project} Prog1 @b{is}
15156 @b{for} Source_Dirs @b{use} ("prog1");
15157 @b{for} Object_Dir @b{use} "prog1/build/" & Common.Build;
15159 @b{package} Compiler @b{renames} Common.Compiler;
15170 @b{project} Prog2 @b{is}
15172 @b{for} Source_Dirs @b{use} ("prog2");
15173 @b{for} Object_Dir @b{use} "prog2/build/" & Common.Build;
15175 @b{package} Compiler @b{renames} Common.Compiler;
15181 @item We create a wrapping project @code{Main}:
15190 @b{project} Main @b{is}
15192 @b{package} Compiler @b{renames} Common.Compiler;
15198 @item Finally we need to create a dummy procedure that @code{with}s (either
15199 explicitly or implicitly) all the sources of our two programs.
15204 Now we can build the programs using the command
15207 gnatmake ^-P^/PROJECT_FILE=^main dummy
15211 for the Debug mode, or
15215 gnatmake -Pmain -XBUILD=release
15221 GNAT MAKE /PROJECT_FILE=main /EXTERNAL_REFERENCE=BUILD=release
15226 for the Release mode.
15228 @c ********************************
15229 @c * Project File Complete Syntax *
15230 @c ********************************
15232 @node Project File Complete Syntax
15233 @section Project File Complete Syntax
15237 context_clause project_declaration
15243 @b{with} path_name @{ , path_name @} ;
15248 project_declaration ::=
15249 simple_project_declaration | project_extension
15251 simple_project_declaration ::=
15252 @b{project} <project_>simple_name @b{is}
15253 @{declarative_item@}
15254 @b{end} <project_>simple_name;
15256 project_extension ::=
15257 @b{project} <project_>simple_name @b{extends} path_name @b{is}
15258 @{declarative_item@}
15259 @b{end} <project_>simple_name;
15261 declarative_item ::=
15262 package_declaration |
15263 typed_string_declaration |
15264 other_declarative_item
15266 package_declaration ::=
15267 package_spec | package_renaming
15270 @b{package} package_identifier @b{is}
15271 @{simple_declarative_item@}
15272 @b{end} package_identifier ;
15274 package_identifier ::=
15275 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
15276 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
15277 @code{^gnatls^gnatls^} | @code{IDE} | @code{Pretty_Printer}
15279 package_renaming ::==
15280 @b{package} package_identifier @b{renames}
15281 <project_>simple_name.package_identifier ;
15283 typed_string_declaration ::=
15284 @b{type} <typed_string_>_simple_name @b{is}
15285 ( string_literal @{, string_literal@} );
15287 other_declarative_item ::=
15288 attribute_declaration |
15289 typed_variable_declaration |
15290 variable_declaration |
15293 attribute_declaration ::=
15294 full_associative_array_declaration |
15295 @b{for} attribute_designator @b{use} expression ;
15297 full_associative_array_declaration ::=
15298 @b{for} <associative_array_attribute_>simple_name @b{use}
15299 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
15301 attribute_designator ::=
15302 <simple_attribute_>simple_name |
15303 <associative_array_attribute_>simple_name ( string_literal )
15305 typed_variable_declaration ::=
15306 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
15308 variable_declaration ::=
15309 <variable_>simple_name := expression;
15319 attribute_reference
15325 ( <string_>expression @{ , <string_>expression @} )
15328 @b{external} ( string_literal [, string_literal] )
15330 attribute_reference ::=
15331 attribute_prefix ' <simple_attribute_>simple_name [ ( literal_string ) ]
15333 attribute_prefix ::=
15335 <project_>simple_name | package_identifier |
15336 <project_>simple_name . package_identifier
15338 case_construction ::=
15339 @b{case} <typed_variable_>name @b{is}
15344 @b{when} discrete_choice_list =>
15345 @{case_construction | attribute_declaration@}
15347 discrete_choice_list ::=
15348 string_literal @{| string_literal@} |
15352 simple_name @{. simple_name@}
15355 identifier (same as Ada)
15359 @node The Cross-Referencing Tools gnatxref and gnatfind
15360 @chapter The Cross-Referencing Tools @code{gnatxref} and @code{gnatfind}
15365 The compiler generates cross-referencing information (unless
15366 you set the @samp{-gnatx} switch), which are saved in the @file{.ali} files.
15367 This information indicates where in the source each entity is declared and
15368 referenced. Note that entities in package Standard are not included, but
15369 entities in all other predefined units are included in the output.
15371 Before using any of these two tools, you need to compile successfully your
15372 application, so that GNAT gets a chance to generate the cross-referencing
15375 The two tools @code{gnatxref} and @code{gnatfind} take advantage of this
15376 information to provide the user with the capability to easily locate the
15377 declaration and references to an entity. These tools are quite similar,
15378 the difference being that @code{gnatfind} is intended for locating
15379 definitions and/or references to a specified entity or entities, whereas
15380 @code{gnatxref} is oriented to generating a full report of all
15383 To use these tools, you must not compile your application using the
15384 @option{-gnatx} switch on the @command{gnatmake} command line
15385 (@pxref{The GNAT Make Program gnatmake}). Otherwise, cross-referencing
15386 information will not be generated.
15388 Note: to invoke @code{gnatxref} or @code{gnatfind} with a project file,
15389 use the @code{gnat} driver (see @ref{The GNAT Driver and Project Files}).
15392 * Switches for gnatxref::
15393 * Switches for gnatfind::
15394 * Project Files for gnatxref and gnatfind::
15395 * Regular Expressions in gnatfind and gnatxref::
15396 * Examples of gnatxref Usage::
15397 * Examples of gnatfind Usage::
15400 @node Switches for gnatxref
15401 @section @code{gnatxref} Switches
15404 The command invocation for @code{gnatxref} is:
15406 $ gnatxref @ovar{switches} @var{sourcefile1} @r{[}@var{sourcefile2} @dots{}@r{]}
15415 identifies the source files for which a report is to be generated. The
15416 ``with''ed units will be processed too. You must provide at least one file.
15418 These file names are considered to be regular expressions, so for instance
15419 specifying @file{source*.adb} is the same as giving every file in the current
15420 directory whose name starts with @file{source} and whose extension is
15423 You shouldn't specify any directory name, just base names. @command{gnatxref}
15424 and @command{gnatfind} will be able to locate these files by themselves using
15425 the source path. If you specify directories, no result is produced.
15430 The switches can be:
15434 @cindex @option{--version} @command{gnatxref}
15435 Display Copyright and version, then exit disregarding all other options.
15438 @cindex @option{--help} @command{gnatxref}
15439 If @option{--version} was not used, display usage, then exit disregarding
15442 @item ^-a^/ALL_FILES^
15443 @cindex @option{^-a^/ALL_FILES^} (@command{gnatxref})
15444 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15445 the read-only files found in the library search path. Otherwise, these files
15446 will be ignored. This option can be used to protect Gnat sources or your own
15447 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15448 much faster, and their output much smaller. Read-only here refers to access
15449 or permissions status in the file system for the current user.
15452 @cindex @option{-aIDIR} (@command{gnatxref})
15453 When looking for source files also look in directory DIR. The order in which
15454 source file search is undertaken is the same as for @command{gnatmake}.
15457 @cindex @option{-aODIR} (@command{gnatxref})
15458 When searching for library and object files, look in directory
15459 DIR. The order in which library files are searched is the same as for
15460 @command{gnatmake}.
15463 @cindex @option{-nostdinc} (@command{gnatxref})
15464 Do not look for sources in the system default directory.
15467 @cindex @option{-nostdlib} (@command{gnatxref})
15468 Do not look for library files in the system default directory.
15470 @item --RTS=@var{rts-path}
15471 @cindex @option{--RTS} (@command{gnatxref})
15472 Specifies the default location of the runtime library. Same meaning as the
15473 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15475 @item ^-d^/DERIVED_TYPES^
15476 @cindex @option{^-d^/DERIVED_TYPES^} (@command{gnatxref})
15477 If this switch is set @code{gnatxref} will output the parent type
15478 reference for each matching derived types.
15480 @item ^-f^/FULL_PATHNAME^
15481 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatxref})
15482 If this switch is set, the output file names will be preceded by their
15483 directory (if the file was found in the search path). If this switch is
15484 not set, the directory will not be printed.
15486 @item ^-g^/IGNORE_LOCALS^
15487 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatxref})
15488 If this switch is set, information is output only for library-level
15489 entities, ignoring local entities. The use of this switch may accelerate
15490 @code{gnatfind} and @code{gnatxref}.
15493 @cindex @option{-IDIR} (@command{gnatxref})
15494 Equivalent to @samp{-aODIR -aIDIR}.
15497 @cindex @option{-pFILE} (@command{gnatxref})
15498 Specify a project file to use @xref{Project Files}.
15499 If you need to use the @file{.gpr}
15500 project files, you should use gnatxref through the GNAT driver
15501 (@command{gnat xref -Pproject}).
15503 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15504 project file in the current directory.
15506 If a project file is either specified or found by the tools, then the content
15507 of the source directory and object directory lines are added as if they
15508 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^}
15509 and @samp{^-aO^OBJECT_SEARCH^}.
15511 Output only unused symbols. This may be really useful if you give your
15512 main compilation unit on the command line, as @code{gnatxref} will then
15513 display every unused entity and 'with'ed package.
15517 Instead of producing the default output, @code{gnatxref} will generate a
15518 @file{tags} file that can be used by vi. For examples how to use this
15519 feature, see @ref{Examples of gnatxref Usage}. The tags file is output
15520 to the standard output, thus you will have to redirect it to a file.
15526 All these switches may be in any order on the command line, and may even
15527 appear after the file names. They need not be separated by spaces, thus
15528 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15529 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15531 @node Switches for gnatfind
15532 @section @code{gnatfind} Switches
15535 The command line for @code{gnatfind} is:
15538 $ gnatfind @ovar{switches} @var{pattern}@r{[}:@var{sourcefile}@r{[}:@var{line}@r{[}:@var{column}@r{]]]}
15539 @r{[}@var{file1} @var{file2} @dots{}]
15547 An entity will be output only if it matches the regular expression found
15548 in @var{pattern}, see @ref{Regular Expressions in gnatfind and gnatxref}.
15550 Omitting the pattern is equivalent to specifying @samp{*}, which
15551 will match any entity. Note that if you do not provide a pattern, you
15552 have to provide both a sourcefile and a line.
15554 Entity names are given in Latin-1, with uppercase/lowercase equivalence
15555 for matching purposes. At the current time there is no support for
15556 8-bit codes other than Latin-1, or for wide characters in identifiers.
15559 @code{gnatfind} will look for references, bodies or declarations
15560 of symbols referenced in @file{@var{sourcefile}}, at line @var{line}
15561 and column @var{column}. See @ref{Examples of gnatfind Usage}
15562 for syntax examples.
15565 is a decimal integer identifying the line number containing
15566 the reference to the entity (or entities) to be located.
15569 is a decimal integer identifying the exact location on the
15570 line of the first character of the identifier for the
15571 entity reference. Columns are numbered from 1.
15573 @item file1 file2 @dots{}
15574 The search will be restricted to these source files. If none are given, then
15575 the search will be done for every library file in the search path.
15576 These file must appear only after the pattern or sourcefile.
15578 These file names are considered to be regular expressions, so for instance
15579 specifying @file{source*.adb} is the same as giving every file in the current
15580 directory whose name starts with @file{source} and whose extension is
15583 The location of the spec of the entity will always be displayed, even if it
15584 isn't in one of @file{@var{file1}}, @file{@var{file2}},@enddots{} The
15585 occurrences of the entity in the separate units of the ones given on the
15586 command line will also be displayed.
15588 Note that if you specify at least one file in this part, @code{gnatfind} may
15589 sometimes not be able to find the body of the subprograms.
15594 At least one of 'sourcefile' or 'pattern' has to be present on
15597 The following switches are available:
15601 @cindex @option{--version} @command{gnatfind}
15602 Display Copyright and version, then exit disregarding all other options.
15605 @cindex @option{--help} @command{gnatfind}
15606 If @option{--version} was not used, display usage, then exit disregarding
15609 @item ^-a^/ALL_FILES^
15610 @cindex @option{^-a^/ALL_FILES^} (@command{gnatfind})
15611 If this switch is present, @code{gnatfind} and @code{gnatxref} will parse
15612 the read-only files found in the library search path. Otherwise, these files
15613 will be ignored. This option can be used to protect Gnat sources or your own
15614 libraries from being parsed, thus making @code{gnatfind} and @code{gnatxref}
15615 much faster, and their output much smaller. Read-only here refers to access
15616 or permission status in the file system for the current user.
15619 @cindex @option{-aIDIR} (@command{gnatfind})
15620 When looking for source files also look in directory DIR. The order in which
15621 source file search is undertaken is the same as for @command{gnatmake}.
15624 @cindex @option{-aODIR} (@command{gnatfind})
15625 When searching for library and object files, look in directory
15626 DIR. The order in which library files are searched is the same as for
15627 @command{gnatmake}.
15630 @cindex @option{-nostdinc} (@command{gnatfind})
15631 Do not look for sources in the system default directory.
15634 @cindex @option{-nostdlib} (@command{gnatfind})
15635 Do not look for library files in the system default directory.
15637 @item --ext=@var{extension}
15638 @cindex @option{--ext} (@command{gnatfind})
15639 Specify an alternate ali file extension. The default is @code{ali} and other
15640 extensions (e.g. @code{sli} for SPARK library files) may be specified via this
15641 switch. Note that if this switch overrides the default, which means that only
15642 the new extension will be considered.
15644 @item --RTS=@var{rts-path}
15645 @cindex @option{--RTS} (@command{gnatfind})
15646 Specifies the default location of the runtime library. Same meaning as the
15647 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
15649 @item ^-d^/DERIVED_TYPE_INFORMATION^
15650 @cindex @option{^-d^/DERIVED_TYPE_INFORMATION^} (@code{gnatfind})
15651 If this switch is set, then @code{gnatfind} will output the parent type
15652 reference for each matching derived types.
15654 @item ^-e^/EXPRESSIONS^
15655 @cindex @option{^-e^/EXPRESSIONS^} (@command{gnatfind})
15656 By default, @code{gnatfind} accept the simple regular expression set for
15657 @samp{pattern}. If this switch is set, then the pattern will be
15658 considered as full Unix-style regular expression.
15660 @item ^-f^/FULL_PATHNAME^
15661 @cindex @option{^-f^/FULL_PATHNAME^} (@command{gnatfind})
15662 If this switch is set, the output file names will be preceded by their
15663 directory (if the file was found in the search path). If this switch is
15664 not set, the directory will not be printed.
15666 @item ^-g^/IGNORE_LOCALS^
15667 @cindex @option{^-g^/IGNORE_LOCALS^} (@command{gnatfind})
15668 If this switch is set, information is output only for library-level
15669 entities, ignoring local entities. The use of this switch may accelerate
15670 @code{gnatfind} and @code{gnatxref}.
15673 @cindex @option{-IDIR} (@command{gnatfind})
15674 Equivalent to @samp{-aODIR -aIDIR}.
15677 @cindex @option{-pFILE} (@command{gnatfind})
15678 Specify a project file (@pxref{Project Files}) to use.
15679 By default, @code{gnatxref} and @code{gnatfind} will try to locate a
15680 project file in the current directory.
15682 If a project file is either specified or found by the tools, then the content
15683 of the source directory and object directory lines are added as if they
15684 had been specified respectively by @samp{^-aI^/SOURCE_SEARCH^} and
15685 @samp{^-aO^/OBJECT_SEARCH^}.
15687 @item ^-r^/REFERENCES^
15688 @cindex @option{^-r^/REFERENCES^} (@command{gnatfind})
15689 By default, @code{gnatfind} will output only the information about the
15690 declaration, body or type completion of the entities. If this switch is
15691 set, the @code{gnatfind} will locate every reference to the entities in
15692 the files specified on the command line (or in every file in the search
15693 path if no file is given on the command line).
15695 @item ^-s^/PRINT_LINES^
15696 @cindex @option{^-s^/PRINT_LINES^} (@command{gnatfind})
15697 If this switch is set, then @code{gnatfind} will output the content
15698 of the Ada source file lines were the entity was found.
15700 @item ^-t^/TYPE_HIERARCHY^
15701 @cindex @option{^-t^/TYPE_HIERARCHY^} (@command{gnatfind})
15702 If this switch is set, then @code{gnatfind} will output the type hierarchy for
15703 the specified type. It act like -d option but recursively from parent
15704 type to parent type. When this switch is set it is not possible to
15705 specify more than one file.
15710 All these switches may be in any order on the command line, and may even
15711 appear after the file names. They need not be separated by spaces, thus
15712 you can say @samp{gnatxref ^-ag^/ALL_FILES/IGNORE_LOCALS^} instead of
15713 @samp{gnatxref ^-a -g^/ALL_FILES /IGNORE_LOCALS^}.
15715 As stated previously, gnatfind will search in every directory in the
15716 search path. You can force it to look only in the current directory if
15717 you specify @code{*} at the end of the command line.
15719 @node Project Files for gnatxref and gnatfind
15720 @section Project Files for @command{gnatxref} and @command{gnatfind}
15723 Project files allow a programmer to specify how to compile its
15724 application, where to find sources, etc. These files are used
15726 primarily by GPS, but they can also be used
15729 @code{gnatxref} and @code{gnatfind}.
15731 A project file name must end with @file{.gpr}. If a single one is
15732 present in the current directory, then @code{gnatxref} and @code{gnatfind} will
15733 extract the information from it. If multiple project files are found, none of
15734 them is read, and you have to use the @samp{-p} switch to specify the one
15737 The following lines can be included, even though most of them have default
15738 values which can be used in most cases.
15739 The lines can be entered in any order in the file.
15740 Except for @file{src_dir} and @file{obj_dir}, you can only have one instance of
15741 each line. If you have multiple instances, only the last one is taken into
15746 [default: @code{"^./^[]^"}]
15747 specifies a directory where to look for source files. Multiple @code{src_dir}
15748 lines can be specified and they will be searched in the order they
15752 [default: @code{"^./^[]^"}]
15753 specifies a directory where to look for object and library files. Multiple
15754 @code{obj_dir} lines can be specified, and they will be searched in the order
15757 @item comp_opt=SWITCHES
15758 [default: @code{""}]
15759 creates a variable which can be referred to subsequently by using
15760 the @code{$@{comp_opt@}} notation. This is intended to store the default
15761 switches given to @command{gnatmake} and @command{gcc}.
15763 @item bind_opt=SWITCHES
15764 [default: @code{""}]
15765 creates a variable which can be referred to subsequently by using
15766 the @samp{$@{bind_opt@}} notation. This is intended to store the default
15767 switches given to @command{gnatbind}.
15769 @item link_opt=SWITCHES
15770 [default: @code{""}]
15771 creates a variable which can be referred to subsequently by using
15772 the @samp{$@{link_opt@}} notation. This is intended to store the default
15773 switches given to @command{gnatlink}.
15775 @item main=EXECUTABLE
15776 [default: @code{""}]
15777 specifies the name of the executable for the application. This variable can
15778 be referred to in the following lines by using the @samp{$@{main@}} notation.
15781 @item comp_cmd=COMMAND
15782 [default: @code{"GNAT COMPILE /SEARCH=$@{src_dir@} /DEBUG /TRY_SEMANTICS"}]
15785 @item comp_cmd=COMMAND
15786 [default: @code{"gcc -c -I$@{src_dir@} -g -gnatq"}]
15788 specifies the command used to compile a single file in the application.
15791 @item make_cmd=COMMAND
15792 [default: @code{"GNAT MAKE $@{main@}
15793 /SOURCE_SEARCH=$@{src_dir@} /OBJECT_SEARCH=$@{obj_dir@}
15794 /DEBUG /TRY_SEMANTICS /COMPILER_QUALIFIERS $@{comp_opt@}
15795 /BINDER_QUALIFIERS $@{bind_opt@} /LINKER_QUALIFIERS $@{link_opt@}"}]
15798 @item make_cmd=COMMAND
15799 [default: @code{"gnatmake $@{main@} -aI$@{src_dir@}
15800 -aO$@{obj_dir@} -g -gnatq -cargs $@{comp_opt@}
15801 -bargs $@{bind_opt@} -largs $@{link_opt@}"}]
15803 specifies the command used to recompile the whole application.
15805 @item run_cmd=COMMAND
15806 [default: @code{"$@{main@}"}]
15807 specifies the command used to run the application.
15809 @item debug_cmd=COMMAND
15810 [default: @code{"gdb $@{main@}"}]
15811 specifies the command used to debug the application
15816 @command{gnatxref} and @command{gnatfind} only take into account the
15817 @code{src_dir} and @code{obj_dir} lines, and ignore the others.
15819 @node Regular Expressions in gnatfind and gnatxref
15820 @section Regular Expressions in @code{gnatfind} and @code{gnatxref}
15823 As specified in the section about @command{gnatfind}, the pattern can be a
15824 regular expression. Actually, there are to set of regular expressions
15825 which are recognized by the program:
15828 @item globbing patterns
15829 These are the most usual regular expression. They are the same that you
15830 generally used in a Unix shell command line, or in a DOS session.
15832 Here is a more formal grammar:
15839 term ::= elmt -- matches elmt
15840 term ::= elmt elmt -- concatenation (elmt then elmt)
15841 term ::= * -- any string of 0 or more characters
15842 term ::= ? -- matches any character
15843 term ::= [char @{char@}] -- matches any character listed
15844 term ::= [char - char] -- matches any character in range
15848 @item full regular expression
15849 The second set of regular expressions is much more powerful. This is the
15850 type of regular expressions recognized by utilities such a @file{grep}.
15852 The following is the form of a regular expression, expressed in Ada
15853 reference manual style BNF is as follows
15860 regexp ::= term @{| term@} -- alternation (term or term @dots{})
15862 term ::= item @{item@} -- concatenation (item then item)
15864 item ::= elmt -- match elmt
15865 item ::= elmt * -- zero or more elmt's
15866 item ::= elmt + -- one or more elmt's
15867 item ::= elmt ? -- matches elmt or nothing
15870 elmt ::= nschar -- matches given character
15871 elmt ::= [nschar @{nschar@}] -- matches any character listed
15872 elmt ::= [^^^ nschar @{nschar@}] -- matches any character not listed
15873 elmt ::= [char - char] -- matches chars in given range
15874 elmt ::= \ char -- matches given character
15875 elmt ::= . -- matches any single character
15876 elmt ::= ( regexp ) -- parens used for grouping
15878 char ::= any character, including special characters
15879 nschar ::= any character except ()[].*+?^^^
15883 Following are a few examples:
15887 will match any of the two strings @samp{abcde} and @samp{fghi},
15890 will match any string like @samp{abd}, @samp{abcd}, @samp{abccd},
15891 @samp{abcccd}, and so on,
15894 will match any string which has only lowercase characters in it (and at
15895 least one character.
15900 @node Examples of gnatxref Usage
15901 @section Examples of @code{gnatxref} Usage
15903 @subsection General Usage
15906 For the following examples, we will consider the following units:
15908 @smallexample @c ada
15914 3: procedure Foo (B : in Integer);
15921 1: package body Main is
15922 2: procedure Foo (B : in Integer) is
15933 2: procedure Print (B : Integer);
15942 The first thing to do is to recompile your application (for instance, in
15943 that case just by doing a @samp{gnatmake main}, so that GNAT generates
15944 the cross-referencing information.
15945 You can then issue any of the following commands:
15947 @item gnatxref main.adb
15948 @code{gnatxref} generates cross-reference information for main.adb
15949 and every unit 'with'ed by main.adb.
15951 The output would be:
15959 Decl: main.ads 3:20
15960 Body: main.adb 2:20
15961 Ref: main.adb 4:13 5:13 6:19
15964 Ref: main.adb 6:8 7:8
15974 Decl: main.ads 3:15
15975 Body: main.adb 2:15
15978 Body: main.adb 1:14
15981 Ref: main.adb 6:12 7:12
15985 that is the entity @code{Main} is declared in main.ads, line 2, column 9,
15986 its body is in main.adb, line 1, column 14 and is not referenced any where.
15988 The entity @code{Print} is declared in bar.ads, line 2, column 15 and it
15989 it referenced in main.adb, line 6 column 12 and line 7 column 12.
15991 @item gnatxref package1.adb package2.ads
15992 @code{gnatxref} will generates cross-reference information for
15993 package1.adb, package2.ads and any other package 'with'ed by any
15999 @subsection Using gnatxref with vi
16001 @code{gnatxref} can generate a tags file output, which can be used
16002 directly from @command{vi}. Note that the standard version of @command{vi}
16003 will not work properly with overloaded symbols. Consider using another
16004 free implementation of @command{vi}, such as @command{vim}.
16007 $ gnatxref -v gnatfind.adb > tags
16011 will generate the tags file for @code{gnatfind} itself (if the sources
16012 are in the search path!).
16014 From @command{vi}, you can then use the command @samp{:tag @var{entity}}
16015 (replacing @var{entity} by whatever you are looking for), and vi will
16016 display a new file with the corresponding declaration of entity.
16019 @node Examples of gnatfind Usage
16020 @section Examples of @code{gnatfind} Usage
16024 @item gnatfind ^-f^/FULL_PATHNAME^ xyz:main.adb
16025 Find declarations for all entities xyz referenced at least once in
16026 main.adb. The references are search in every library file in the search
16029 The directories will be printed as well (as the @samp{^-f^/FULL_PATHNAME^}
16032 The output will look like:
16034 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
16035 ^directory/^[directory]^main.adb:24:10: xyz <= body
16036 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
16040 that is to say, one of the entities xyz found in main.adb is declared at
16041 line 12 of main.ads (and its body is in main.adb), and another one is
16042 declared at line 45 of foo.ads
16044 @item gnatfind ^-fs^/FULL_PATHNAME/SOURCE_LINE^ xyz:main.adb
16045 This is the same command as the previous one, instead @code{gnatfind} will
16046 display the content of the Ada source file lines.
16048 The output will look like:
16051 ^directory/^[directory]^main.ads:106:14: xyz <= declaration
16053 ^directory/^[directory]^main.adb:24:10: xyz <= body
16055 ^directory/^[directory]^foo.ads:45:23: xyz <= declaration
16060 This can make it easier to find exactly the location your are looking
16063 @item gnatfind ^-r^/REFERENCES^ "*x*":main.ads:123 foo.adb
16064 Find references to all entities containing an x that are
16065 referenced on line 123 of main.ads.
16066 The references will be searched only in main.ads and foo.adb.
16068 @item gnatfind main.ads:123
16069 Find declarations and bodies for all entities that are referenced on
16070 line 123 of main.ads.
16072 This is the same as @code{gnatfind "*":main.adb:123}.
16074 @item gnatfind ^mydir/^[mydir]^main.adb:123:45
16075 Find the declaration for the entity referenced at column 45 in
16076 line 123 of file main.adb in directory mydir. Note that it
16077 is usual to omit the identifier name when the column is given,
16078 since the column position identifies a unique reference.
16080 The column has to be the beginning of the identifier, and should not
16081 point to any character in the middle of the identifier.
16085 @c *********************************
16086 @node The GNAT Pretty-Printer gnatpp
16087 @chapter The GNAT Pretty-Printer @command{gnatpp}
16089 @cindex Pretty-Printer
16092 ^The @command{gnatpp} tool^GNAT PRETTY^ is an ASIS-based utility
16093 for source reformatting / pretty-printing.
16094 It takes an Ada source file as input and generates a reformatted
16096 You can specify various style directives via switches; e.g.,
16097 identifier case conventions, rules of indentation, and comment layout.
16099 To produce a reformatted file, @command{gnatpp} generates and uses the ASIS
16100 tree for the input source and thus requires the input to be syntactically and
16101 semantically legal.
16102 If this condition is not met, @command{gnatpp} will terminate with an
16103 error message; no output file will be generated.
16105 If the source files presented to @command{gnatpp} contain
16106 preprocessing directives, then the output file will
16107 correspond to the generated source after all
16108 preprocessing is carried out. There is no way
16109 using @command{gnatpp} to obtain pretty printed files that
16110 include the preprocessing directives.
16112 If the compilation unit
16113 contained in the input source depends semantically upon units located
16114 outside the current directory, you have to provide the source search path
16115 when invoking @command{gnatpp}, if these units are contained in files with
16116 names that do not follow the GNAT file naming rules, you have to provide
16117 the configuration file describing the corresponding naming scheme;
16118 see the description of the @command{gnatpp}
16119 switches below. Another possibility is to use a project file and to
16120 call @command{gnatpp} through the @command{gnat} driver
16122 The @command{gnatpp} command has the form
16125 $ gnatpp @ovar{switches} @var{filename}
16132 @var{switches} is an optional sequence of switches defining such properties as
16133 the formatting rules, the source search path, and the destination for the
16137 @var{filename} is the name (including the extension) of the source file to
16138 reformat; ``wildcards'' or several file names on the same gnatpp command are
16139 allowed. The file name may contain path information; it does not have to
16140 follow the GNAT file naming rules
16144 * Switches for gnatpp::
16145 * Formatting Rules::
16148 @node Switches for gnatpp
16149 @section Switches for @command{gnatpp}
16152 The following subsections describe the various switches accepted by
16153 @command{gnatpp}, organized by category.
16156 You specify a switch by supplying a name and generally also a value.
16157 In many cases the values for a switch with a given name are incompatible with
16159 (for example the switch that controls the casing of a reserved word may have
16160 exactly one value: upper case, lower case, or
16161 mixed case) and thus exactly one such switch can be in effect for an
16162 invocation of @command{gnatpp}.
16163 If more than one is supplied, the last one is used.
16164 However, some values for the same switch are mutually compatible.
16165 You may supply several such switches to @command{gnatpp}, but then
16166 each must be specified in full, with both the name and the value.
16167 Abbreviated forms (the name appearing once, followed by each value) are
16169 For example, to set
16170 the alignment of the assignment delimiter both in declarations and in
16171 assignment statements, you must write @option{-A2A3}
16172 (or @option{-A2 -A3}), but not @option{-A23}.
16176 In many cases the set of options for a given qualifier are incompatible with
16177 each other (for example the qualifier that controls the casing of a reserved
16178 word may have exactly one option, which specifies either upper case, lower
16179 case, or mixed case), and thus exactly one such option can be in effect for
16180 an invocation of @command{gnatpp}.
16181 If more than one is supplied, the last one is used.
16182 However, some qualifiers have options that are mutually compatible,
16183 and then you may then supply several such options when invoking
16187 In most cases, it is obvious whether or not the
16188 ^values for a switch with a given name^options for a given qualifier^
16189 are compatible with each other.
16190 When the semantics might not be evident, the summaries below explicitly
16191 indicate the effect.
16194 * Alignment Control::
16196 * Construct Layout Control::
16197 * General Text Layout Control::
16198 * Other Formatting Options::
16199 * Setting the Source Search Path::
16200 * Output File Control::
16201 * Other gnatpp Switches::
16204 @node Alignment Control
16205 @subsection Alignment Control
16206 @cindex Alignment control in @command{gnatpp}
16209 Programs can be easier to read if certain constructs are vertically aligned.
16210 By default all alignments are set ON.
16211 Through the @option{^-A0^/ALIGN=OFF^} switch you may reset the default to
16212 OFF, and then use one or more of the other
16213 ^@option{-A@var{n}} switches^@option{/ALIGN} options^
16214 to activate alignment for specific constructs.
16217 @cindex @option{^-A@var{n}^/ALIGN^} (@command{gnatpp})
16221 Set all alignments to ON
16224 @item ^-A0^/ALIGN=OFF^
16225 Set all alignments to OFF
16227 @item ^-A1^/ALIGN=COLONS^
16228 Align @code{:} in declarations
16230 @item ^-A2^/ALIGN=DECLARATIONS^
16231 Align @code{:=} in initializations in declarations
16233 @item ^-A3^/ALIGN=STATEMENTS^
16234 Align @code{:=} in assignment statements
16236 @item ^-A4^/ALIGN=ARROWS^
16237 Align @code{=>} in associations
16239 @item ^-A5^/ALIGN=COMPONENT_CLAUSES^
16240 Align @code{at} keywords in the component clauses in record
16241 representation clauses
16245 The @option{^-A^/ALIGN^} switches are mutually compatible; any combination
16248 @node Casing Control
16249 @subsection Casing Control
16250 @cindex Casing control in @command{gnatpp}
16253 @command{gnatpp} allows you to specify the casing for reserved words,
16254 pragma names, attribute designators and identifiers.
16255 For identifiers you may define a
16256 general rule for name casing but also override this rule
16257 via a set of dictionary files.
16259 Three types of casing are supported: lower case, upper case, and mixed case.
16260 Lower and upper case are self-explanatory (but since some letters in
16261 Latin1 and other GNAT-supported character sets
16262 exist only in lower-case form, an upper case conversion will have no
16264 ``Mixed case'' means that the first letter, and also each letter immediately
16265 following an underscore, are converted to their uppercase forms;
16266 all the other letters are converted to their lowercase forms.
16269 @cindex @option{^-a@var{x}^/ATTRIBUTE^} (@command{gnatpp})
16270 @item ^-aL^/ATTRIBUTE_CASING=LOWER_CASE^
16271 Attribute designators are lower case
16273 @item ^-aU^/ATTRIBUTE_CASING=UPPER_CASE^
16274 Attribute designators are upper case
16276 @item ^-aM^/ATTRIBUTE_CASING=MIXED_CASE^
16277 Attribute designators are mixed case (this is the default)
16279 @cindex @option{^-k@var{x}^/KEYWORD_CASING^} (@command{gnatpp})
16280 @item ^-kL^/KEYWORD_CASING=LOWER_CASE^
16281 Keywords (technically, these are known in Ada as @emph{reserved words}) are
16282 lower case (this is the default)
16284 @item ^-kU^/KEYWORD_CASING=UPPER_CASE^
16285 Keywords are upper case
16287 @cindex @option{^-n@var{x}^/NAME_CASING^} (@command{gnatpp})
16288 @item ^-nD^/NAME_CASING=AS_DECLARED^
16289 Name casing for defining occurrences are as they appear in the source file
16290 (this is the default)
16292 @item ^-nU^/NAME_CASING=UPPER_CASE^
16293 Names are in upper case
16295 @item ^-nL^/NAME_CASING=LOWER_CASE^
16296 Names are in lower case
16298 @item ^-nM^/NAME_CASING=MIXED_CASE^
16299 Names are in mixed case
16301 @cindex @option{^-p@var{x}^/PRAGMA_CASING^} (@command{gnatpp})
16302 @item ^-pL^/PRAGMA_CASING=LOWER_CASE^
16303 Pragma names are lower case
16305 @item ^-pU^/PRAGMA_CASING=UPPER_CASE^
16306 Pragma names are upper case
16308 @item ^-pM^/PRAGMA_CASING=MIXED_CASE^
16309 Pragma names are mixed case (this is the default)
16311 @item ^-D@var{file}^/DICTIONARY=@var{file}^
16312 @cindex @option{^-D^/DICTIONARY^} (@command{gnatpp})
16313 Use @var{file} as a @emph{dictionary file} that defines
16314 the casing for a set of specified names,
16315 thereby overriding the effect on these names by
16316 any explicit or implicit
16317 ^-n^/NAME_CASING^ switch.
16318 To supply more than one dictionary file,
16319 use ^several @option{-D} switches^a list of files as options^.
16322 @option{gnatpp} implicitly uses a @emph{default dictionary file}
16323 to define the casing for the Ada predefined names and
16324 the names declared in the GNAT libraries.
16326 @item ^-D-^/SPECIFIC_CASING^
16327 @cindex @option{^-D-^/SPECIFIC_CASING^} (@command{gnatpp})
16328 Do not use the default dictionary file;
16329 instead, use the casing
16330 defined by a @option{^-n^/NAME_CASING^} switch and any explicit
16335 The structure of a dictionary file, and details on the conventions
16336 used in the default dictionary file, are defined in @ref{Name Casing}.
16338 The @option{^-D-^/SPECIFIC_CASING^} and
16339 @option{^-D@var{file}^/DICTIONARY=@var{file}^} switches are mutually
16342 @node Construct Layout Control
16343 @subsection Construct Layout Control
16344 @cindex Layout control in @command{gnatpp}
16347 This group of @command{gnatpp} switches controls the layout of comments and
16348 complex syntactic constructs. See @ref{Formatting Comments} for details
16352 @cindex @option{^-c@var{n}^/COMMENTS_LAYOUT^} (@command{gnatpp})
16353 @item ^-c0^/COMMENTS_LAYOUT=UNTOUCHED^
16354 All the comments remain unchanged
16356 @item ^-c1^/COMMENTS_LAYOUT=DEFAULT^
16357 GNAT-style comment line indentation (this is the default).
16359 @item ^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^
16360 Reference-manual comment line indentation.
16362 @item ^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^
16363 GNAT-style comment beginning
16365 @item ^-c4^/COMMENTS_LAYOUT=REFORMAT^
16366 Reformat comment blocks
16368 @item ^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^
16369 Keep unchanged special form comments
16371 Reformat comment blocks
16373 @cindex @option{^-l@var{n}^/CONSTRUCT_LAYOUT^} (@command{gnatpp})
16374 @item ^-l1^/CONSTRUCT_LAYOUT=GNAT^
16375 GNAT-style layout (this is the default)
16377 @item ^-l2^/CONSTRUCT_LAYOUT=COMPACT^
16380 @item ^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^
16383 @cindex @option{^-N^/NOTABS^} (@command{gnatpp})
16385 All the VT characters are removed from the comment text. All the HT characters
16386 are expanded with the sequences of space characters to get to the next tab
16389 @cindex @option{^--no-separate-is^/NO_SEPARATE_IS^} (@command{gnatpp})
16390 @item ^--no-separate-is^/NO_SEPARATE_IS^
16391 Do not place the keyword @code{is} on a separate line in a subprogram body in
16392 case if the spec occupies more then one line.
16394 @cindex @option{^--separate-label^/SEPARATE_LABEL^} (@command{gnatpp})
16395 @item ^--separate-label^/SEPARATE_LABEL^
16396 Place statement label(s) on a separate line, with the following statement
16399 @cindex @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} (@command{gnatpp})
16400 @item ^--separate-loop-then^/SEPARATE_LOOP_THEN^
16401 Place the keyword @code{loop} in FOR and WHILE loop statements and the
16402 keyword @code{then} in IF statements on a separate line.
16404 @cindex @option{^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^} (@command{gnatpp})
16405 @item ^--no-separate-loop-then^/NO_SEPARATE_LOOP_THEN^
16406 Do not place the keyword @code{loop} in FOR and WHILE loop statements and the
16407 keyword @code{then} in IF statements on a separate line. This option is
16408 incompatible with @option{^--separate-loop-then^/SEPARATE_LOOP_THEN^} option.
16410 @cindex @option{^--use-on-new-line^/USE_ON_NEW_LINE^} (@command{gnatpp})
16411 @item ^--use-on-new-line^/USE_ON_NEW_LINE^
16412 Start each USE clause in a context clause from a separate line.
16414 @cindex @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^} (@command{gnatpp})
16415 @item ^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^
16416 Use a separate line for a loop or block statement name, but do not use an extra
16417 indentation level for the statement itself.
16423 The @option{-c1} and @option{-c2} switches are incompatible.
16424 The @option{-c3} and @option{-c4} switches are compatible with each other and
16425 also with @option{-c1} and @option{-c2}. The @option{-c0} switch disables all
16426 the other comment formatting switches.
16428 The @option{-l1}, @option{-l2}, and @option{-l3} switches are incompatible.
16433 For the @option{/COMMENTS_LAYOUT} qualifier:
16436 The @option{DEFAULT} and @option{STANDARD_INDENT} options are incompatible.
16438 The @option{GNAT_BEGINNING} and @option{REFORMAT} options are compatible with
16439 each other and also with @option{DEFAULT} and @option{STANDARD_INDENT}.
16443 The @option{GNAT}, @option{COMPACT}, and @option{UNCOMPACT} options for the
16444 @option{/CONSTRUCT_LAYOUT} qualifier are incompatible.
16447 @node General Text Layout Control
16448 @subsection General Text Layout Control
16451 These switches allow control over line length and indentation.
16454 @item ^-M@var{nnn}^/LINE_LENGTH_MAX=@var{nnn}^
16455 @cindex @option{^-M^/LINE_LENGTH^} (@command{gnatpp})
16456 Maximum line length, @var{nnn} from 32@dots{}256, the default value is 79
16458 @item ^-i@var{nnn}^/INDENTATION_LEVEL=@var{nnn}^
16459 @cindex @option{^-i^/INDENTATION_LEVEL^} (@command{gnatpp})
16460 Indentation level, @var{nnn} from 1@dots{}9, the default value is 3
16462 @item ^-cl@var{nnn}^/CONTINUATION_INDENT=@var{nnn}^
16463 @cindex @option{^-cl^/CONTINUATION_INDENT^} (@command{gnatpp})
16464 Indentation level for continuation lines (relative to the line being
16465 continued), @var{nnn} from 1@dots{}9.
16467 value is one less then the (normal) indentation level, unless the
16468 indentation is set to 1 (in which case the default value for continuation
16469 line indentation is also 1)
16472 @node Other Formatting Options
16473 @subsection Other Formatting Options
16476 These switches control the inclusion of missing end/exit labels, and
16477 the indentation level in @b{case} statements.
16480 @item ^-e^/NO_MISSED_LABELS^
16481 @cindex @option{^-e^/NO_MISSED_LABELS^} (@command{gnatpp})
16482 Do not insert missing end/exit labels. An end label is the name of
16483 a construct that may optionally be repeated at the end of the
16484 construct's declaration;
16485 e.g., the names of packages, subprograms, and tasks.
16486 An exit label is the name of a loop that may appear as target
16487 of an exit statement within the loop.
16488 By default, @command{gnatpp} inserts these end/exit labels when
16489 they are absent from the original source. This option suppresses such
16490 insertion, so that the formatted source reflects the original.
16492 @item ^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^
16493 @cindex @option{^-ff^/FORM_FEED_AFTER_PRAGMA_PAGE^} (@command{gnatpp})
16494 Insert a Form Feed character after a pragma Page.
16496 @item ^-T@var{nnn}^/MAX_INDENT=@var{nnn}^
16497 @cindex @option{^-T^/MAX_INDENT^} (@command{gnatpp})
16498 Do not use an additional indentation level for @b{case} alternatives
16499 and variants if there are @var{nnn} or more (the default
16501 If @var{nnn} is 0, an additional indentation level is
16502 used for @b{case} alternatives and variants regardless of their number.
16505 @node Setting the Source Search Path
16506 @subsection Setting the Source Search Path
16509 To define the search path for the input source file, @command{gnatpp}
16510 uses the same switches as the GNAT compiler, with the same effects.
16513 @item ^-I^/SEARCH=^@var{dir}
16514 @cindex @option{^-I^/SEARCH^} (@code{gnatpp})
16515 The same as the corresponding gcc switch
16517 @item ^-I-^/NOCURRENT_DIRECTORY^
16518 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatpp})
16519 The same as the corresponding gcc switch
16521 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE^=@var{path}
16522 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@code{gnatpp})
16523 The same as the corresponding gcc switch
16525 @item ^--RTS^/RUNTIME_SYSTEM^=@var{path}
16526 @cindex @option{^--RTS^/RUNTIME_SYSTEM^} (@code{gnatpp})
16527 The same as the corresponding gcc switch
16531 @node Output File Control
16532 @subsection Output File Control
16535 By default the output is sent to the file whose name is obtained by appending
16536 the ^@file{.pp}^@file{$PP}^ suffix to the name of the input file
16537 (if the file with this name already exists, it is unconditionally overwritten).
16538 Thus if the input file is @file{^my_ada_proc.adb^MY_ADA_PROC.ADB^} then
16539 @command{gnatpp} will produce @file{^my_ada_proc.adb.pp^MY_ADA_PROC.ADB$PP^}
16541 The output may be redirected by the following switches:
16544 @item ^-pipe^/STANDARD_OUTPUT^
16545 @cindex @option{^-pipe^/STANDARD_OUTPUT^} (@code{gnatpp})
16546 Send the output to @code{Standard_Output}
16548 @item ^-o @var{output_file}^/OUTPUT=@var{output_file}^
16549 @cindex @option{^-o^/OUTPUT^} (@code{gnatpp})
16550 Write the output into @var{output_file}.
16551 If @var{output_file} already exists, @command{gnatpp} terminates without
16552 reading or processing the input file.
16554 @item ^-of ^/FORCED_OUTPUT=^@var{output_file}
16555 @cindex @option{^-of^/FORCED_OUTPUT^} (@code{gnatpp})
16556 Write the output into @var{output_file}, overwriting the existing file
16557 (if one is present).
16559 @item ^-r^/REPLACE^
16560 @cindex @option{^-r^/REPLACE^} (@code{gnatpp})
16561 Replace the input source file with the reformatted output, and copy the
16562 original input source into the file whose name is obtained by appending the
16563 ^@file{.npp}^@file{$NPP}^ suffix to the name of the input file.
16564 If a file with this name already exists, @command{gnatpp} terminates without
16565 reading or processing the input file.
16567 @item ^-rf^/OVERRIDING_REPLACE^
16568 @cindex @option{^-rf^/OVERRIDING_REPLACE^} (@code{gnatpp})
16569 Like @option{^-r^/REPLACE^} except that if the file with the specified name
16570 already exists, it is overwritten.
16572 @item ^-rnb^/REPLACE_NO_BACKUP^
16573 @cindex @option{^-rnb^/REPLACE_NO_BACKUP^} (@code{gnatpp})
16574 Replace the input source file with the reformatted output without
16575 creating any backup copy of the input source.
16577 @item ^--eol=@var{xxx}^/END_OF_LINE=@var{xxx}^
16578 @cindex @option{^--eol^/END_OF_LINE^} (@code{gnatpp})
16579 Specifies the format of the reformatted output file. The @var{xxx}
16580 ^string specified with the switch^option^ may be either
16582 @item ``@option{^dos^DOS^}'' MS DOS style, lines end with CR LF characters
16583 @item ``@option{^crlf^CRLF^}''
16584 the same as @option{^crlf^CRLF^}
16585 @item ``@option{^unix^UNIX^}'' UNIX style, lines end with LF character
16586 @item ``@option{^lf^LF^}''
16587 the same as @option{^unix^UNIX^}
16590 @item ^-W^/RESULT_ENCODING=^@var{e}
16591 @cindex @option{^-W^/RESULT_ENCODING=^} (@command{gnatpp})
16592 Specify the wide character encoding method used to write the code in the
16594 @var{e} is one of the following:
16602 Upper half encoding
16604 @item ^s^SHIFT_JIS^
16614 Brackets encoding (default value)
16620 Options @option{^-pipe^/STANDARD_OUTPUT^},
16621 @option{^-o^/OUTPUT^} and
16622 @option{^-of^/FORCED_OUTPUT^} are allowed only if the call to gnatpp
16623 contains only one file to reformat.
16625 @option{^--eol^/END_OF_LINE^}
16627 @option{^-W^/RESULT_ENCODING^}
16628 cannot be used together
16629 with @option{^-pipe^/STANDARD_OUTPUT^} option.
16631 @node Other gnatpp Switches
16632 @subsection Other @code{gnatpp} Switches
16635 The additional @command{gnatpp} switches are defined in this subsection.
16638 @item ^-files @var{filename}^/FILES=@var{output_file}^
16639 @cindex @option{^-files^/FILES^} (@code{gnatpp})
16640 Take the argument source files from the specified file. This file should be an
16641 ordinary textual file containing file names separated by spaces or
16642 line breaks. You can use this switch more then once in the same call to
16643 @command{gnatpp}. You also can combine this switch with explicit list of
16646 @item ^-v^/VERBOSE^
16647 @cindex @option{^-v^/VERBOSE^} (@code{gnatpp})
16649 @command{gnatpp} generates version information and then
16650 a trace of the actions it takes to produce or obtain the ASIS tree.
16652 @item ^-w^/WARNINGS^
16653 @cindex @option{^-w^/WARNINGS^} (@code{gnatpp})
16655 @command{gnatpp} generates a warning whenever it cannot provide
16656 a required layout in the result source.
16659 @node Formatting Rules
16660 @section Formatting Rules
16663 The following subsections show how @command{gnatpp} treats ``white space'',
16664 comments, program layout, and name casing.
16665 They provide the detailed descriptions of the switches shown above.
16668 * White Space and Empty Lines::
16669 * Formatting Comments::
16670 * Construct Layout::
16674 @node White Space and Empty Lines
16675 @subsection White Space and Empty Lines
16678 @command{gnatpp} does not have an option to control space characters.
16679 It will add or remove spaces according to the style illustrated by the
16680 examples in the @cite{Ada Reference Manual}.
16682 The only format effectors
16683 (see @cite{Ada Reference Manual}, paragraph 2.1(13))
16684 that will appear in the output file are platform-specific line breaks,
16685 and also format effectors within (but not at the end of) comments.
16686 In particular, each horizontal tab character that is not inside
16687 a comment will be treated as a space and thus will appear in the
16688 output file as zero or more spaces depending on
16689 the reformatting of the line in which it appears.
16690 The only exception is a Form Feed character, which is inserted after a
16691 pragma @code{Page} when @option{-ff} is set.
16693 The output file will contain no lines with trailing ``white space'' (spaces,
16696 Empty lines in the original source are preserved
16697 only if they separate declarations or statements.
16698 In such contexts, a
16699 sequence of two or more empty lines is replaced by exactly one empty line.
16700 Note that a blank line will be removed if it separates two ``comment blocks''
16701 (a comment block is a sequence of whole-line comments).
16702 In order to preserve a visual separation between comment blocks, use an
16703 ``empty comment'' (a line comprising only hyphens) rather than an empty line.
16704 Likewise, if for some reason you wish to have a sequence of empty lines,
16705 use a sequence of empty comments instead.
16707 @node Formatting Comments
16708 @subsection Formatting Comments
16711 Comments in Ada code are of two kinds:
16714 a @emph{whole-line comment}, which appears by itself (possibly preceded by
16715 ``white space'') on a line
16718 an @emph{end-of-line comment}, which follows some other Ada lexical element
16723 The indentation of a whole-line comment is that of either
16724 the preceding or following line in
16725 the formatted source, depending on switch settings as will be described below.
16727 For an end-of-line comment, @command{gnatpp} leaves the same number of spaces
16728 between the end of the preceding Ada lexical element and the beginning
16729 of the comment as appear in the original source,
16730 unless either the comment has to be split to
16731 satisfy the line length limitation, or else the next line contains a
16732 whole line comment that is considered a continuation of this end-of-line
16733 comment (because it starts at the same position).
16735 cases, the start of the end-of-line comment is moved right to the nearest
16736 multiple of the indentation level.
16737 This may result in a ``line overflow'' (the right-shifted comment extending
16738 beyond the maximum line length), in which case the comment is split as
16741 There is a difference between @option{^-c1^/COMMENTS_LAYOUT=DEFAULT^}
16742 (GNAT-style comment line indentation)
16743 and @option{^-c2^/COMMENTS_LAYOUT=STANDARD_INDENT^}
16744 (reference-manual comment line indentation).
16745 With reference-manual style, a whole-line comment is indented as if it
16746 were a declaration or statement at the same place
16747 (i.e., according to the indentation of the preceding line(s)).
16748 With GNAT style, a whole-line comment that is immediately followed by an
16749 @b{if} or @b{case} statement alternative, a record variant, or the reserved
16750 word @b{begin}, is indented based on the construct that follows it.
16753 @smallexample @c ada
16765 Reference-manual indentation produces:
16767 @smallexample @c ada
16779 while GNAT-style indentation produces:
16781 @smallexample @c ada
16793 The @option{^-c3^/COMMENTS_LAYOUT=GNAT_BEGINNING^} switch
16794 (GNAT style comment beginning) has the following
16799 For each whole-line comment that does not end with two hyphens,
16800 @command{gnatpp} inserts spaces if necessary after the starting two hyphens
16801 to ensure that there are at least two spaces between these hyphens and the
16802 first non-blank character of the comment.
16806 For an end-of-line comment, if in the original source the next line is a
16807 whole-line comment that starts at the same position
16808 as the end-of-line comment,
16809 then the whole-line comment (and all whole-line comments
16810 that follow it and that start at the same position)
16811 will start at this position in the output file.
16814 That is, if in the original source we have:
16816 @smallexample @c ada
16819 A := B + C; -- B must be in the range Low1..High1
16820 -- C must be in the range Low2..High2
16821 --B+C will be in the range Low1+Low2..High1+High2
16827 Then in the formatted source we get
16829 @smallexample @c ada
16832 A := B + C; -- B must be in the range Low1..High1
16833 -- C must be in the range Low2..High2
16834 -- B+C will be in the range Low1+Low2..High1+High2
16840 A comment that exceeds the line length limit will be split.
16842 @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} (reformat comment blocks) is set and
16843 the line belongs to a reformattable block, splitting the line generates a
16844 @command{gnatpp} warning.
16845 The @option{^-c4^/COMMENTS_LAYOUT=REFORMAT^} switch specifies that whole-line
16846 comments may be reformatted in typical
16847 word processor style (that is, moving words between lines and putting as
16848 many words in a line as possible).
16851 The @option{^-c5^/COMMENTS_LAYOUT=KEEP_SPECIAL^} switch specifies, that comments
16852 that has a special format (that is, a character that is neither a letter nor digit
16853 not white space nor line break immediately following the leading @code{--} of
16854 the comment) should be without any change moved from the argument source
16855 into reformatted source. This switch allows to preserve comments that are used
16856 as a special marks in the code (e.g.@: SPARK annotation).
16858 @node Construct Layout
16859 @subsection Construct Layout
16862 In several cases the suggested layout in the Ada Reference Manual includes
16863 an extra level of indentation that many programmers prefer to avoid. The
16864 affected cases include:
16868 @item Record type declaration (RM 3.8)
16870 @item Record representation clause (RM 13.5.1)
16872 @item Loop statement in case if a loop has a statement identifier (RM 5.6)
16874 @item Block statement in case if a block has a statement identifier (RM 5.6)
16878 In compact mode (when GNAT style layout or compact layout is set),
16879 the pretty printer uses one level of indentation instead
16880 of two. This is achieved in the record definition and record representation
16881 clause cases by putting the @code{record} keyword on the same line as the
16882 start of the declaration or representation clause, and in the block and loop
16883 case by putting the block or loop header on the same line as the statement
16887 The difference between GNAT style @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^}
16888 and compact @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^}
16889 layout on the one hand, and uncompact layout
16890 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} on the other hand,
16891 can be illustrated by the following examples:
16895 @multitable @columnfractions .5 .5
16896 @item @i{GNAT style, compact layout} @tab @i{Uncompact layout}
16899 @smallexample @c ada
16906 @smallexample @c ada
16915 @smallexample @c ada
16917 a at 0 range 0 .. 31;
16918 b at 4 range 0 .. 31;
16922 @smallexample @c ada
16925 a at 0 range 0 .. 31;
16926 b at 4 range 0 .. 31;
16931 @smallexample @c ada
16939 @smallexample @c ada
16949 @smallexample @c ada
16950 Clear : for J in 1 .. 10 loop
16955 @smallexample @c ada
16957 for J in 1 .. 10 loop
16968 GNAT style, compact layout Uncompact layout
16970 type q is record type q is
16971 a : integer; record
16972 b : integer; a : integer;
16973 end record; b : integer;
16976 for q use record for q use
16977 a at 0 range 0 .. 31; record
16978 b at 4 range 0 .. 31; a at 0 range 0 .. 31;
16979 end record; b at 4 range 0 .. 31;
16982 Block : declare Block :
16983 A : Integer := 3; declare
16984 begin A : Integer := 3;
16986 end Block; Proc (A, A);
16989 Clear : for J in 1 .. 10 loop Clear :
16990 A (J) := 0; for J in 1 .. 10 loop
16991 end loop Clear; A (J) := 0;
16998 A further difference between GNAT style layout and compact layout is that
16999 GNAT style layout inserts empty lines as separation for
17000 compound statements, return statements and bodies.
17002 Note that the layout specified by
17003 @option{^--separate-stmt-name^/STMT_NAME_ON_NEW_LINE^}
17004 for named block and loop statements overrides the layout defined by these
17005 constructs by @option{^-l1^/CONSTRUCT_LAYOUT=GNAT^},
17006 @option{^-l2^/CONSTRUCT_LAYOUT=COMPACT^} or
17007 @option{^-l3^/CONSTRUCT_LAYOUT=UNCOMPACT^} option.
17010 @subsection Name Casing
17013 @command{gnatpp} always converts the usage occurrence of a (simple) name to
17014 the same casing as the corresponding defining identifier.
17016 You control the casing for defining occurrences via the
17017 @option{^-n^/NAME_CASING^} switch.
17019 With @option{-nD} (``as declared'', which is the default),
17022 With @option{/NAME_CASING=AS_DECLARED}, which is the default,
17024 defining occurrences appear exactly as in the source file
17025 where they are declared.
17026 The other ^values for this switch^options for this qualifier^ ---
17027 @option{^-nU^UPPER_CASE^},
17028 @option{^-nL^LOWER_CASE^},
17029 @option{^-nM^MIXED_CASE^} ---
17031 ^upper, lower, or mixed case, respectively^the corresponding casing^.
17032 If @command{gnatpp} changes the casing of a defining
17033 occurrence, it analogously changes the casing of all the
17034 usage occurrences of this name.
17036 If the defining occurrence of a name is not in the source compilation unit
17037 currently being processed by @command{gnatpp}, the casing of each reference to
17038 this name is changed according to the value of the @option{^-n^/NAME_CASING^}
17039 switch (subject to the dictionary file mechanism described below).
17040 Thus @command{gnatpp} acts as though the @option{^-n^/NAME_CASING^} switch
17042 casing for the defining occurrence of the name.
17044 Some names may need to be spelled with casing conventions that are not
17045 covered by the upper-, lower-, and mixed-case transformations.
17046 You can arrange correct casing by placing such names in a
17047 @emph{dictionary file},
17048 and then supplying a @option{^-D^/DICTIONARY^} switch.
17049 The casing of names from dictionary files overrides
17050 any @option{^-n^/NAME_CASING^} switch.
17052 To handle the casing of Ada predefined names and the names from GNAT libraries,
17053 @command{gnatpp} assumes a default dictionary file.
17054 The name of each predefined entity is spelled with the same casing as is used
17055 for the entity in the @cite{Ada Reference Manual}.
17056 The name of each entity in the GNAT libraries is spelled with the same casing
17057 as is used in the declaration of that entity.
17059 The @w{@option{^-D-^/SPECIFIC_CASING^}} switch suppresses the use of the
17060 default dictionary file.
17061 Instead, the casing for predefined and GNAT-defined names will be established
17062 by the @option{^-n^/NAME_CASING^} switch or explicit dictionary files.
17063 For example, by default the names @code{Ada.Text_IO} and @code{GNAT.OS_Lib}
17064 will appear as just shown,
17065 even in the presence of a @option{^-nU^/NAME_CASING=UPPER_CASE^} switch.
17066 To ensure that even such names are rendered in uppercase,
17067 additionally supply the @w{@option{^-D-^/SPECIFIC_CASING^}} switch
17068 (or else, less conveniently, place these names in upper case in a dictionary
17071 A dictionary file is
17072 a plain text file; each line in this file can be either a blank line
17073 (containing only space characters and ASCII.HT characters), an Ada comment
17074 line, or the specification of exactly one @emph{casing schema}.
17076 A casing schema is a string that has the following syntax:
17080 @var{casing_schema} ::= @var{identifier} | *@var{simple_identifier}*
17082 @var{simple_identifier} ::= @var{letter}@{@var{letter_or_digit}@}
17087 (See @cite{Ada Reference Manual}, Section 2.3) for the definition of the
17088 @var{identifier} lexical element and the @var{letter_or_digit} category.)
17090 The casing schema string can be followed by white space and/or an Ada-style
17091 comment; any amount of white space is allowed before the string.
17093 If a dictionary file is passed as
17095 the value of a @option{-D@var{file}} switch
17098 an option to the @option{/DICTIONARY} qualifier
17101 simple name and every identifier, @command{gnatpp} checks if the dictionary
17102 defines the casing for the name or for some of its parts (the term ``subword''
17103 is used below to denote the part of a name which is delimited by ``_'' or by
17104 the beginning or end of the word and which does not contain any ``_'' inside):
17108 if the whole name is in the dictionary, @command{gnatpp} uses for this name
17109 the casing defined by the dictionary; no subwords are checked for this word
17112 for every subword @command{gnatpp} checks if the dictionary contains the
17113 corresponding string of the form @code{*@var{simple_identifier}*},
17114 and if it does, the casing of this @var{simple_identifier} is used
17118 if the whole name does not contain any ``_'' inside, and if for this name
17119 the dictionary contains two entries - one of the form @var{identifier},
17120 and another - of the form *@var{simple_identifier}*, then the first one
17121 is applied to define the casing of this name
17124 if more than one dictionary file is passed as @command{gnatpp} switches, each
17125 dictionary adds new casing exceptions and overrides all the existing casing
17126 exceptions set by the previous dictionaries
17129 when @command{gnatpp} checks if the word or subword is in the dictionary,
17130 this check is not case sensitive
17134 For example, suppose we have the following source to reformat:
17136 @smallexample @c ada
17139 name1 : integer := 1;
17140 name4_name3_name2 : integer := 2;
17141 name2_name3_name4 : Boolean;
17144 name2_name3_name4 := name4_name3_name2 > name1;
17150 And suppose we have two dictionaries:
17167 If @command{gnatpp} is called with the following switches:
17171 @command{gnatpp -nM -D dict1 -D dict2 test.adb}
17174 @command{gnatpp test.adb /NAME_CASING=MIXED_CASE /DICTIONARY=(dict1, dict2)}
17179 then we will get the following name casing in the @command{gnatpp} output:
17181 @smallexample @c ada
17184 NAME1 : Integer := 1;
17185 Name4_NAME3_Name2 : Integer := 2;
17186 Name2_NAME3_Name4 : Boolean;
17189 Name2_NAME3_Name4 := Name4_NAME3_Name2 > NAME1;
17194 @c *********************************
17195 @node The GNAT Metric Tool gnatmetric
17196 @chapter The GNAT Metric Tool @command{gnatmetric}
17198 @cindex Metric tool
17201 ^The @command{gnatmetric} tool^@command{GNAT METRIC}^ is an ASIS-based utility
17202 for computing various program metrics.
17203 It takes an Ada source file as input and generates a file containing the
17204 metrics data as output. Various switches control which
17205 metrics are computed and output.
17207 @command{gnatmetric} generates and uses the ASIS
17208 tree for the input source and thus requires the input to be syntactically and
17209 semantically legal.
17210 If this condition is not met, @command{gnatmetric} will generate
17211 an error message; no metric information for this file will be
17212 computed and reported.
17214 If the compilation unit contained in the input source depends semantically
17215 upon units in files located outside the current directory, you have to provide
17216 the source search path when invoking @command{gnatmetric}.
17217 If it depends semantically upon units that are contained
17218 in files with names that do not follow the GNAT file naming rules, you have to
17219 provide the configuration file describing the corresponding naming scheme (see
17220 the description of the @command{gnatmetric} switches below.)
17221 Alternatively, you may use a project file and invoke @command{gnatmetric}
17222 through the @command{gnat} driver.
17224 The @command{gnatmetric} command has the form
17227 $ gnatmetric @ovar{switches} @{@var{filename}@} @r{[}-cargs @var{gcc_switches}@r{]}
17234 @var{switches} specify the metrics to compute and define the destination for
17238 Each @var{filename} is the name (including the extension) of a source
17239 file to process. ``Wildcards'' are allowed, and
17240 the file name may contain path information.
17241 If no @var{filename} is supplied, then the @var{switches} list must contain
17243 @option{-files} switch (@pxref{Other gnatmetric Switches}).
17244 Including both a @option{-files} switch and one or more
17245 @var{filename} arguments is permitted.
17248 @samp{-cargs @var{gcc_switches}} is a list of switches for
17249 @command{gcc}. They will be passed on to all compiler invocations made by
17250 @command{gnatmetric} to generate the ASIS trees. Here you can provide
17251 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
17252 and use the @option{-gnatec} switch to set the configuration file.
17256 * Switches for gnatmetric::
17259 @node Switches for gnatmetric
17260 @section Switches for @command{gnatmetric}
17263 The following subsections describe the various switches accepted by
17264 @command{gnatmetric}, organized by category.
17267 * Output Files Control::
17268 * Disable Metrics For Local Units::
17269 * Specifying a set of metrics to compute::
17270 * Other gnatmetric Switches::
17271 * Generate project-wide metrics::
17274 @node Output Files Control
17275 @subsection Output File Control
17276 @cindex Output file control in @command{gnatmetric}
17279 @command{gnatmetric} has two output formats. It can generate a
17280 textual (human-readable) form, and also XML. By default only textual
17281 output is generated.
17283 When generating the output in textual form, @command{gnatmetric} creates
17284 for each Ada source file a corresponding text file
17285 containing the computed metrics, except for the case when the set of metrics
17286 specified by gnatmetric parameters consists only of metrics that are computed
17287 for the whole set of analyzed sources, but not for each Ada source.
17288 By default, this file is placed in the same directory as where the source
17289 file is located, and its name is obtained
17290 by appending the ^@file{.metrix}^@file{$METRIX}^ suffix to the name of the
17293 All the output information generated in XML format is placed in a single
17294 file. By default this file is placed in the current directory and has the
17295 name ^@file{metrix.xml}^@file{METRIX$XML}^.
17297 Some of the computed metrics are summed over the units passed to
17298 @command{gnatmetric}; for example, the total number of lines of code.
17299 By default this information is sent to @file{stdout}, but a file
17300 can be specified with the @option{-og} switch.
17302 The following switches control the @command{gnatmetric} output:
17305 @cindex @option{^-x^/XML^} (@command{gnatmetric})
17307 Generate the XML output
17309 @cindex @option{^-xs^/XSD^} (@command{gnatmetric})
17311 Generate the XML output and the XML schema file that describes the structure
17312 of the XML metric report, this schema is assigned to the XML file. The schema
17313 file has the same name as the XML output file with @file{.xml} suffix replaced
17316 @cindex @option{^-nt^/NO_TEXT^} (@command{gnatmetric})
17317 @item ^-nt^/NO_TEXT^
17318 Do not generate the output in text form (implies @option{^-x^/XML^})
17320 @cindex @option{^-d^/DIRECTORY^} (@command{gnatmetric})
17321 @item ^-d @var{output_dir}^/DIRECTORY=@var{output_dir}^
17322 Put textual files with detailed metrics into @var{output_dir}
17324 @cindex @option{^-o^/SUFFIX_DETAILS^} (@command{gnatmetric})
17325 @item ^-o @var{file_suffix}^/SUFFIX_DETAILS=@var{file_suffix}^
17326 Use @var{file_suffix}, instead of ^@file{.metrix}^@file{$METRIX}^
17327 in the name of the output file.
17329 @cindex @option{^-og^/GLOBAL_OUTPUT^} (@command{gnatmetric})
17330 @item ^-og @var{file_name}^/GLOBAL_OUTPUT=@var{file_name}^
17331 Put global metrics into @var{file_name}
17333 @cindex @option{^-ox^/XML_OUTPUT^} (@command{gnatmetric})
17334 @item ^-ox @var{file_name}^/XML_OUTPUT=@var{file_name}^
17335 Put the XML output into @var{file_name} (also implies @option{^-x^/XML^})
17337 @cindex @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} (@command{gnatmetric})
17338 @item ^-sfn^/SHORT_SOURCE_FILE_NAME^
17339 Use ``short'' source file names in the output. (The @command{gnatmetric}
17340 output includes the name(s) of the Ada source file(s) from which the metrics
17341 are computed. By default each name includes the absolute path. The
17342 @option{^-sfn^/SHORT_SOURCE_FILE_NAME^} switch causes @command{gnatmetric}
17343 to exclude all directory information from the file names that are output.)
17347 @node Disable Metrics For Local Units
17348 @subsection Disable Metrics For Local Units
17349 @cindex Disable Metrics For Local Units in @command{gnatmetric}
17352 @command{gnatmetric} relies on the GNAT compilation model @minus{}
17354 unit per one source file. It computes line metrics for the whole source
17355 file, and it also computes syntax
17356 and complexity metrics for the file's outermost unit.
17358 By default, @command{gnatmetric} will also compute all metrics for certain
17359 kinds of locally declared program units:
17363 subprogram (and generic subprogram) bodies;
17366 package (and generic package) specs and bodies;
17369 task object and type specifications and bodies;
17372 protected object and type specifications and bodies.
17376 These kinds of entities will be referred to as
17377 @emph{eligible local program units}, or simply @emph{eligible local units},
17378 @cindex Eligible local unit (for @command{gnatmetric})
17379 in the discussion below.
17381 Note that a subprogram declaration, generic instantiation,
17382 or renaming declaration only receives metrics
17383 computation when it appear as the outermost entity
17386 Suppression of metrics computation for eligible local units can be
17387 obtained via the following switch:
17390 @cindex @option{^-n@var{x}^/SUPPRESS^} (@command{gnatmetric})
17391 @item ^-nolocal^/SUPPRESS=LOCAL_DETAILS^
17392 Do not compute detailed metrics for eligible local program units
17396 @node Specifying a set of metrics to compute
17397 @subsection Specifying a set of metrics to compute
17400 By default all the metrics are computed and reported. The switches
17401 described in this subsection allow you to control, on an individual
17402 basis, whether metrics are computed and
17403 reported. If at least one positive metric
17404 switch is specified (that is, a switch that defines that a given
17405 metric or set of metrics is to be computed), then only
17406 explicitly specified metrics are reported.
17409 * Line Metrics Control::
17410 * Syntax Metrics Control::
17411 * Complexity Metrics Control::
17412 * Object-Oriented Metrics Control::
17415 @node Line Metrics Control
17416 @subsubsection Line Metrics Control
17417 @cindex Line metrics control in @command{gnatmetric}
17420 For any (legal) source file, and for each of its
17421 eligible local program units, @command{gnatmetric} computes the following
17426 the total number of lines;
17429 the total number of code lines (i.e., non-blank lines that are not comments)
17432 the number of comment lines
17435 the number of code lines containing end-of-line comments;
17438 the comment percentage: the ratio between the number of lines that contain
17439 comments and the number of all non-blank lines, expressed as a percentage;
17442 the number of empty lines and lines containing only space characters and/or
17443 format effectors (blank lines)
17446 the average number of code lines in subprogram bodies, task bodies, entry
17447 bodies and statement sequences in package bodies (this metric is only computed
17448 across the whole set of the analyzed units)
17453 @command{gnatmetric} sums the values of the line metrics for all the
17454 files being processed and then generates the cumulative results. The tool
17455 also computes for all the files being processed the average number of code
17458 You can use the following switches to select the specific line metrics
17459 to be computed and reported.
17462 @cindex @option{^--lines@var{x}^/LINE_COUNT_METRICS^} (@command{gnatmetric})
17465 @cindex @option{--no-lines@var{x}}
17468 @item ^--lines-all^/LINE_COUNT_METRICS=ALL^
17469 Report all the line metrics
17471 @item ^--no-lines-all^/LINE_COUNT_METRICS=NONE^
17472 Do not report any of line metrics
17474 @item ^--lines^/LINE_COUNT_METRICS=ALL_LINES^
17475 Report the number of all lines
17477 @item ^--no-lines^/LINE_COUNT_METRICS=NOALL_LINES^
17478 Do not report the number of all lines
17480 @item ^--lines-code^/LINE_COUNT_METRICS=CODE_LINES^
17481 Report the number of code lines
17483 @item ^--no-lines-code^/LINE_COUNT_METRICS=NOCODE_LINES^
17484 Do not report the number of code lines
17486 @item ^--lines-comment^/LINE_COUNT_METRICS=COMMENT_LINES^
17487 Report the number of comment lines
17489 @item ^--no-lines-comment^/LINE_COUNT_METRICS=NOCOMMENT_LINES^
17490 Do not report the number of comment lines
17492 @item ^--lines-eol-comment^/LINE_COUNT_METRICS=CODE_COMMENT_LINES^
17493 Report the number of code lines containing
17494 end-of-line comments
17496 @item ^--no-lines-eol-comment^/LINE_COUNT_METRICS=NOCODE_COMMENT_LINES^
17497 Do not report the number of code lines containing
17498 end-of-line comments
17500 @item ^--lines-ratio^/LINE_COUNT_METRICS=COMMENT_PERCENTAGE^
17501 Report the comment percentage in the program text
17503 @item ^--no-lines-ratio^/LINE_COUNT_METRICS=NOCOMMENT_PERCENTAGE^
17504 Do not report the comment percentage in the program text
17506 @item ^--lines-blank^/LINE_COUNT_METRICS=BLANK_LINES^
17507 Report the number of blank lines
17509 @item ^--no-lines-blank^/LINE_COUNT_METRICS=NOBLANK_LINES^
17510 Do not report the number of blank lines
17512 @item ^--lines-average^/LINE_COUNT_METRICS=AVERAGE_BODY_LINES^
17513 Report the average number of code lines in subprogram bodies, task bodies,
17514 entry bodies and statement sequences in package bodies. The metric is computed
17515 and reported for the whole set of processed Ada sources only.
17517 @item ^--no-lines-average^/LINE_COUNT_METRICS=NOAVERAGE_BODY_LINES^
17518 Do not report the average number of code lines in subprogram bodies,
17519 task bodies, entry bodies and statement sequences in package bodies.
17523 @node Syntax Metrics Control
17524 @subsubsection Syntax Metrics Control
17525 @cindex Syntax metrics control in @command{gnatmetric}
17528 @command{gnatmetric} computes various syntactic metrics for the
17529 outermost unit and for each eligible local unit:
17532 @item LSLOC (``Logical Source Lines Of Code'')
17533 The total number of declarations and the total number of statements
17535 @item Maximal static nesting level of inner program units
17537 @cite{Ada Reference Manual}, 10.1(1), ``A program unit is either a
17538 package, a task unit, a protected unit, a
17539 protected entry, a generic unit, or an explicitly declared subprogram other
17540 than an enumeration literal.''
17542 @item Maximal nesting level of composite syntactic constructs
17543 This corresponds to the notion of the
17544 maximum nesting level in the GNAT built-in style checks
17545 (@pxref{Style Checking})
17549 For the outermost unit in the file, @command{gnatmetric} additionally computes
17550 the following metrics:
17553 @item Public subprograms
17554 This metric is computed for package specs. It is the
17555 number of subprograms and generic subprograms declared in the visible
17556 part (including the visible part of nested packages, protected objects, and
17559 @item All subprograms
17560 This metric is computed for bodies and subunits. The
17561 metric is equal to a total number of subprogram bodies in the compilation
17563 Neither generic instantiations nor renamings-as-a-body nor body stubs
17564 are counted. Any subprogram body is counted, independently of its nesting
17565 level and enclosing constructs. Generic bodies and bodies of protected
17566 subprograms are counted in the same way as ``usual'' subprogram bodies.
17569 This metric is computed for package specs and
17570 generic package declarations. It is the total number of types
17571 that can be referenced from outside this compilation unit, plus the
17572 number of types from all the visible parts of all the visible generic
17573 packages. Generic formal types are not counted. Only types, not subtypes,
17577 Along with the total number of public types, the following
17578 types are counted and reported separately:
17585 Root tagged types (abstract, non-abstract, private, non-private). Type
17586 extensions are @emph{not} counted
17589 Private types (including private extensions)
17600 This metric is computed for any compilation unit. It is equal to the total
17601 number of the declarations of different types given in the compilation unit.
17602 The private and the corresponding full type declaration are counted as one
17603 type declaration. Incomplete type declarations and generic formal types
17605 No distinction is made among different kinds of types (abstract,
17606 private etc.); the total number of types is computed and reported.
17611 By default, all the syntax metrics are computed and reported. You can use the
17612 following switches to select specific syntax metrics.
17616 @cindex @option{^--syntax@var{x}^/SYNTAX_METRICS^} (@command{gnatmetric})
17619 @cindex @option{--no-syntax@var{x}} (@command{gnatmetric})
17622 @item ^--syntax-all^/SYNTAX_METRICS=ALL^
17623 Report all the syntax metrics
17625 @item ^--no-syntax-all^/SYNTAX_METRICS=NONE^
17626 Do not report any of syntax metrics
17628 @item ^--declarations^/SYNTAX_METRICS=DECLARATIONS^
17629 Report the total number of declarations
17631 @item ^--no-declarations^/SYNTAX_METRICS=NODECLARATIONS^
17632 Do not report the total number of declarations
17634 @item ^--statements^/SYNTAX_METRICS=STATEMENTS^
17635 Report the total number of statements
17637 @item ^--no-statements^/SYNTAX_METRICS=NOSTATEMENTS^
17638 Do not report the total number of statements
17640 @item ^--public-subprograms^/SYNTAX_METRICS=PUBLIC_SUBPROGRAMS^
17641 Report the number of public subprograms in a compilation unit
17643 @item ^--no-public-subprograms^/SYNTAX_METRICS=NOPUBLIC_SUBPROGRAMS^
17644 Do not report the number of public subprograms in a compilation unit
17646 @item ^--all-subprograms^/SYNTAX_METRICS=ALL_SUBPROGRAMS^
17647 Report the number of all the subprograms in a compilation unit
17649 @item ^--no-all-subprograms^/SYNTAX_METRICS=NOALL_SUBPROGRAMS^
17650 Do not report the number of all the subprograms in a compilation unit
17652 @item ^--public-types^/SYNTAX_METRICS=PUBLIC_TYPES^
17653 Report the number of public types in a compilation unit
17655 @item ^--no-public-types^/SYNTAX_METRICS=NOPUBLIC_TYPES^
17656 Do not report the number of public types in a compilation unit
17658 @item ^--all-types^/SYNTAX_METRICS=ALL_TYPES^
17659 Report the number of all the types in a compilation unit
17661 @item ^--no-all-types^/SYNTAX_METRICS=NOALL_TYPES^
17662 Do not report the number of all the types in a compilation unit
17664 @item ^--unit-nesting^/SYNTAX_METRICS=UNIT_NESTING^
17665 Report the maximal program unit nesting level
17667 @item ^--no-unit-nesting^/SYNTAX_METRICS=UNIT_NESTING_OFF^
17668 Do not report the maximal program unit nesting level
17670 @item ^--construct-nesting^/SYNTAX_METRICS=CONSTRUCT_NESTING^
17671 Report the maximal construct nesting level
17673 @item ^--no-construct-nesting^/SYNTAX_METRICS=NOCONSTRUCT_NESTING^
17674 Do not report the maximal construct nesting level
17678 @node Complexity Metrics Control
17679 @subsubsection Complexity Metrics Control
17680 @cindex Complexity metrics control in @command{gnatmetric}
17683 For a program unit that is an executable body (a subprogram body (including
17684 generic bodies), task body, entry body or a package body containing
17685 its own statement sequence) @command{gnatmetric} computes the following
17686 complexity metrics:
17690 McCabe cyclomatic complexity;
17693 McCabe essential complexity;
17696 maximal loop nesting level
17701 The McCabe complexity metrics are defined
17702 in @url{http://www.mccabe.com/pdf/nist235r.pdf}
17704 According to McCabe, both control statements and short-circuit control forms
17705 should be taken into account when computing cyclomatic complexity. For each
17706 body, we compute three metric values:
17710 the complexity introduced by control
17711 statements only, without taking into account short-circuit forms,
17714 the complexity introduced by short-circuit control forms only, and
17718 cyclomatic complexity, which is the sum of these two values.
17722 When computing cyclomatic and essential complexity, @command{gnatmetric} skips
17723 the code in the exception handlers and in all the nested program units.
17725 By default, all the complexity metrics are computed and reported.
17726 For more fine-grained control you can use
17727 the following switches:
17730 @cindex @option{^-complexity@var{x}^/COMPLEXITY_METRICS^} (@command{gnatmetric})
17733 @cindex @option{--no-complexity@var{x}}
17736 @item ^--complexity-all^/COMPLEXITY_METRICS=ALL^
17737 Report all the complexity metrics
17739 @item ^--no-complexity-all^/COMPLEXITY_METRICS=NONE^
17740 Do not report any of complexity metrics
17742 @item ^--complexity-cyclomatic^/COMPLEXITY_METRICS=CYCLOMATIC^
17743 Report the McCabe Cyclomatic Complexity
17745 @item ^--no-complexity-cyclomatic^/COMPLEXITY_METRICS=NOCYCLOMATIC^
17746 Do not report the McCabe Cyclomatic Complexity
17748 @item ^--complexity-essential^/COMPLEXITY_METRICS=ESSENTIAL^
17749 Report the Essential Complexity
17751 @item ^--no-complexity-essential^/COMPLEXITY_METRICS=NOESSENTIAL^
17752 Do not report the Essential Complexity
17754 @item ^--loop-nesting^/COMPLEXITY_METRICS=LOOP_NESTING_ON^
17755 Report maximal loop nesting level
17757 @item ^--no-loop-nesting^/COMPLEXITY_METRICS=NOLOOP_NESTING^
17758 Do not report maximal loop nesting level
17760 @item ^--complexity-average^/COMPLEXITY_METRICS=AVERAGE_COMPLEXITY^
17761 Report the average McCabe Cyclomatic Complexity for all the subprogram bodies,
17762 task bodies, entry bodies and statement sequences in package bodies.
17763 The metric is computed and reported for whole set of processed Ada sources
17766 @item ^--no-complexity-average^/COMPLEXITY_METRICS=NOAVERAGE_COMPLEXITY^
17767 Do not report the average McCabe Cyclomatic Complexity for all the subprogram
17768 bodies, task bodies, entry bodies and statement sequences in package bodies
17770 @cindex @option{^-ne^/NO_EXITS_AS_GOTOS^} (@command{gnatmetric})
17771 @item ^-ne^/NO_EXITS_AS_GOTOS^
17772 Do not consider @code{exit} statements as @code{goto}s when
17773 computing Essential Complexity
17775 @item ^--extra-exit-points^/EXTRA_EXIT_POINTS^
17776 Report the extra exit points for subprogram bodies. As an exit point, this
17777 metric counts @code{return} statements and raise statements in case when the
17778 raised exception is not handled in the same body. In case of a function this
17779 metric subtracts 1 from the number of exit points, because a function body
17780 must contain at least one @code{return} statement.
17782 @item ^--no-extra-exit-points^/NOEXTRA_EXIT_POINTS^
17783 Do not report the extra exit points for subprogram bodies
17787 @node Object-Oriented Metrics Control
17788 @subsubsection Object-Oriented Metrics Control
17789 @cindex Object-Oriented metrics control in @command{gnatmetric}
17792 @cindex Coupling metrics (in in @command{gnatmetric})
17793 Coupling metrics are object-oriented metrics that measure the
17794 dependencies between a given class (or a group of classes) and the
17795 ``external world'' (that is, the other classes in the program). In this
17796 subsection the term ``class'' is used in its
17797 traditional object-oriented programming sense
17798 (an instantiable module that contains data and/or method members).
17799 A @emph{category} (of classes)
17800 is a group of closely related classes that are reused and/or
17803 A class @code{K}'s @emph{efferent coupling} is the number of classes
17804 that @code{K} depends upon.
17805 A category's efferent coupling is the number of classes outside the
17806 category that the classes inside the category depend upon.
17808 A class @code{K}'s @emph{afferent coupling} is the number of classes
17809 that depend upon @code{K}.
17810 A category's afferent coupling is the number of classes outside the
17811 category that depend on classes belonging to the category.
17813 Ada's implementation of the object-oriented paradigm does not use the
17814 traditional class notion, so the definition of the coupling
17815 metrics for Ada maps the class and class category notions
17816 onto Ada constructs.
17818 For the coupling metrics, several kinds of modules -- a library package,
17819 a library generic package, and a library generic package instantiation --
17820 that define a tagged type or an interface type are
17821 considered to be a class. A category consists of a library package (or
17822 a library generic package) that defines a tagged or an interface type,
17823 together with all its descendant (generic) packages that define tagged
17824 or interface types. For any package counted as a class,
17825 its body and subunits (if any) are considered
17826 together with its spec when counting the dependencies, and coupling
17827 metrics are reported for spec units only. For dependencies
17828 between classes, the Ada semantic dependencies are considered.
17829 For coupling metrics, only dependencies on units that are considered as
17830 classes, are considered.
17832 When computing coupling metrics, @command{gnatmetric} counts only
17833 dependencies between units that are arguments of the gnatmetric call.
17834 Coupling metrics are program-wide (or project-wide) metrics, so to
17835 get a valid result, you should call @command{gnatmetric} for
17836 the whole set of sources that make up your program. It can be done
17837 by calling @command{gnatmetric} from the GNAT driver with @option{-U}
17838 option (see See @ref{The GNAT Driver and Project Files} for details.
17840 By default, all the coupling metrics are disabled. You can use the following
17841 switches to specify the coupling metrics to be computed and reported:
17846 @cindex @option{--package@var{x}} (@command{gnatmetric})
17847 @cindex @option{--no-package@var{x}} (@command{gnatmetric})
17848 @cindex @option{--category@var{x}} (@command{gnatmetric})
17849 @cindex @option{--no-category@var{x}} (@command{gnatmetric})
17853 @cindex @option{/COUPLING_METRICS} (@command{gnatmetric})
17856 @item ^--coupling-all^/COUPLING_METRICS=ALL^
17857 Report all the coupling metrics
17859 @item ^--no-coupling-all^/COUPLING_METRICS=NONE^
17860 Do not report any of metrics
17862 @item ^--package-efferent-coupling^/COUPLING_METRICS=PACKAGE_EFFERENT^
17863 Report package efferent coupling
17865 @item ^--no-package-efferent-coupling^/COUPLING_METRICS=NOPACKAGE_EFFERENT^
17866 Do not report package efferent coupling
17868 @item ^--package-afferent-coupling^/COUPLING_METRICS=PACKAGE_AFFERENT^
17869 Report package afferent coupling
17871 @item ^--no-package-afferent-coupling^/COUPLING_METRICS=NOPACKAGE_AFFERENT^
17872 Do not report package afferent coupling
17874 @item ^--category-efferent-coupling^/COUPLING_METRICS=CATEGORY_EFFERENT^
17875 Report category efferent coupling
17877 @item ^--no-category-efferent-coupling^/COUPLING_METRICS=NOCATEGORY_EFFERENT^
17878 Do not report category efferent coupling
17880 @item ^--category-afferent-coupling^/COUPLING_METRICS=CATEGORY_AFFERENT^
17881 Report category afferent coupling
17883 @item ^--no-category-afferent-coupling^/COUPLING_METRICS=NOCATEGORY_AFFERENT^
17884 Do not report category afferent coupling
17888 @node Other gnatmetric Switches
17889 @subsection Other @code{gnatmetric} Switches
17892 Additional @command{gnatmetric} switches are as follows:
17895 @item ^-files @var{filename}^/FILES=@var{filename}^
17896 @cindex @option{^-files^/FILES^} (@code{gnatmetric})
17897 Take the argument source files from the specified file. This file should be an
17898 ordinary text file containing file names separated by spaces or
17899 line breaks. You can use this switch more then once in the same call to
17900 @command{gnatmetric}. You also can combine this switch with
17901 an explicit list of files.
17903 @item ^-v^/VERBOSE^
17904 @cindex @option{^-v^/VERBOSE^} (@code{gnatmetric})
17906 @command{gnatmetric} generates version information and then
17907 a trace of sources being processed.
17909 @item ^-dv^/DEBUG_OUTPUT^
17910 @cindex @option{^-dv^/DEBUG_OUTPUT^} (@code{gnatmetric})
17912 @command{gnatmetric} generates various messages useful to understand what
17913 happens during the metrics computation
17916 @cindex @option{^-q^/QUIET^} (@code{gnatmetric})
17920 @node Generate project-wide metrics
17921 @subsection Generate project-wide metrics
17923 In order to compute metrics on all units of a given project, you can use
17924 the @command{gnat} driver along with the @option{-P} option:
17930 If the project @code{proj} depends upon other projects, you can compute
17931 the metrics on the project closure using the @option{-U} option:
17933 gnat metric -Pproj -U
17937 Finally, if not all the units are relevant to a particular main
17938 program in the project closure, you can generate metrics for the set
17939 of units needed to create a given main program (unit closure) using
17940 the @option{-U} option followed by the name of the main unit:
17942 gnat metric -Pproj -U main
17946 @c ***********************************
17947 @node File Name Krunching Using gnatkr
17948 @chapter File Name Krunching Using @code{gnatkr}
17952 This chapter discusses the method used by the compiler to shorten
17953 the default file names chosen for Ada units so that they do not
17954 exceed the maximum length permitted. It also describes the
17955 @code{gnatkr} utility that can be used to determine the result of
17956 applying this shortening.
17960 * Krunching Method::
17961 * Examples of gnatkr Usage::
17965 @section About @code{gnatkr}
17968 The default file naming rule in GNAT
17969 is that the file name must be derived from
17970 the unit name. The exact default rule is as follows:
17973 Take the unit name and replace all dots by hyphens.
17975 If such a replacement occurs in the
17976 second character position of a name, and the first character is
17977 ^@samp{a}, @samp{g}, @samp{s}, or @samp{i}, ^@samp{A}, @samp{G}, @samp{S}, or @samp{I},^
17978 then replace the dot by the character
17979 ^@samp{~} (tilde)^@samp{$} (dollar sign)^
17980 instead of a minus.
17982 The reason for this exception is to avoid clashes
17983 with the standard names for children of System, Ada, Interfaces,
17984 and GNAT, which use the prefixes
17985 ^@samp{s-}, @samp{a-}, @samp{i-}, and @samp{g-},^@samp{S-}, @samp{A-}, @samp{I-}, and @samp{G-},^
17988 The @option{^-gnatk^/FILE_NAME_MAX_LENGTH=^@var{nn}}
17989 switch of the compiler activates a ``krunching''
17990 circuit that limits file names to nn characters (where nn is a decimal
17991 integer). For example, using OpenVMS,
17992 where the maximum file name length is
17993 39, the value of nn is usually set to 39, but if you want to generate
17994 a set of files that would be usable if ported to a system with some
17995 different maximum file length, then a different value can be specified.
17996 The default value of 39 for OpenVMS need not be specified.
17998 The @code{gnatkr} utility can be used to determine the krunched name for
17999 a given file, when krunched to a specified maximum length.
18002 @section Using @code{gnatkr}
18005 The @code{gnatkr} command has the form
18009 $ gnatkr @var{name} @ovar{length}
18015 $ gnatkr @var{name} /COUNT=nn
18020 @var{name} is the uncrunched file name, derived from the name of the unit
18021 in the standard manner described in the previous section (i.e., in particular
18022 all dots are replaced by hyphens). The file name may or may not have an
18023 extension (defined as a suffix of the form period followed by arbitrary
18024 characters other than period). If an extension is present then it will
18025 be preserved in the output. For example, when krunching @file{hellofile.ads}
18026 to eight characters, the result will be hellofil.ads.
18028 Note: for compatibility with previous versions of @code{gnatkr} dots may
18029 appear in the name instead of hyphens, but the last dot will always be
18030 taken as the start of an extension. So if @code{gnatkr} is given an argument
18031 such as @file{Hello.World.adb} it will be treated exactly as if the first
18032 period had been a hyphen, and for example krunching to eight characters
18033 gives the result @file{hellworl.adb}.
18035 Note that the result is always all lower case (except on OpenVMS where it is
18036 all upper case). Characters of the other case are folded as required.
18038 @var{length} represents the length of the krunched name. The default
18039 when no argument is given is ^8^39^ characters. A length of zero stands for
18040 unlimited, in other words do not chop except for system files where the
18041 implied crunching length is always eight characters.
18044 The output is the krunched name. The output has an extension only if the
18045 original argument was a file name with an extension.
18047 @node Krunching Method
18048 @section Krunching Method
18051 The initial file name is determined by the name of the unit that the file
18052 contains. The name is formed by taking the full expanded name of the
18053 unit and replacing the separating dots with hyphens and
18054 using ^lowercase^uppercase^
18055 for all letters, except that a hyphen in the second character position is
18056 replaced by a ^tilde^dollar sign^ if the first character is
18057 ^@samp{a}, @samp{i}, @samp{g}, or @samp{s}^@samp{A}, @samp{I}, @samp{G}, or @samp{S}^.
18058 The extension is @code{.ads} for a
18059 spec and @code{.adb} for a body.
18060 Krunching does not affect the extension, but the file name is shortened to
18061 the specified length by following these rules:
18065 The name is divided into segments separated by hyphens, tildes or
18066 underscores and all hyphens, tildes, and underscores are
18067 eliminated. If this leaves the name short enough, we are done.
18070 If the name is too long, the longest segment is located (left-most
18071 if there are two of equal length), and shortened by dropping
18072 its last character. This is repeated until the name is short enough.
18074 As an example, consider the krunching of @*@file{our-strings-wide_fixed.adb}
18075 to fit the name into 8 characters as required by some operating systems.
18078 our-strings-wide_fixed 22
18079 our strings wide fixed 19
18080 our string wide fixed 18
18081 our strin wide fixed 17
18082 our stri wide fixed 16
18083 our stri wide fixe 15
18084 our str wide fixe 14
18085 our str wid fixe 13
18091 Final file name: oustwifi.adb
18095 The file names for all predefined units are always krunched to eight
18096 characters. The krunching of these predefined units uses the following
18097 special prefix replacements:
18101 replaced by @file{^a^A^-}
18104 replaced by @file{^g^G^-}
18107 replaced by @file{^i^I^-}
18110 replaced by @file{^s^S^-}
18113 These system files have a hyphen in the second character position. That
18114 is why normal user files replace such a character with a
18115 ^tilde^dollar sign^, to
18116 avoid confusion with system file names.
18118 As an example of this special rule, consider
18119 @*@file{ada-strings-wide_fixed.adb}, which gets krunched as follows:
18122 ada-strings-wide_fixed 22
18123 a- strings wide fixed 18
18124 a- string wide fixed 17
18125 a- strin wide fixed 16
18126 a- stri wide fixed 15
18127 a- stri wide fixe 14
18128 a- str wide fixe 13
18134 Final file name: a-stwifi.adb
18138 Of course no file shortening algorithm can guarantee uniqueness over all
18139 possible unit names, and if file name krunching is used then it is your
18140 responsibility to ensure that no name clashes occur. The utility
18141 program @code{gnatkr} is supplied for conveniently determining the
18142 krunched name of a file.
18144 @node Examples of gnatkr Usage
18145 @section Examples of @code{gnatkr} Usage
18152 $ gnatkr very_long_unit_name.ads --> velounna.ads
18153 $ gnatkr grandparent-parent-child.ads --> grparchi.ads
18154 $ gnatkr Grandparent.Parent.Child.ads --> grparchi.ads
18155 $ gnatkr grandparent-parent-child --> grparchi
18157 $ gnatkr very_long_unit_name.ads/count=6 --> vlunna.ads
18158 $ gnatkr very_long_unit_name.ads/count=0 --> very_long_unit_name.ads
18161 @node Preprocessing Using gnatprep
18162 @chapter Preprocessing Using @code{gnatprep}
18166 This chapter discusses how to use GNAT's @code{gnatprep} utility for simple
18168 Although designed for use with GNAT, @code{gnatprep} does not depend on any
18169 special GNAT features.
18170 For further discussion of conditional compilation in general, see
18171 @ref{Conditional Compilation}.
18174 * Preprocessing Symbols::
18176 * Switches for gnatprep::
18177 * Form of Definitions File::
18178 * Form of Input Text for gnatprep::
18181 @node Preprocessing Symbols
18182 @section Preprocessing Symbols
18185 Preprocessing symbols are defined in definition files and referred to in
18186 sources to be preprocessed. A Preprocessing symbol is an identifier, following
18187 normal Ada (case-insensitive) rules for its syntax, with the restriction that
18188 all characters need to be in the ASCII set (no accented letters).
18190 @node Using gnatprep
18191 @section Using @code{gnatprep}
18194 To call @code{gnatprep} use
18197 $ gnatprep @ovar{switches} @var{infile} @var{outfile} @ovar{deffile}
18204 is an optional sequence of switches as described in the next section.
18207 is the full name of the input file, which is an Ada source
18208 file containing preprocessor directives.
18211 is the full name of the output file, which is an Ada source
18212 in standard Ada form. When used with GNAT, this file name will
18213 normally have an ads or adb suffix.
18216 is the full name of a text file containing definitions of
18217 preprocessing symbols to be referenced by the preprocessor. This argument is
18218 optional, and can be replaced by the use of the @option{-D} switch.
18222 @node Switches for gnatprep
18223 @section Switches for @code{gnatprep}
18228 @item ^-b^/BLANK_LINES^
18229 @cindex @option{^-b^/BLANK_LINES^} (@command{gnatprep})
18230 Causes both preprocessor lines and the lines deleted by
18231 preprocessing to be replaced by blank lines in the output source file,
18232 preserving line numbers in the output file.
18234 @item ^-c^/COMMENTS^
18235 @cindex @option{^-c^/COMMENTS^} (@command{gnatprep})
18236 Causes both preprocessor lines and the lines deleted
18237 by preprocessing to be retained in the output source as comments marked
18238 with the special string @code{"--! "}. This option will result in line numbers
18239 being preserved in the output file.
18241 @item ^-C^/REPLACE_IN_COMMENTS^
18242 @cindex @option{^-C^/REPLACE_IN_COMMENTS^} (@command{gnatprep})
18243 Causes comments to be scanned. Normally comments are ignored by gnatprep.
18244 If this option is specified, then comments are scanned and any $symbol
18245 substitutions performed as in program text. This is particularly useful
18246 when structured comments are used (e.g., when writing programs in the
18247 SPARK dialect of Ada). Note that this switch is not available when
18248 doing integrated preprocessing (it would be useless in this context
18249 since comments are ignored by the compiler in any case).
18251 @item ^-Dsymbol=value^/ASSOCIATE="symbol=value"^
18252 @cindex @option{^-D^/ASSOCIATE^} (@command{gnatprep})
18253 Defines a new preprocessing symbol, associated with value. If no value is given
18254 on the command line, then symbol is considered to be @code{True}. This switch
18255 can be used in place of a definition file.
18259 @cindex @option{/REMOVE} (@command{gnatprep})
18260 This is the default setting which causes lines deleted by preprocessing
18261 to be entirely removed from the output file.
18264 @item ^-r^/REFERENCE^
18265 @cindex @option{^-r^/REFERENCE^} (@command{gnatprep})
18266 Causes a @code{Source_Reference} pragma to be generated that
18267 references the original input file, so that error messages will use
18268 the file name of this original file. The use of this switch implies
18269 that preprocessor lines are not to be removed from the file, so its
18270 use will force @option{^-b^/BLANK_LINES^} mode if
18271 @option{^-c^/COMMENTS^}
18272 has not been specified explicitly.
18274 Note that if the file to be preprocessed contains multiple units, then
18275 it will be necessary to @code{gnatchop} the output file from
18276 @code{gnatprep}. If a @code{Source_Reference} pragma is present
18277 in the preprocessed file, it will be respected by
18278 @code{gnatchop ^-r^/REFERENCE^}
18279 so that the final chopped files will correctly refer to the original
18280 input source file for @code{gnatprep}.
18282 @item ^-s^/SYMBOLS^
18283 @cindex @option{^-s^/SYMBOLS^} (@command{gnatprep})
18284 Causes a sorted list of symbol names and values to be
18285 listed on the standard output file.
18287 @item ^-u^/UNDEFINED^
18288 @cindex @option{^-u^/UNDEFINED^} (@command{gnatprep})
18289 Causes undefined symbols to be treated as having the value FALSE in the context
18290 of a preprocessor test. In the absence of this option, an undefined symbol in
18291 a @code{#if} or @code{#elsif} test will be treated as an error.
18297 Note: if neither @option{-b} nor @option{-c} is present,
18298 then preprocessor lines and
18299 deleted lines are completely removed from the output, unless -r is
18300 specified, in which case -b is assumed.
18303 @node Form of Definitions File
18304 @section Form of Definitions File
18307 The definitions file contains lines of the form
18314 where symbol is a preprocessing symbol, and value is one of the following:
18318 Empty, corresponding to a null substitution
18320 A string literal using normal Ada syntax
18322 Any sequence of characters from the set
18323 (letters, digits, period, underline).
18327 Comment lines may also appear in the definitions file, starting with
18328 the usual @code{--},
18329 and comments may be added to the definitions lines.
18331 @node Form of Input Text for gnatprep
18332 @section Form of Input Text for @code{gnatprep}
18335 The input text may contain preprocessor conditional inclusion lines,
18336 as well as general symbol substitution sequences.
18338 The preprocessor conditional inclusion commands have the form
18343 #if @i{expression} @r{[}then@r{]}
18345 #elsif @i{expression} @r{[}then@r{]}
18347 #elsif @i{expression} @r{[}then@r{]}
18358 In this example, @i{expression} is defined by the following grammar:
18360 @i{expression} ::= <symbol>
18361 @i{expression} ::= <symbol> = "<value>"
18362 @i{expression} ::= <symbol> = <symbol>
18363 @i{expression} ::= <symbol> 'Defined
18364 @i{expression} ::= not @i{expression}
18365 @i{expression} ::= @i{expression} and @i{expression}
18366 @i{expression} ::= @i{expression} or @i{expression}
18367 @i{expression} ::= @i{expression} and then @i{expression}
18368 @i{expression} ::= @i{expression} or else @i{expression}
18369 @i{expression} ::= ( @i{expression} )
18372 The following restriction exists: it is not allowed to have "and" or "or"
18373 following "not" in the same expression without parentheses. For example, this
18380 This should be one of the following:
18388 For the first test (@i{expression} ::= <symbol>) the symbol must have
18389 either the value true or false, that is to say the right-hand of the
18390 symbol definition must be one of the (case-insensitive) literals
18391 @code{True} or @code{False}. If the value is true, then the
18392 corresponding lines are included, and if the value is false, they are
18395 The test (@i{expression} ::= <symbol> @code{'Defined}) is true only if
18396 the symbol has been defined in the definition file or by a @option{-D}
18397 switch on the command line. Otherwise, the test is false.
18399 The equality tests are case insensitive, as are all the preprocessor lines.
18401 If the symbol referenced is not defined in the symbol definitions file,
18402 then the effect depends on whether or not switch @option{-u}
18403 is specified. If so, then the symbol is treated as if it had the value
18404 false and the test fails. If this switch is not specified, then
18405 it is an error to reference an undefined symbol. It is also an error to
18406 reference a symbol that is defined with a value other than @code{True}
18409 The use of the @code{not} operator inverts the sense of this logical test.
18410 The @code{not} operator cannot be combined with the @code{or} or @code{and}
18411 operators, without parentheses. For example, "if not X or Y then" is not
18412 allowed, but "if (not X) or Y then" and "if not (X or Y) then" are.
18414 The @code{then} keyword is optional as shown
18416 The @code{#} must be the first non-blank character on a line, but
18417 otherwise the format is free form. Spaces or tabs may appear between
18418 the @code{#} and the keyword. The keywords and the symbols are case
18419 insensitive as in normal Ada code. Comments may be used on a
18420 preprocessor line, but other than that, no other tokens may appear on a
18421 preprocessor line. Any number of @code{elsif} clauses can be present,
18422 including none at all. The @code{else} is optional, as in Ada.
18424 The @code{#} marking the start of a preprocessor line must be the first
18425 non-blank character on the line, i.e., it must be preceded only by
18426 spaces or horizontal tabs.
18428 Symbol substitution outside of preprocessor lines is obtained by using
18436 anywhere within a source line, except in a comment or within a
18437 string literal. The identifier
18438 following the @code{$} must match one of the symbols defined in the symbol
18439 definition file, and the result is to substitute the value of the
18440 symbol in place of @code{$symbol} in the output file.
18442 Note that although the substitution of strings within a string literal
18443 is not possible, it is possible to have a symbol whose defined value is
18444 a string literal. So instead of setting XYZ to @code{hello} and writing:
18447 Header : String := "$XYZ";
18451 you should set XYZ to @code{"hello"} and write:
18454 Header : String := $XYZ;
18458 and then the substitution will occur as desired.
18461 @node The GNAT Run-Time Library Builder gnatlbr
18462 @chapter The GNAT Run-Time Library Builder @code{gnatlbr}
18464 @cindex Library builder
18467 @code{gnatlbr} is a tool for rebuilding the GNAT run time with user
18468 supplied configuration pragmas.
18471 * Running gnatlbr::
18472 * Switches for gnatlbr::
18473 * Examples of gnatlbr Usage::
18476 @node Running gnatlbr
18477 @section Running @code{gnatlbr}
18480 The @code{gnatlbr} command has the form
18483 $ GNAT LIBRARY /@r{[}CREATE@r{|}SET@r{|}DELETE@r{]}=directory @r{[}/CONFIG=file@r{]}
18486 @node Switches for gnatlbr
18487 @section Switches for @code{gnatlbr}
18490 @code{gnatlbr} recognizes the following switches:
18494 @item /CREATE=directory
18495 @cindex @code{/CREATE} (@code{gnatlbr})
18496 Create the new run-time library in the specified directory.
18498 @item /SET=directory
18499 @cindex @code{/SET} (@code{gnatlbr})
18500 Make the library in the specified directory the current run-time library.
18502 @item /DELETE=directory
18503 @cindex @code{/DELETE} (@code{gnatlbr})
18504 Delete the run-time library in the specified directory.
18507 @cindex @code{/CONFIG} (@code{gnatlbr})
18508 With /CREATE: Use the configuration pragmas in the specified file when
18509 building the library.
18511 With /SET: Use the configuration pragmas in the specified file when
18516 @node Examples of gnatlbr Usage
18517 @section Example of @code{gnatlbr} Usage
18520 Contents of VAXFLOAT.ADC:
18521 pragma Float_Representation (VAX_Float);
18523 $ GNAT LIBRARY /CREATE=[.VAXFLOAT] /CONFIG=VAXFLOAT.ADC
18525 GNAT LIBRARY rebuilds the run-time library in directory [.VAXFLOAT]
18530 @node The GNAT Library Browser gnatls
18531 @chapter The GNAT Library Browser @code{gnatls}
18533 @cindex Library browser
18536 @code{gnatls} is a tool that outputs information about compiled
18537 units. It gives the relationship between objects, unit names and source
18538 files. It can also be used to check the source dependencies of a unit
18539 as well as various characteristics.
18541 Note: to invoke @code{gnatls} with a project file, use the @code{gnat}
18542 driver (see @ref{The GNAT Driver and Project Files}).
18546 * Switches for gnatls::
18547 * Examples of gnatls Usage::
18550 @node Running gnatls
18551 @section Running @code{gnatls}
18554 The @code{gnatls} command has the form
18557 $ gnatls switches @var{object_or_ali_file}
18561 The main argument is the list of object or @file{ali} files
18562 (@pxref{The Ada Library Information Files})
18563 for which information is requested.
18565 In normal mode, without additional option, @code{gnatls} produces a
18566 four-column listing. Each line represents information for a specific
18567 object. The first column gives the full path of the object, the second
18568 column gives the name of the principal unit in this object, the third
18569 column gives the status of the source and the fourth column gives the
18570 full path of the source representing this unit.
18571 Here is a simple example of use:
18575 ^./^[]^demo1.o demo1 DIF demo1.adb
18576 ^./^[]^demo2.o demo2 OK demo2.adb
18577 ^./^[]^hello.o h1 OK hello.adb
18578 ^./^[]^instr-child.o instr.child MOK instr-child.adb
18579 ^./^[]^instr.o instr OK instr.adb
18580 ^./^[]^tef.o tef DIF tef.adb
18581 ^./^[]^text_io_example.o text_io_example OK text_io_example.adb
18582 ^./^[]^tgef.o tgef DIF tgef.adb
18586 The first line can be interpreted as follows: the main unit which is
18588 object file @file{demo1.o} is demo1, whose main source is in
18589 @file{demo1.adb}. Furthermore, the version of the source used for the
18590 compilation of demo1 has been modified (DIF). Each source file has a status
18591 qualifier which can be:
18594 @item OK (unchanged)
18595 The version of the source file used for the compilation of the
18596 specified unit corresponds exactly to the actual source file.
18598 @item MOK (slightly modified)
18599 The version of the source file used for the compilation of the
18600 specified unit differs from the actual source file but not enough to
18601 require recompilation. If you use gnatmake with the qualifier
18602 @option{^-m (minimal recompilation)^/MINIMAL_RECOMPILATION^}, a file marked
18603 MOK will not be recompiled.
18605 @item DIF (modified)
18606 No version of the source found on the path corresponds to the source
18607 used to build this object.
18609 @item ??? (file not found)
18610 No source file was found for this unit.
18612 @item HID (hidden, unchanged version not first on PATH)
18613 The version of the source that corresponds exactly to the source used
18614 for compilation has been found on the path but it is hidden by another
18615 version of the same source that has been modified.
18619 @node Switches for gnatls
18620 @section Switches for @code{gnatls}
18623 @code{gnatls} recognizes the following switches:
18627 @cindex @option{--version} @command{gnatls}
18628 Display Copyright and version, then exit disregarding all other options.
18631 @cindex @option{--help} @command{gnatls}
18632 If @option{--version} was not used, display usage, then exit disregarding
18635 @item ^-a^/ALL_UNITS^
18636 @cindex @option{^-a^/ALL_UNITS^} (@code{gnatls})
18637 Consider all units, including those of the predefined Ada library.
18638 Especially useful with @option{^-d^/DEPENDENCIES^}.
18640 @item ^-d^/DEPENDENCIES^
18641 @cindex @option{^-d^/DEPENDENCIES^} (@code{gnatls})
18642 List sources from which specified units depend on.
18644 @item ^-h^/OUTPUT=OPTIONS^
18645 @cindex @option{^-h^/OUTPUT=OPTIONS^} (@code{gnatls})
18646 Output the list of options.
18648 @item ^-o^/OUTPUT=OBJECTS^
18649 @cindex @option{^-o^/OUTPUT=OBJECTS^} (@code{gnatls})
18650 Only output information about object files.
18652 @item ^-s^/OUTPUT=SOURCES^
18653 @cindex @option{^-s^/OUTPUT=SOURCES^} (@code{gnatls})
18654 Only output information about source files.
18656 @item ^-u^/OUTPUT=UNITS^
18657 @cindex @option{^-u^/OUTPUT=UNITS^} (@code{gnatls})
18658 Only output information about compilation units.
18660 @item ^-files^/FILES^=@var{file}
18661 @cindex @option{^-files^/FILES^} (@code{gnatls})
18662 Take as arguments the files listed in text file @var{file}.
18663 Text file @var{file} may contain empty lines that are ignored.
18664 Each nonempty line should contain the name of an existing file.
18665 Several such switches may be specified simultaneously.
18667 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18668 @itemx ^-aI^/SOURCE_SEARCH=^@var{dir}
18669 @itemx ^-I^/SEARCH=^@var{dir}
18670 @itemx ^-I-^/NOCURRENT_DIRECTORY^
18672 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatls})
18673 @cindex @option{^-aI^/SOURCE_SEARCH^} (@code{gnatls})
18674 @cindex @option{^-I^/SEARCH^} (@code{gnatls})
18675 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatls})
18676 Source path manipulation. Same meaning as the equivalent @command{gnatmake}
18677 flags (@pxref{Switches for gnatmake}).
18679 @item --RTS=@var{rts-path}
18680 @cindex @option{--RTS} (@code{gnatls})
18681 Specifies the default location of the runtime library. Same meaning as the
18682 equivalent @command{gnatmake} flag (@pxref{Switches for gnatmake}).
18684 @item ^-v^/OUTPUT=VERBOSE^
18685 @cindex @option{^-v^/OUTPUT=VERBOSE^} (@code{gnatls})
18686 Verbose mode. Output the complete source, object and project paths. Do not use
18687 the default column layout but instead use long format giving as much as
18688 information possible on each requested units, including special
18689 characteristics such as:
18692 @item Preelaborable
18693 The unit is preelaborable in the Ada sense.
18696 No elaboration code has been produced by the compiler for this unit.
18699 The unit is pure in the Ada sense.
18701 @item Elaborate_Body
18702 The unit contains a pragma Elaborate_Body.
18705 The unit contains a pragma Remote_Types.
18707 @item Shared_Passive
18708 The unit contains a pragma Shared_Passive.
18711 This unit is part of the predefined environment and cannot be modified
18714 @item Remote_Call_Interface
18715 The unit contains a pragma Remote_Call_Interface.
18721 @node Examples of gnatls Usage
18722 @section Example of @code{gnatls} Usage
18726 Example of using the verbose switch. Note how the source and
18727 object paths are affected by the -I switch.
18730 $ gnatls -v -I.. demo1.o
18732 GNATLS 5.03w (20041123-34)
18733 Copyright 1997-2004 Free Software Foundation, Inc.
18735 Source Search Path:
18736 <Current_Directory>
18738 /home/comar/local/adainclude/
18740 Object Search Path:
18741 <Current_Directory>
18743 /home/comar/local/lib/gcc-lib/x86-linux/3.4.3/adalib/
18745 Project Search Path:
18746 <Current_Directory>
18747 /home/comar/local/lib/gnat/
18752 Kind => subprogram body
18753 Flags => No_Elab_Code
18754 Source => demo1.adb modified
18758 The following is an example of use of the dependency list.
18759 Note the use of the -s switch
18760 which gives a straight list of source files. This can be useful for
18761 building specialized scripts.
18764 $ gnatls -d demo2.o
18765 ./demo2.o demo2 OK demo2.adb
18771 $ gnatls -d -s -a demo1.o
18773 /home/comar/local/adainclude/ada.ads
18774 /home/comar/local/adainclude/a-finali.ads
18775 /home/comar/local/adainclude/a-filico.ads
18776 /home/comar/local/adainclude/a-stream.ads
18777 /home/comar/local/adainclude/a-tags.ads
18780 /home/comar/local/adainclude/gnat.ads
18781 /home/comar/local/adainclude/g-io.ads
18783 /home/comar/local/adainclude/system.ads
18784 /home/comar/local/adainclude/s-exctab.ads
18785 /home/comar/local/adainclude/s-finimp.ads
18786 /home/comar/local/adainclude/s-finroo.ads
18787 /home/comar/local/adainclude/s-secsta.ads
18788 /home/comar/local/adainclude/s-stalib.ads
18789 /home/comar/local/adainclude/s-stoele.ads
18790 /home/comar/local/adainclude/s-stratt.ads
18791 /home/comar/local/adainclude/s-tasoli.ads
18792 /home/comar/local/adainclude/s-unstyp.ads
18793 /home/comar/local/adainclude/unchconv.ads
18799 GNAT LIST /DEPENDENCIES /OUTPUT=SOURCES /ALL_UNITS DEMO1.ADB
18801 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]ada.ads
18802 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-finali.ads
18803 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-filico.ads
18804 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-stream.ads
18805 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]a-tags.ads
18809 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]gnat.ads
18810 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]g-io.ads
18812 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]system.ads
18813 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-exctab.ads
18814 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finimp.ads
18815 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-finroo.ads
18816 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-secsta.ads
18817 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stalib.ads
18818 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stoele.ads
18819 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-stratt.ads
18820 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-tasoli.ads
18821 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]s-unstyp.ads
18822 GNU:[LIB.OPENVMS7_1.2_8_1.ADALIB]unchconv.ads
18826 @node Cleaning Up Using gnatclean
18827 @chapter Cleaning Up Using @code{gnatclean}
18829 @cindex Cleaning tool
18832 @code{gnatclean} is a tool that allows the deletion of files produced by the
18833 compiler, binder and linker, including ALI files, object files, tree files,
18834 expanded source files, library files, interface copy source files, binder
18835 generated files and executable files.
18838 * Running gnatclean::
18839 * Switches for gnatclean::
18840 @c * Examples of gnatclean Usage::
18843 @node Running gnatclean
18844 @section Running @code{gnatclean}
18847 The @code{gnatclean} command has the form:
18850 $ gnatclean switches @var{names}
18854 @var{names} is a list of source file names. Suffixes @code{.^ads^ADS^} and
18855 @code{^adb^ADB^} may be omitted. If a project file is specified using switch
18856 @code{^-P^/PROJECT_FILE=^}, then @var{names} may be completely omitted.
18859 In normal mode, @code{gnatclean} delete the files produced by the compiler and,
18860 if switch @code{^-c^/COMPILER_FILES_ONLY^} is not specified, by the binder and
18861 the linker. In informative-only mode, specified by switch
18862 @code{^-n^/NODELETE^}, the list of files that would have been deleted in
18863 normal mode is listed, but no file is actually deleted.
18865 @node Switches for gnatclean
18866 @section Switches for @code{gnatclean}
18869 @code{gnatclean} recognizes the following switches:
18873 @cindex @option{--version} @command{gnatclean}
18874 Display Copyright and version, then exit disregarding all other options.
18877 @cindex @option{--help} @command{gnatclean}
18878 If @option{--version} was not used, display usage, then exit disregarding
18881 @item ^-c^/COMPILER_FILES_ONLY^
18882 @cindex @option{^-c^/COMPILER_FILES_ONLY^} (@code{gnatclean})
18883 Only attempt to delete the files produced by the compiler, not those produced
18884 by the binder or the linker. The files that are not to be deleted are library
18885 files, interface copy files, binder generated files and executable files.
18887 @item ^-D ^/DIRECTORY_OBJECTS=^@var{dir}
18888 @cindex @option{^-D^/DIRECTORY_OBJECTS^} (@code{gnatclean})
18889 Indicate that ALI and object files should normally be found in directory
18892 @item ^-F^/FULL_PATH_IN_BRIEF_MESSAGES^
18893 @cindex @option{^-F^/FULL_PATH_IN_BRIEF_MESSAGES^} (@code{gnatclean})
18894 When using project files, if some errors or warnings are detected during
18895 parsing and verbose mode is not in effect (no use of switch
18896 ^-v^/VERBOSE^), then error lines start with the full path name of the project
18897 file, rather than its simple file name.
18900 @cindex @option{^-h^/HELP^} (@code{gnatclean})
18901 Output a message explaining the usage of @code{^gnatclean^gnatclean^}.
18903 @item ^-n^/NODELETE^
18904 @cindex @option{^-n^/NODELETE^} (@code{gnatclean})
18905 Informative-only mode. Do not delete any files. Output the list of the files
18906 that would have been deleted if this switch was not specified.
18908 @item ^-P^/PROJECT_FILE=^@var{project}
18909 @cindex @option{^-P^/PROJECT_FILE^} (@code{gnatclean})
18910 Use project file @var{project}. Only one such switch can be used.
18911 When cleaning a project file, the files produced by the compilation of the
18912 immediate sources or inherited sources of the project files are to be
18913 deleted. This is not depending on the presence or not of executable names
18914 on the command line.
18917 @cindex @option{^-q^/QUIET^} (@code{gnatclean})
18918 Quiet output. If there are no errors, do not output anything, except in
18919 verbose mode (switch ^-v^/VERBOSE^) or in informative-only mode
18920 (switch ^-n^/NODELETE^).
18922 @item ^-r^/RECURSIVE^
18923 @cindex @option{^-r^/RECURSIVE^} (@code{gnatclean})
18924 When a project file is specified (using switch ^-P^/PROJECT_FILE=^),
18925 clean all imported and extended project files, recursively. If this switch
18926 is not specified, only the files related to the main project file are to be
18927 deleted. This switch has no effect if no project file is specified.
18929 @item ^-v^/VERBOSE^
18930 @cindex @option{^-v^/VERBOSE^} (@code{gnatclean})
18933 @item ^-vP^/MESSAGES_PROJECT_FILE=^@emph{x}
18934 @cindex @option{^-vP^/MESSAGES_PROJECT_FILE^} (@code{gnatclean})
18935 Indicates the verbosity of the parsing of GNAT project files.
18936 @xref{Switches Related to Project Files}.
18938 @item ^-X^/EXTERNAL_REFERENCE=^@var{name=value}
18939 @cindex @option{^-X^/EXTERNAL_REFERENCE^} (@code{gnatclean})
18940 Indicates that external variable @var{name} has the value @var{value}.
18941 The Project Manager will use this value for occurrences of
18942 @code{external(name)} when parsing the project file.
18943 @xref{Switches Related to Project Files}.
18945 @item ^-aO^/OBJECT_SEARCH=^@var{dir}
18946 @cindex @option{^-aO^/OBJECT_SEARCH^} (@code{gnatclean})
18947 When searching for ALI and object files, look in directory
18950 @item ^-I^/SEARCH=^@var{dir}
18951 @cindex @option{^-I^/SEARCH^} (@code{gnatclean})
18952 Equivalent to @option{^-aO^/OBJECT_SEARCH=^@var{dir}}.
18954 @item ^-I-^/NOCURRENT_DIRECTORY^
18955 @cindex @option{^-I-^/NOCURRENT_DIRECTORY^} (@code{gnatclean})
18956 @cindex Source files, suppressing search
18957 Do not look for ALI or object files in the directory
18958 where @code{gnatclean} was invoked.
18962 @c @node Examples of gnatclean Usage
18963 @c @section Examples of @code{gnatclean} Usage
18966 @node GNAT and Libraries
18967 @chapter GNAT and Libraries
18968 @cindex Library, building, installing, using
18971 This chapter describes how to build and use libraries with GNAT, and also shows
18972 how to recompile the GNAT run-time library. You should be familiar with the
18973 Project Manager facility (@pxref{GNAT Project Manager}) before reading this
18977 * Introduction to Libraries in GNAT::
18978 * General Ada Libraries::
18979 * Stand-alone Ada Libraries::
18980 * Rebuilding the GNAT Run-Time Library::
18983 @node Introduction to Libraries in GNAT
18984 @section Introduction to Libraries in GNAT
18987 A library is, conceptually, a collection of objects which does not have its
18988 own main thread of execution, but rather provides certain services to the
18989 applications that use it. A library can be either statically linked with the
18990 application, in which case its code is directly included in the application,
18991 or, on platforms that support it, be dynamically linked, in which case
18992 its code is shared by all applications making use of this library.
18994 GNAT supports both types of libraries.
18995 In the static case, the compiled code can be provided in different ways. The
18996 simplest approach is to provide directly the set of objects resulting from
18997 compilation of the library source files. Alternatively, you can group the
18998 objects into an archive using whatever commands are provided by the operating
18999 system. For the latter case, the objects are grouped into a shared library.
19001 In the GNAT environment, a library has three types of components:
19007 @xref{The Ada Library Information Files}.
19009 Object files, an archive or a shared library.
19013 A GNAT library may expose all its source files, which is useful for
19014 documentation purposes. Alternatively, it may expose only the units needed by
19015 an external user to make use of the library. That is to say, the specs
19016 reflecting the library services along with all the units needed to compile
19017 those specs, which can include generic bodies or any body implementing an
19018 inlined routine. In the case of @emph{stand-alone libraries} those exposed
19019 units are called @emph{interface units} (@pxref{Stand-alone Ada Libraries}).
19021 All compilation units comprising an application, including those in a library,
19022 need to be elaborated in an order partially defined by Ada's semantics. GNAT
19023 computes the elaboration order from the @file{ALI} files and this is why they
19024 constitute a mandatory part of GNAT libraries.
19025 @emph{Stand-alone libraries} are the exception to this rule because a specific
19026 library elaboration routine is produced independently of the application(s)
19029 @node General Ada Libraries
19030 @section General Ada Libraries
19033 * Building a library::
19034 * Installing a library::
19035 * Using a library::
19038 @node Building a library
19039 @subsection Building a library
19042 The easiest way to build a library is to use the Project Manager,
19043 which supports a special type of project called a @emph{Library Project}
19044 (@pxref{Library Projects}).
19046 A project is considered a library project, when two project-level attributes
19047 are defined in it: @code{Library_Name} and @code{Library_Dir}. In order to
19048 control different aspects of library configuration, additional optional
19049 project-level attributes can be specified:
19052 This attribute controls whether the library is to be static or dynamic
19054 @item Library_Version
19055 This attribute specifies the library version; this value is used
19056 during dynamic linking of shared libraries to determine if the currently
19057 installed versions of the binaries are compatible.
19059 @item Library_Options
19061 These attributes specify additional low-level options to be used during
19062 library generation, and redefine the actual application used to generate
19067 The GNAT Project Manager takes full care of the library maintenance task,
19068 including recompilation of the source files for which objects do not exist
19069 or are not up to date, assembly of the library archive, and installation of
19070 the library (i.e., copying associated source, object and @file{ALI} files
19071 to the specified location).
19073 Here is a simple library project file:
19074 @smallexample @c ada
19076 for Source_Dirs use ("src1", "src2");
19077 for Object_Dir use "obj";
19078 for Library_Name use "mylib";
19079 for Library_Dir use "lib";
19080 for Library_Kind use "dynamic";
19085 and the compilation command to build and install the library:
19087 @smallexample @c ada
19088 $ gnatmake -Pmy_lib
19092 It is not entirely trivial to perform manually all the steps required to
19093 produce a library. We recommend that you use the GNAT Project Manager
19094 for this task. In special cases where this is not desired, the necessary
19095 steps are discussed below.
19097 There are various possibilities for compiling the units that make up the
19098 library: for example with a Makefile (@pxref{Using the GNU make Utility}) or
19099 with a conventional script. For simple libraries, it is also possible to create
19100 a dummy main program which depends upon all the packages that comprise the
19101 interface of the library. This dummy main program can then be given to
19102 @command{gnatmake}, which will ensure that all necessary objects are built.
19104 After this task is accomplished, you should follow the standard procedure
19105 of the underlying operating system to produce the static or shared library.
19107 Here is an example of such a dummy program:
19108 @smallexample @c ada
19110 with My_Lib.Service1;
19111 with My_Lib.Service2;
19112 with My_Lib.Service3;
19113 procedure My_Lib_Dummy is
19121 Here are the generic commands that will build an archive or a shared library.
19124 # compiling the library
19125 $ gnatmake -c my_lib_dummy.adb
19127 # we don't need the dummy object itself
19128 $ rm my_lib_dummy.o my_lib_dummy.ali
19130 # create an archive with the remaining objects
19131 $ ar rc libmy_lib.a *.o
19132 # some systems may require "ranlib" to be run as well
19134 # or create a shared library
19135 $ gcc -shared -o libmy_lib.so *.o
19136 # some systems may require the code to have been compiled with -fPIC
19138 # remove the object files that are now in the library
19141 # Make the ALI files read-only so that gnatmake will not try to
19142 # regenerate the objects that are in the library
19147 Please note that the library must have a name of the form @file{lib@var{xxx}.a}
19148 or @file{lib@var{xxx}.so} (or @file{lib@var{xxx}.dll} on Windows) in order to
19149 be accessed by the directive @option{-l@var{xxx}} at link time.
19151 @node Installing a library
19152 @subsection Installing a library
19153 @cindex @code{ADA_PROJECT_PATH}
19154 @cindex @code{GPR_PROJECT_PATH}
19157 If you use project files, library installation is part of the library build
19158 process. Thus no further action is needed in order to make use of the
19159 libraries that are built as part of the general application build. A usable
19160 version of the library is installed in the directory specified by the
19161 @code{Library_Dir} attribute of the library project file.
19163 You may want to install a library in a context different from where the library
19164 is built. This situation arises with third party suppliers, who may want
19165 to distribute a library in binary form where the user is not expected to be
19166 able to recompile the library. The simplest option in this case is to provide
19167 a project file slightly different from the one used to build the library, by
19168 using the @code{externally_built} attribute. For instance, the project
19169 file used to build the library in the previous section can be changed into the
19170 following one when the library is installed:
19172 @smallexample @c projectfile
19174 for Source_Dirs use ("src1", "src2");
19175 for Library_Name use "mylib";
19176 for Library_Dir use "lib";
19177 for Library_Kind use "dynamic";
19178 for Externally_Built use "true";
19183 This project file assumes that the directories @file{src1},
19184 @file{src2}, and @file{lib} exist in
19185 the directory containing the project file. The @code{externally_built}
19186 attribute makes it clear to the GNAT builder that it should not attempt to
19187 recompile any of the units from this library. It allows the library provider to
19188 restrict the source set to the minimum necessary for clients to make use of the
19189 library as described in the first section of this chapter. It is the
19190 responsibility of the library provider to install the necessary sources, ALI
19191 files and libraries in the directories mentioned in the project file. For
19192 convenience, the user's library project file should be installed in a location
19193 that will be searched automatically by the GNAT
19194 builder. These are the directories referenced in the @env{GPR_PROJECT_PATH}
19195 environment variable (@pxref{Importing Projects}), and also the default GNAT
19196 library location that can be queried with @command{gnatls -v} and is usually of
19197 the form $gnat_install_root/lib/gnat.
19199 When project files are not an option, it is also possible, but not recommended,
19200 to install the library so that the sources needed to use the library are on the
19201 Ada source path and the ALI files & libraries be on the Ada Object path (see
19202 @ref{Search Paths and the Run-Time Library (RTL)}. Alternatively, the system
19203 administrator can place general-purpose libraries in the default compiler
19204 paths, by specifying the libraries' location in the configuration files
19205 @file{ada_source_path} and @file{ada_object_path}. These configuration files
19206 must be located in the GNAT installation tree at the same place as the gcc spec
19207 file. The location of the gcc spec file can be determined as follows:
19213 The configuration files mentioned above have a simple format: each line
19214 must contain one unique directory name.
19215 Those names are added to the corresponding path
19216 in their order of appearance in the file. The names can be either absolute
19217 or relative; in the latter case, they are relative to where theses files
19220 The files @file{ada_source_path} and @file{ada_object_path} might not be
19222 GNAT installation, in which case, GNAT will look for its run-time library in
19223 the directories @file{adainclude} (for the sources) and @file{adalib} (for the
19224 objects and @file{ALI} files). When the files exist, the compiler does not
19225 look in @file{adainclude} and @file{adalib}, and thus the
19226 @file{ada_source_path} file
19227 must contain the location for the GNAT run-time sources (which can simply
19228 be @file{adainclude}). In the same way, the @file{ada_object_path} file must
19229 contain the location for the GNAT run-time objects (which can simply
19232 You can also specify a new default path to the run-time library at compilation
19233 time with the switch @option{--RTS=rts-path}. You can thus choose / change
19234 the run-time library you want your program to be compiled with. This switch is
19235 recognized by @command{gcc}, @command{gnatmake}, @command{gnatbind},
19236 @command{gnatls}, @command{gnatfind} and @command{gnatxref}.
19238 It is possible to install a library before or after the standard GNAT
19239 library, by reordering the lines in the configuration files. In general, a
19240 library must be installed before the GNAT library if it redefines
19243 @node Using a library
19244 @subsection Using a library
19246 @noindent Once again, the project facility greatly simplifies the use of
19247 libraries. In this context, using a library is just a matter of adding a
19248 @code{with} clause in the user project. For instance, to make use of the
19249 library @code{My_Lib} shown in examples in earlier sections, you can
19252 @smallexample @c projectfile
19259 Even if you have a third-party, non-Ada library, you can still use GNAT's
19260 Project Manager facility to provide a wrapper for it. For example, the
19261 following project, when @code{with}ed by your main project, will link with the
19262 third-party library @file{liba.a}:
19264 @smallexample @c projectfile
19267 for Externally_Built use "true";
19268 for Source_Files use ();
19269 for Library_Dir use "lib";
19270 for Library_Name use "a";
19271 for Library_Kind use "static";
19275 This is an alternative to the use of @code{pragma Linker_Options}. It is
19276 especially interesting in the context of systems with several interdependent
19277 static libraries where finding a proper linker order is not easy and best be
19278 left to the tools having visibility over project dependence information.
19281 In order to use an Ada library manually, you need to make sure that this
19282 library is on both your source and object path
19283 (see @ref{Search Paths and the Run-Time Library (RTL)}
19284 and @ref{Search Paths for gnatbind}). Furthermore, when the objects are grouped
19285 in an archive or a shared library, you need to specify the desired
19286 library at link time.
19288 For example, you can use the library @file{mylib} installed in
19289 @file{/dir/my_lib_src} and @file{/dir/my_lib_obj} with the following commands:
19292 $ gnatmake -aI/dir/my_lib_src -aO/dir/my_lib_obj my_appl \
19297 This can be expressed more simply:
19302 when the following conditions are met:
19305 @file{/dir/my_lib_src} has been added by the user to the environment
19306 variable @env{ADA_INCLUDE_PATH}, or by the administrator to the file
19307 @file{ada_source_path}
19309 @file{/dir/my_lib_obj} has been added by the user to the environment
19310 variable @env{ADA_OBJECTS_PATH}, or by the administrator to the file
19311 @file{ada_object_path}
19313 a pragma @code{Linker_Options} has been added to one of the sources.
19316 @smallexample @c ada
19317 pragma Linker_Options ("-lmy_lib");
19321 @node Stand-alone Ada Libraries
19322 @section Stand-alone Ada Libraries
19323 @cindex Stand-alone library, building, using
19326 * Introduction to Stand-alone Libraries::
19327 * Building a Stand-alone Library::
19328 * Creating a Stand-alone Library to be used in a non-Ada context::
19329 * Restrictions in Stand-alone Libraries::
19332 @node Introduction to Stand-alone Libraries
19333 @subsection Introduction to Stand-alone Libraries
19336 A Stand-alone Library (abbreviated ``SAL'') is a library that contains the
19338 elaborate the Ada units that are included in the library. In contrast with
19339 an ordinary library, which consists of all sources, objects and @file{ALI}
19341 library, a SAL may specify a restricted subset of compilation units
19342 to serve as a library interface. In this case, the fully
19343 self-sufficient set of files will normally consist of an objects
19344 archive, the sources of interface units' specs, and the @file{ALI}
19345 files of interface units.
19346 If an interface spec contains a generic unit or an inlined subprogram,
19348 source must also be provided; if the units that must be provided in the source
19349 form depend on other units, the source and @file{ALI} files of those must
19352 The main purpose of a SAL is to minimize the recompilation overhead of client
19353 applications when a new version of the library is installed. Specifically,
19354 if the interface sources have not changed, client applications do not need to
19355 be recompiled. If, furthermore, a SAL is provided in the shared form and its
19356 version, controlled by @code{Library_Version} attribute, is not changed,
19357 then the clients do not need to be relinked.
19359 SALs also allow the library providers to minimize the amount of library source
19360 text exposed to the clients. Such ``information hiding'' might be useful or
19361 necessary for various reasons.
19363 Stand-alone libraries are also well suited to be used in an executable whose
19364 main routine is not written in Ada.
19366 @node Building a Stand-alone Library
19367 @subsection Building a Stand-alone Library
19370 GNAT's Project facility provides a simple way of building and installing
19371 stand-alone libraries; see @ref{Stand-alone Library Projects}.
19372 To be a Stand-alone Library Project, in addition to the two attributes
19373 that make a project a Library Project (@code{Library_Name} and
19374 @code{Library_Dir}; see @ref{Library Projects}), the attribute
19375 @code{Library_Interface} must be defined. For example:
19377 @smallexample @c projectfile
19379 for Library_Dir use "lib_dir";
19380 for Library_Name use "dummy";
19381 for Library_Interface use ("int1", "int1.child");
19386 Attribute @code{Library_Interface} has a non-empty string list value,
19387 each string in the list designating a unit contained in an immediate source
19388 of the project file.
19390 When a Stand-alone Library is built, first the binder is invoked to build
19391 a package whose name depends on the library name
19392 (@file{^b~dummy.ads/b^B$DUMMY.ADS/B^} in the example above).
19393 This binder-generated package includes initialization and
19394 finalization procedures whose
19395 names depend on the library name (@code{dummyinit} and @code{dummyfinal}
19397 above). The object corresponding to this package is included in the library.
19399 You must ensure timely (e.g., prior to any use of interfaces in the SAL)
19400 calling of these procedures if a static SAL is built, or if a shared SAL
19402 with the project-level attribute @code{Library_Auto_Init} set to
19405 For a Stand-Alone Library, only the @file{ALI} files of the Interface Units
19406 (those that are listed in attribute @code{Library_Interface}) are copied to
19407 the Library Directory. As a consequence, only the Interface Units may be
19408 imported from Ada units outside of the library. If other units are imported,
19409 the binding phase will fail.
19411 The attribute @code{Library_Src_Dir} may be specified for a
19412 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
19413 single string value. Its value must be the path (absolute or relative to the
19414 project directory) of an existing directory. This directory cannot be the
19415 object directory or one of the source directories, but it can be the same as
19416 the library directory. The sources of the Interface
19417 Units of the library that are needed by an Ada client of the library will be
19418 copied to the designated directory, called the Interface Copy directory.
19419 These sources include the specs of the Interface Units, but they may also
19420 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
19421 are used, or when there is a generic unit in the spec. Before the sources
19422 are copied to the Interface Copy directory, an attempt is made to delete all
19423 files in the Interface Copy directory.
19425 Building stand-alone libraries by hand is somewhat tedious, but for those
19426 occasions when it is necessary here are the steps that you need to perform:
19429 Compile all library sources.
19432 Invoke the binder with the switch @option{-n} (No Ada main program),
19433 with all the @file{ALI} files of the interfaces, and
19434 with the switch @option{-L} to give specific names to the @code{init}
19435 and @code{final} procedures. For example:
19437 gnatbind -n int1.ali int2.ali -Lsal1
19441 Compile the binder generated file:
19447 Link the dynamic library with all the necessary object files,
19448 indicating to the linker the names of the @code{init} (and possibly
19449 @code{final}) procedures for automatic initialization (and finalization).
19450 The built library should be placed in a directory different from
19451 the object directory.
19454 Copy the @code{ALI} files of the interface to the library directory,
19455 add in this copy an indication that it is an interface to a SAL
19456 (i.e., add a word @option{SL} on the line in the @file{ALI} file that starts
19457 with letter ``P'') and make the modified copy of the @file{ALI} file
19462 Using SALs is not different from using other libraries
19463 (see @ref{Using a library}).
19465 @node Creating a Stand-alone Library to be used in a non-Ada context
19466 @subsection Creating a Stand-alone Library to be used in a non-Ada context
19469 It is easy to adapt the SAL build procedure discussed above for use of a SAL in
19472 The only extra step required is to ensure that library interface subprograms
19473 are compatible with the main program, by means of @code{pragma Export}
19474 or @code{pragma Convention}.
19476 Here is an example of simple library interface for use with C main program:
19478 @smallexample @c ada
19479 package My_Package is
19481 procedure Do_Something;
19482 pragma Export (C, Do_Something, "do_something");
19484 procedure Do_Something_Else;
19485 pragma Export (C, Do_Something_Else, "do_something_else");
19491 On the foreign language side, you must provide a ``foreign'' view of the
19492 library interface; remember that it should contain elaboration routines in
19493 addition to interface subprograms.
19495 The example below shows the content of @code{mylib_interface.h} (note
19496 that there is no rule for the naming of this file, any name can be used)
19498 /* the library elaboration procedure */
19499 extern void mylibinit (void);
19501 /* the library finalization procedure */
19502 extern void mylibfinal (void);
19504 /* the interface exported by the library */
19505 extern void do_something (void);
19506 extern void do_something_else (void);
19510 Libraries built as explained above can be used from any program, provided
19511 that the elaboration procedures (named @code{mylibinit} in the previous
19512 example) are called before the library services are used. Any number of
19513 libraries can be used simultaneously, as long as the elaboration
19514 procedure of each library is called.
19516 Below is an example of a C program that uses the @code{mylib} library.
19519 #include "mylib_interface.h"
19524 /* First, elaborate the library before using it */
19527 /* Main program, using the library exported entities */
19529 do_something_else ();
19531 /* Library finalization at the end of the program */
19538 Note that invoking any library finalization procedure generated by
19539 @code{gnatbind} shuts down the Ada run-time environment.
19541 finalization of all Ada libraries must be performed at the end of the program.
19542 No call to these libraries or to the Ada run-time library should be made
19543 after the finalization phase.
19545 @node Restrictions in Stand-alone Libraries
19546 @subsection Restrictions in Stand-alone Libraries
19549 The pragmas listed below should be used with caution inside libraries,
19550 as they can create incompatibilities with other Ada libraries:
19552 @item pragma @code{Locking_Policy}
19553 @item pragma @code{Queuing_Policy}
19554 @item pragma @code{Task_Dispatching_Policy}
19555 @item pragma @code{Unreserve_All_Interrupts}
19559 When using a library that contains such pragmas, the user must make sure
19560 that all libraries use the same pragmas with the same values. Otherwise,
19561 @code{Program_Error} will
19562 be raised during the elaboration of the conflicting
19563 libraries. The usage of these pragmas and its consequences for the user
19564 should therefore be well documented.
19566 Similarly, the traceback in the exception occurrence mechanism should be
19567 enabled or disabled in a consistent manner across all libraries.
19568 Otherwise, Program_Error will be raised during the elaboration of the
19569 conflicting libraries.
19571 If the @code{Version} or @code{Body_Version}
19572 attributes are used inside a library, then you need to
19573 perform a @code{gnatbind} step that specifies all @file{ALI} files in all
19574 libraries, so that version identifiers can be properly computed.
19575 In practice these attributes are rarely used, so this is unlikely
19576 to be a consideration.
19578 @node Rebuilding the GNAT Run-Time Library
19579 @section Rebuilding the GNAT Run-Time Library
19580 @cindex GNAT Run-Time Library, rebuilding
19581 @cindex Building the GNAT Run-Time Library
19582 @cindex Rebuilding the GNAT Run-Time Library
19583 @cindex Run-Time Library, rebuilding
19586 It may be useful to recompile the GNAT library in various contexts, the
19587 most important one being the use of partition-wide configuration pragmas
19588 such as @code{Normalize_Scalars}. A special Makefile called
19589 @code{Makefile.adalib} is provided to that effect and can be found in
19590 the directory containing the GNAT library. The location of this
19591 directory depends on the way the GNAT environment has been installed and can
19592 be determined by means of the command:
19599 The last entry in the object search path usually contains the
19600 gnat library. This Makefile contains its own documentation and in
19601 particular the set of instructions needed to rebuild a new library and
19604 @node Using the GNU make Utility
19605 @chapter Using the GNU @code{make} Utility
19609 This chapter offers some examples of makefiles that solve specific
19610 problems. It does not explain how to write a makefile (@pxref{Top,, GNU
19611 make, make, GNU @code{make}}), nor does it try to replace the
19612 @command{gnatmake} utility (@pxref{The GNAT Make Program gnatmake}).
19614 All the examples in this section are specific to the GNU version of
19615 make. Although @command{make} is a standard utility, and the basic language
19616 is the same, these examples use some advanced features found only in
19620 * Using gnatmake in a Makefile::
19621 * Automatically Creating a List of Directories::
19622 * Generating the Command Line Switches::
19623 * Overcoming Command Line Length Limits::
19626 @node Using gnatmake in a Makefile
19627 @section Using gnatmake in a Makefile
19632 Complex project organizations can be handled in a very powerful way by
19633 using GNU make combined with gnatmake. For instance, here is a Makefile
19634 which allows you to build each subsystem of a big project into a separate
19635 shared library. Such a makefile allows you to significantly reduce the link
19636 time of very big applications while maintaining full coherence at
19637 each step of the build process.
19639 The list of dependencies are handled automatically by
19640 @command{gnatmake}. The Makefile is simply used to call gnatmake in each of
19641 the appropriate directories.
19643 Note that you should also read the example on how to automatically
19644 create the list of directories
19645 (@pxref{Automatically Creating a List of Directories})
19646 which might help you in case your project has a lot of subdirectories.
19651 @font@heightrm=cmr8
19654 ## This Makefile is intended to be used with the following directory
19656 ## - The sources are split into a series of csc (computer software components)
19657 ## Each of these csc is put in its own directory.
19658 ## Their name are referenced by the directory names.
19659 ## They will be compiled into shared library (although this would also work
19660 ## with static libraries
19661 ## - The main program (and possibly other packages that do not belong to any
19662 ## csc is put in the top level directory (where the Makefile is).
19663 ## toplevel_dir __ first_csc (sources) __ lib (will contain the library)
19664 ## \_ second_csc (sources) __ lib (will contain the library)
19666 ## Although this Makefile is build for shared library, it is easy to modify
19667 ## to build partial link objects instead (modify the lines with -shared and
19670 ## With this makefile, you can change any file in the system or add any new
19671 ## file, and everything will be recompiled correctly (only the relevant shared
19672 ## objects will be recompiled, and the main program will be re-linked).
19674 # The list of computer software component for your project. This might be
19675 # generated automatically.
19678 # Name of the main program (no extension)
19681 # If we need to build objects with -fPIC, uncomment the following line
19684 # The following variable should give the directory containing libgnat.so
19685 # You can get this directory through 'gnatls -v'. This is usually the last
19686 # directory in the Object_Path.
19689 # The directories for the libraries
19690 # (This macro expands the list of CSC to the list of shared libraries, you
19691 # could simply use the expanded form:
19692 # LIB_DIR=aa/lib/libaa.so bb/lib/libbb.so cc/lib/libcc.so
19693 LIB_DIR=$@{foreach dir,$@{CSC_LIST@},$@{dir@}/lib/lib$@{dir@}.so@}
19695 $@{MAIN@}: objects $@{LIB_DIR@}
19696 gnatbind $@{MAIN@} $@{CSC_LIST:%=-aO%/lib@} -shared
19697 gnatlink $@{MAIN@} $@{CSC_LIST:%=-l%@}
19700 # recompile the sources
19701 gnatmake -c -i $@{MAIN@}.adb $@{NEED_FPIC@} $@{CSC_LIST:%=-I%@}
19703 # Note: In a future version of GNAT, the following commands will be simplified
19704 # by a new tool, gnatmlib
19706 mkdir -p $@{dir $@@ @}
19707 cd $@{dir $@@ @} && gcc -shared -o $@{notdir $@@ @} ../*.o -L$@{GLIB@} -lgnat
19708 cd $@{dir $@@ @} && cp -f ../*.ali .
19710 # The dependencies for the modules
19711 # Note that we have to force the expansion of *.o, since in some cases
19712 # make won't be able to do it itself.
19713 aa/lib/libaa.so: $@{wildcard aa/*.o@}
19714 bb/lib/libbb.so: $@{wildcard bb/*.o@}
19715 cc/lib/libcc.so: $@{wildcard cc/*.o@}
19717 # Make sure all of the shared libraries are in the path before starting the
19720 LD_LIBRARY_PATH=`pwd`/aa/lib:`pwd`/bb/lib:`pwd`/cc/lib ./$@{MAIN@}
19723 $@{RM@} -rf $@{CSC_LIST:%=%/lib@}
19724 $@{RM@} $@{CSC_LIST:%=%/*.ali@}
19725 $@{RM@} $@{CSC_LIST:%=%/*.o@}
19726 $@{RM@} *.o *.ali $@{MAIN@}
19729 @node Automatically Creating a List of Directories
19730 @section Automatically Creating a List of Directories
19733 In most makefiles, you will have to specify a list of directories, and
19734 store it in a variable. For small projects, it is often easier to
19735 specify each of them by hand, since you then have full control over what
19736 is the proper order for these directories, which ones should be
19739 However, in larger projects, which might involve hundreds of
19740 subdirectories, it might be more convenient to generate this list
19743 The example below presents two methods. The first one, although less
19744 general, gives you more control over the list. It involves wildcard
19745 characters, that are automatically expanded by @command{make}. Its
19746 shortcoming is that you need to explicitly specify some of the
19747 organization of your project, such as for instance the directory tree
19748 depth, whether some directories are found in a separate tree, @enddots{}
19750 The second method is the most general one. It requires an external
19751 program, called @command{find}, which is standard on all Unix systems. All
19752 the directories found under a given root directory will be added to the
19758 @font@heightrm=cmr8
19761 # The examples below are based on the following directory hierarchy:
19762 # All the directories can contain any number of files
19763 # ROOT_DIRECTORY -> a -> aa -> aaa
19766 # -> b -> ba -> baa
19769 # This Makefile creates a variable called DIRS, that can be reused any time
19770 # you need this list (see the other examples in this section)
19772 # The root of your project's directory hierarchy
19776 # First method: specify explicitly the list of directories
19777 # This allows you to specify any subset of all the directories you need.
19780 DIRS := a/aa/ a/ab/ b/ba/
19783 # Second method: use wildcards
19784 # Note that the argument(s) to wildcard below should end with a '/'.
19785 # Since wildcards also return file names, we have to filter them out
19786 # to avoid duplicate directory names.
19787 # We thus use make's @code{dir} and @code{sort} functions.
19788 # It sets DIRs to the following value (note that the directories aaa and baa
19789 # are not given, unless you change the arguments to wildcard).
19790 # DIRS= ./a/a/ ./b/ ./a/aa/ ./a/ab/ ./a/ac/ ./b/ba/ ./b/bb/ ./b/bc/
19793 DIRS := $@{sort $@{dir $@{wildcard $@{ROOT_DIRECTORY@}/*/
19794 $@{ROOT_DIRECTORY@}/*/*/@}@}@}
19797 # Third method: use an external program
19798 # This command is much faster if run on local disks, avoiding NFS slowdowns.
19799 # This is the most complete command: it sets DIRs to the following value:
19800 # DIRS= ./a ./a/aa ./a/aa/aaa ./a/ab ./a/ac ./b ./b/ba ./b/ba/baa ./b/bb ./b/bc
19803 DIRS := $@{shell find $@{ROOT_DIRECTORY@} -type d -print@}
19807 @node Generating the Command Line Switches
19808 @section Generating the Command Line Switches
19811 Once you have created the list of directories as explained in the
19812 previous section (@pxref{Automatically Creating a List of Directories}),
19813 you can easily generate the command line arguments to pass to gnatmake.
19815 For the sake of completeness, this example assumes that the source path
19816 is not the same as the object path, and that you have two separate lists
19820 # see "Automatically creating a list of directories" to create
19825 GNATMAKE_SWITCHES := $@{patsubst %,-aI%,$@{SOURCE_DIRS@}@}
19826 GNATMAKE_SWITCHES += $@{patsubst %,-aO%,$@{OBJECT_DIRS@}@}
19829 gnatmake $@{GNATMAKE_SWITCHES@} main_unit
19832 @node Overcoming Command Line Length Limits
19833 @section Overcoming Command Line Length Limits
19836 One problem that might be encountered on big projects is that many
19837 operating systems limit the length of the command line. It is thus hard to give
19838 gnatmake the list of source and object directories.
19840 This example shows how you can set up environment variables, which will
19841 make @command{gnatmake} behave exactly as if the directories had been
19842 specified on the command line, but have a much higher length limit (or
19843 even none on most systems).
19845 It assumes that you have created a list of directories in your Makefile,
19846 using one of the methods presented in
19847 @ref{Automatically Creating a List of Directories}.
19848 For the sake of completeness, we assume that the object
19849 path (where the ALI files are found) is different from the sources patch.
19851 Note a small trick in the Makefile below: for efficiency reasons, we
19852 create two temporary variables (SOURCE_LIST and OBJECT_LIST), that are
19853 expanded immediately by @code{make}. This way we overcome the standard
19854 make behavior which is to expand the variables only when they are
19857 On Windows, if you are using the standard Windows command shell, you must
19858 replace colons with semicolons in the assignments to these variables.
19863 @font@heightrm=cmr8
19866 # In this example, we create both ADA_INCLUDE_PATH and ADA_OBJECT_PATH.
19867 # This is the same thing as putting the -I arguments on the command line.
19868 # (the equivalent of using -aI on the command line would be to define
19869 # only ADA_INCLUDE_PATH, the equivalent of -aO is ADA_OBJECT_PATH).
19870 # You can of course have different values for these variables.
19872 # Note also that we need to keep the previous values of these variables, since
19873 # they might have been set before running 'make' to specify where the GNAT
19874 # library is installed.
19876 # see "Automatically creating a list of directories" to create these
19882 space:=$@{empty@} $@{empty@}
19883 SOURCE_LIST := $@{subst $@{space@},:,$@{SOURCE_DIRS@}@}
19884 OBJECT_LIST := $@{subst $@{space@},:,$@{OBJECT_DIRS@}@}
19885 ADA_INCLUDE_PATH += $@{SOURCE_LIST@}
19886 ADA_OBJECT_PATH += $@{OBJECT_LIST@}
19887 export ADA_INCLUDE_PATH
19888 export ADA_OBJECT_PATH
19895 @node Memory Management Issues
19896 @chapter Memory Management Issues
19899 This chapter describes some useful memory pools provided in the GNAT library
19900 and in particular the GNAT Debug Pool facility, which can be used to detect
19901 incorrect uses of access values (including ``dangling references'').
19903 It also describes the @command{gnatmem} tool, which can be used to track down
19908 * Some Useful Memory Pools::
19909 * The GNAT Debug Pool Facility::
19911 * The gnatmem Tool::
19915 @node Some Useful Memory Pools
19916 @section Some Useful Memory Pools
19917 @findex Memory Pool
19918 @cindex storage, pool
19921 The @code{System.Pool_Global} package offers the Unbounded_No_Reclaim_Pool
19922 storage pool. Allocations use the standard system call @code{malloc} while
19923 deallocations use the standard system call @code{free}. No reclamation is
19924 performed when the pool goes out of scope. For performance reasons, the
19925 standard default Ada allocators/deallocators do not use any explicit storage
19926 pools but if they did, they could use this storage pool without any change in
19927 behavior. That is why this storage pool is used when the user
19928 manages to make the default implicit allocator explicit as in this example:
19929 @smallexample @c ada
19930 type T1 is access Something;
19931 -- no Storage pool is defined for T2
19932 type T2 is access Something_Else;
19933 for T2'Storage_Pool use T1'Storage_Pool;
19934 -- the above is equivalent to
19935 for T2'Storage_Pool use System.Pool_Global.Global_Pool_Object;
19939 The @code{System.Pool_Local} package offers the Unbounded_Reclaim_Pool storage
19940 pool. The allocation strategy is similar to @code{Pool_Local}'s
19941 except that the all
19942 storage allocated with this pool is reclaimed when the pool object goes out of
19943 scope. This pool provides a explicit mechanism similar to the implicit one
19944 provided by several Ada 83 compilers for allocations performed through a local
19945 access type and whose purpose was to reclaim memory when exiting the
19946 scope of a given local access. As an example, the following program does not
19947 leak memory even though it does not perform explicit deallocation:
19949 @smallexample @c ada
19950 with System.Pool_Local;
19951 procedure Pooloc1 is
19952 procedure Internal is
19953 type A is access Integer;
19954 X : System.Pool_Local.Unbounded_Reclaim_Pool;
19955 for A'Storage_Pool use X;
19958 for I in 1 .. 50 loop
19963 for I in 1 .. 100 loop
19970 The @code{System.Pool_Size} package implements the Stack_Bounded_Pool used when
19971 @code{Storage_Size} is specified for an access type.
19972 The whole storage for the pool is
19973 allocated at once, usually on the stack at the point where the access type is
19974 elaborated. It is automatically reclaimed when exiting the scope where the
19975 access type is defined. This package is not intended to be used directly by the
19976 user and it is implicitly used for each such declaration:
19978 @smallexample @c ada
19979 type T1 is access Something;
19980 for T1'Storage_Size use 10_000;
19983 @node The GNAT Debug Pool Facility
19984 @section The GNAT Debug Pool Facility
19986 @cindex storage, pool, memory corruption
19989 The use of unchecked deallocation and unchecked conversion can easily
19990 lead to incorrect memory references. The problems generated by such
19991 references are usually difficult to tackle because the symptoms can be
19992 very remote from the origin of the problem. In such cases, it is
19993 very helpful to detect the problem as early as possible. This is the
19994 purpose of the Storage Pool provided by @code{GNAT.Debug_Pools}.
19996 In order to use the GNAT specific debugging pool, the user must
19997 associate a debug pool object with each of the access types that may be
19998 related to suspected memory problems. See Ada Reference Manual 13.11.
19999 @smallexample @c ada
20000 type Ptr is access Some_Type;
20001 Pool : GNAT.Debug_Pools.Debug_Pool;
20002 for Ptr'Storage_Pool use Pool;
20006 @code{GNAT.Debug_Pools} is derived from a GNAT-specific kind of
20007 pool: the @code{Checked_Pool}. Such pools, like standard Ada storage pools,
20008 allow the user to redefine allocation and deallocation strategies. They
20009 also provide a checkpoint for each dereference, through the use of
20010 the primitive operation @code{Dereference} which is implicitly called at
20011 each dereference of an access value.
20013 Once an access type has been associated with a debug pool, operations on
20014 values of the type may raise four distinct exceptions,
20015 which correspond to four potential kinds of memory corruption:
20018 @code{GNAT.Debug_Pools.Accessing_Not_Allocated_Storage}
20020 @code{GNAT.Debug_Pools.Accessing_Deallocated_Storage}
20022 @code{GNAT.Debug_Pools.Freeing_Not_Allocated_Storage}
20024 @code{GNAT.Debug_Pools.Freeing_Deallocated_Storage }
20028 For types associated with a Debug_Pool, dynamic allocation is performed using
20029 the standard GNAT allocation routine. References to all allocated chunks of
20030 memory are kept in an internal dictionary. Several deallocation strategies are
20031 provided, whereupon the user can choose to release the memory to the system,
20032 keep it allocated for further invalid access checks, or fill it with an easily
20033 recognizable pattern for debug sessions. The memory pattern is the old IBM
20034 hexadecimal convention: @code{16#DEADBEEF#}.
20036 See the documentation in the file g-debpoo.ads for more information on the
20037 various strategies.
20039 Upon each dereference, a check is made that the access value denotes a
20040 properly allocated memory location. Here is a complete example of use of
20041 @code{Debug_Pools}, that includes typical instances of memory corruption:
20042 @smallexample @c ada
20046 with Gnat.Io; use Gnat.Io;
20047 with Unchecked_Deallocation;
20048 with Unchecked_Conversion;
20049 with GNAT.Debug_Pools;
20050 with System.Storage_Elements;
20051 with Ada.Exceptions; use Ada.Exceptions;
20052 procedure Debug_Pool_Test is
20054 type T is access Integer;
20055 type U is access all T;
20057 P : GNAT.Debug_Pools.Debug_Pool;
20058 for T'Storage_Pool use P;
20060 procedure Free is new Unchecked_Deallocation (Integer, T);
20061 function UC is new Unchecked_Conversion (U, T);
20064 procedure Info is new GNAT.Debug_Pools.Print_Info(Put_Line);
20074 Put_Line (Integer'Image(B.all));
20076 when E : others => Put_Line ("raised: " & Exception_Name (E));
20081 when E : others => Put_Line ("raised: " & Exception_Name (E));
20085 Put_Line (Integer'Image(B.all));
20087 when E : others => Put_Line ("raised: " & Exception_Name (E));
20092 when E : others => Put_Line ("raised: " & Exception_Name (E));
20095 end Debug_Pool_Test;
20099 The debug pool mechanism provides the following precise diagnostics on the
20100 execution of this erroneous program:
20103 Total allocated bytes : 0
20104 Total deallocated bytes : 0
20105 Current Water Mark: 0
20109 Total allocated bytes : 8
20110 Total deallocated bytes : 0
20111 Current Water Mark: 8
20114 raised: GNAT.DEBUG_POOLS.ACCESSING_DEALLOCATED_STORAGE
20115 raised: GNAT.DEBUG_POOLS.FREEING_DEALLOCATED_STORAGE
20116 raised: GNAT.DEBUG_POOLS.ACCESSING_NOT_ALLOCATED_STORAGE
20117 raised: GNAT.DEBUG_POOLS.FREEING_NOT_ALLOCATED_STORAGE
20119 Total allocated bytes : 8
20120 Total deallocated bytes : 4
20121 Current Water Mark: 4
20126 @node The gnatmem Tool
20127 @section The @command{gnatmem} Tool
20131 The @code{gnatmem} utility monitors dynamic allocation and
20132 deallocation activity in a program, and displays information about
20133 incorrect deallocations and possible sources of memory leaks.
20134 It is designed to work in association with a static runtime library
20135 only and in this context provides three types of information:
20138 General information concerning memory management, such as the total
20139 number of allocations and deallocations, the amount of allocated
20140 memory and the high water mark, i.e.@: the largest amount of allocated
20141 memory in the course of program execution.
20144 Backtraces for all incorrect deallocations, that is to say deallocations
20145 which do not correspond to a valid allocation.
20148 Information on each allocation that is potentially the origin of a memory
20153 * Running gnatmem::
20154 * Switches for gnatmem::
20155 * Example of gnatmem Usage::
20158 @node Running gnatmem
20159 @subsection Running @code{gnatmem}
20162 @code{gnatmem} makes use of the output created by the special version of
20163 allocation and deallocation routines that record call information. This
20164 allows to obtain accurate dynamic memory usage history at a minimal cost to
20165 the execution speed. Note however, that @code{gnatmem} is not supported on
20166 all platforms (currently, it is supported on AIX, HP-UX, GNU/Linux,
20167 Solaris and Windows NT/2000/XP (x86).
20170 The @code{gnatmem} command has the form
20173 $ gnatmem @ovar{switches} user_program
20177 The program must have been linked with the instrumented version of the
20178 allocation and deallocation routines. This is done by linking with the
20179 @file{libgmem.a} library. For correct symbolic backtrace information,
20180 the user program should be compiled with debugging options
20181 (see @ref{Switches for gcc}). For example to build @file{my_program}:
20184 $ gnatmake -g my_program -largs -lgmem
20188 As library @file{libgmem.a} contains an alternate body for package
20189 @code{System.Memory}, @file{s-memory.adb} should not be compiled and linked
20190 when an executable is linked with library @file{libgmem.a}. It is then not
20191 recommended to use @command{gnatmake} with switch @option{^-a^/ALL_FILES^}.
20194 When @file{my_program} is executed, the file @file{gmem.out} is produced.
20195 This file contains information about all allocations and deallocations
20196 performed by the program. It is produced by the instrumented allocations and
20197 deallocations routines and will be used by @code{gnatmem}.
20199 In order to produce symbolic backtrace information for allocations and
20200 deallocations performed by the GNAT run-time library, you need to use a
20201 version of that library that has been compiled with the @option{-g} switch
20202 (see @ref{Rebuilding the GNAT Run-Time Library}).
20204 Gnatmem must be supplied with the @file{gmem.out} file and the executable to
20205 examine. If the location of @file{gmem.out} file was not explicitly supplied by
20206 @option{-i} switch, gnatmem will assume that this file can be found in the
20207 current directory. For example, after you have executed @file{my_program},
20208 @file{gmem.out} can be analyzed by @code{gnatmem} using the command:
20211 $ gnatmem my_program
20215 This will produce the output with the following format:
20217 *************** debut cc
20219 $ gnatmem my_program
20223 Total number of allocations : 45
20224 Total number of deallocations : 6
20225 Final Water Mark (non freed mem) : 11.29 Kilobytes
20226 High Water Mark : 11.40 Kilobytes
20231 Allocation Root # 2
20232 -------------------
20233 Number of non freed allocations : 11
20234 Final Water Mark (non freed mem) : 1.16 Kilobytes
20235 High Water Mark : 1.27 Kilobytes
20237 my_program.adb:23 my_program.alloc
20243 The first block of output gives general information. In this case, the
20244 Ada construct ``@code{@b{new}}'' was executed 45 times, and only 6 calls to an
20245 Unchecked_Deallocation routine occurred.
20248 Subsequent paragraphs display information on all allocation roots.
20249 An allocation root is a specific point in the execution of the program
20250 that generates some dynamic allocation, such as a ``@code{@b{new}}''
20251 construct. This root is represented by an execution backtrace (or subprogram
20252 call stack). By default the backtrace depth for allocations roots is 1, so
20253 that a root corresponds exactly to a source location. The backtrace can
20254 be made deeper, to make the root more specific.
20256 @node Switches for gnatmem
20257 @subsection Switches for @code{gnatmem}
20260 @code{gnatmem} recognizes the following switches:
20265 @cindex @option{-q} (@code{gnatmem})
20266 Quiet. Gives the minimum output needed to identify the origin of the
20267 memory leaks. Omits statistical information.
20270 @cindex @var{N} (@code{gnatmem})
20271 N is an integer literal (usually between 1 and 10) which controls the
20272 depth of the backtraces defining allocation root. The default value for
20273 N is 1. The deeper the backtrace, the more precise the localization of
20274 the root. Note that the total number of roots can depend on this
20275 parameter. This parameter must be specified @emph{before} the name of the
20276 executable to be analyzed, to avoid ambiguity.
20279 @cindex @option{-b} (@code{gnatmem})
20280 This switch has the same effect as just depth parameter.
20282 @item -i @var{file}
20283 @cindex @option{-i} (@code{gnatmem})
20284 Do the @code{gnatmem} processing starting from @file{file}, rather than
20285 @file{gmem.out} in the current directory.
20288 @cindex @option{-m} (@code{gnatmem})
20289 This switch causes @code{gnatmem} to mask the allocation roots that have less
20290 than n leaks. The default value is 1. Specifying the value of 0 will allow to
20291 examine even the roots that didn't result in leaks.
20294 @cindex @option{-s} (@code{gnatmem})
20295 This switch causes @code{gnatmem} to sort the allocation roots according to the
20296 specified order of sort criteria, each identified by a single letter. The
20297 currently supported criteria are @code{n, h, w} standing respectively for
20298 number of unfreed allocations, high watermark, and final watermark
20299 corresponding to a specific root. The default order is @code{nwh}.
20303 @node Example of gnatmem Usage
20304 @subsection Example of @code{gnatmem} Usage
20307 The following example shows the use of @code{gnatmem}
20308 on a simple memory-leaking program.
20309 Suppose that we have the following Ada program:
20311 @smallexample @c ada
20314 with Unchecked_Deallocation;
20315 procedure Test_Gm is
20317 type T is array (1..1000) of Integer;
20318 type Ptr is access T;
20319 procedure Free is new Unchecked_Deallocation (T, Ptr);
20322 procedure My_Alloc is
20327 procedure My_DeAlloc is
20335 for I in 1 .. 5 loop
20336 for J in I .. 5 loop
20347 The program needs to be compiled with debugging option and linked with
20348 @code{gmem} library:
20351 $ gnatmake -g test_gm -largs -lgmem
20355 Then we execute the program as usual:
20362 Then @code{gnatmem} is invoked simply with
20368 which produces the following output (result may vary on different platforms):
20373 Total number of allocations : 18
20374 Total number of deallocations : 5
20375 Final Water Mark (non freed mem) : 53.00 Kilobytes
20376 High Water Mark : 56.90 Kilobytes
20378 Allocation Root # 1
20379 -------------------
20380 Number of non freed allocations : 11
20381 Final Water Mark (non freed mem) : 42.97 Kilobytes
20382 High Water Mark : 46.88 Kilobytes
20384 test_gm.adb:11 test_gm.my_alloc
20386 Allocation Root # 2
20387 -------------------
20388 Number of non freed allocations : 1
20389 Final Water Mark (non freed mem) : 10.02 Kilobytes
20390 High Water Mark : 10.02 Kilobytes
20392 s-secsta.adb:81 system.secondary_stack.ss_init
20394 Allocation Root # 3
20395 -------------------
20396 Number of non freed allocations : 1
20397 Final Water Mark (non freed mem) : 12 Bytes
20398 High Water Mark : 12 Bytes
20400 s-secsta.adb:181 system.secondary_stack.ss_init
20404 Note that the GNAT run time contains itself a certain number of
20405 allocations that have no corresponding deallocation,
20406 as shown here for root #2 and root
20407 #3. This is a normal behavior when the number of non-freed allocations
20408 is one, it allocates dynamic data structures that the run time needs for
20409 the complete lifetime of the program. Note also that there is only one
20410 allocation root in the user program with a single line back trace:
20411 test_gm.adb:11 test_gm.my_alloc, whereas a careful analysis of the
20412 program shows that 'My_Alloc' is called at 2 different points in the
20413 source (line 21 and line 24). If those two allocation roots need to be
20414 distinguished, the backtrace depth parameter can be used:
20417 $ gnatmem 3 test_gm
20421 which will give the following output:
20426 Total number of allocations : 18
20427 Total number of deallocations : 5
20428 Final Water Mark (non freed mem) : 53.00 Kilobytes
20429 High Water Mark : 56.90 Kilobytes
20431 Allocation Root # 1
20432 -------------------
20433 Number of non freed allocations : 10
20434 Final Water Mark (non freed mem) : 39.06 Kilobytes
20435 High Water Mark : 42.97 Kilobytes
20437 test_gm.adb:11 test_gm.my_alloc
20438 test_gm.adb:24 test_gm
20439 b_test_gm.c:52 main
20441 Allocation Root # 2
20442 -------------------
20443 Number of non freed allocations : 1
20444 Final Water Mark (non freed mem) : 10.02 Kilobytes
20445 High Water Mark : 10.02 Kilobytes
20447 s-secsta.adb:81 system.secondary_stack.ss_init
20448 s-secsta.adb:283 <system__secondary_stack___elabb>
20449 b_test_gm.c:33 adainit
20451 Allocation Root # 3
20452 -------------------
20453 Number of non freed allocations : 1
20454 Final Water Mark (non freed mem) : 3.91 Kilobytes
20455 High Water Mark : 3.91 Kilobytes
20457 test_gm.adb:11 test_gm.my_alloc
20458 test_gm.adb:21 test_gm
20459 b_test_gm.c:52 main
20461 Allocation Root # 4
20462 -------------------
20463 Number of non freed allocations : 1
20464 Final Water Mark (non freed mem) : 12 Bytes
20465 High Water Mark : 12 Bytes
20467 s-secsta.adb:181 system.secondary_stack.ss_init
20468 s-secsta.adb:283 <system__secondary_stack___elabb>
20469 b_test_gm.c:33 adainit
20473 The allocation root #1 of the first example has been split in 2 roots #1
20474 and #3 thanks to the more precise associated backtrace.
20478 @node Stack Related Facilities
20479 @chapter Stack Related Facilities
20482 This chapter describes some useful tools associated with stack
20483 checking and analysis. In
20484 particular, it deals with dynamic and static stack usage measurements.
20487 * Stack Overflow Checking::
20488 * Static Stack Usage Analysis::
20489 * Dynamic Stack Usage Analysis::
20492 @node Stack Overflow Checking
20493 @section Stack Overflow Checking
20494 @cindex Stack Overflow Checking
20495 @cindex -fstack-check
20498 For most operating systems, @command{gcc} does not perform stack overflow
20499 checking by default. This means that if the main environment task or
20500 some other task exceeds the available stack space, then unpredictable
20501 behavior will occur. Most native systems offer some level of protection by
20502 adding a guard page at the end of each task stack. This mechanism is usually
20503 not enough for dealing properly with stack overflow situations because
20504 a large local variable could ``jump'' above the guard page.
20505 Furthermore, when the
20506 guard page is hit, there may not be any space left on the stack for executing
20507 the exception propagation code. Enabling stack checking avoids
20510 To activate stack checking, compile all units with the gcc option
20511 @option{-fstack-check}. For example:
20514 gcc -c -fstack-check package1.adb
20518 Units compiled with this option will generate extra instructions to check
20519 that any use of the stack (for procedure calls or for declaring local
20520 variables in declare blocks) does not exceed the available stack space.
20521 If the space is exceeded, then a @code{Storage_Error} exception is raised.
20523 For declared tasks, the stack size is controlled by the size
20524 given in an applicable @code{Storage_Size} pragma or by the value specified
20525 at bind time with @option{-d} (@pxref{Switches for gnatbind}) or is set to
20526 the default size as defined in the GNAT runtime otherwise.
20528 For the environment task, the stack size depends on
20529 system defaults and is unknown to the compiler. Stack checking
20530 may still work correctly if a fixed
20531 size stack is allocated, but this cannot be guaranteed.
20533 To ensure that a clean exception is signalled for stack
20534 overflow, set the environment variable
20535 @env{GNAT_STACK_LIMIT} to indicate the maximum
20536 stack area that can be used, as in:
20537 @cindex GNAT_STACK_LIMIT
20540 SET GNAT_STACK_LIMIT 1600
20544 The limit is given in kilobytes, so the above declaration would
20545 set the stack limit of the environment task to 1.6 megabytes.
20546 Note that the only purpose of this usage is to limit the amount
20547 of stack used by the environment task. If it is necessary to
20548 increase the amount of stack for the environment task, then this
20549 is an operating systems issue, and must be addressed with the
20550 appropriate operating systems commands.
20553 To have a fixed size stack in the environment task, the stack must be put
20554 in the P0 address space and its size specified. Use these switches to
20558 gnatmake my_progs -largs "-Wl,--opt=STACK=4000,/p0image"
20562 The quotes are required to keep case. The number after @samp{STACK=} is the
20563 size of the environmental task stack in pagelets (512 bytes). In this example
20564 the stack size is about 2 megabytes.
20567 A consequence of the @option{/p0image} qualifier is also to makes RMS buffers
20568 be placed in P0 space. Refer to @cite{HP OpenVMS Linker Utility Manual} for
20569 more details about the @option{/p0image} qualifier and the @option{stack}
20573 @node Static Stack Usage Analysis
20574 @section Static Stack Usage Analysis
20575 @cindex Static Stack Usage Analysis
20576 @cindex -fstack-usage
20579 A unit compiled with @option{-fstack-usage} will generate an extra file
20581 the maximum amount of stack used, on a per-function basis.
20582 The file has the same
20583 basename as the target object file with a @file{.su} extension.
20584 Each line of this file is made up of three fields:
20588 The name of the function.
20592 One or more qualifiers: @code{static}, @code{dynamic}, @code{bounded}.
20595 The second field corresponds to the size of the known part of the function
20598 The qualifier @code{static} means that the function frame size
20600 It usually means that all local variables have a static size.
20601 In this case, the second field is a reliable measure of the function stack
20604 The qualifier @code{dynamic} means that the function frame size is not static.
20605 It happens mainly when some local variables have a dynamic size. When this
20606 qualifier appears alone, the second field is not a reliable measure
20607 of the function stack analysis. When it is qualified with @code{bounded}, it
20608 means that the second field is a reliable maximum of the function stack
20611 @node Dynamic Stack Usage Analysis
20612 @section Dynamic Stack Usage Analysis
20615 It is possible to measure the maximum amount of stack used by a task, by
20616 adding a switch to @command{gnatbind}, as:
20619 $ gnatbind -u0 file
20623 With this option, at each task termination, its stack usage is output on
20625 It is not always convenient to output the stack usage when the program
20626 is still running. Hence, it is possible to delay this output until program
20627 termination. for a given number of tasks specified as the argument of the
20628 @option{-u} option. For instance:
20631 $ gnatbind -u100 file
20635 will buffer the stack usage information of the first 100 tasks to terminate and
20636 output this info at program termination. Results are displayed in four
20640 Index | Task Name | Stack Size | Stack Usage [Value +/- Variation]
20647 is a number associated with each task.
20650 is the name of the task analyzed.
20653 is the maximum size for the stack.
20656 is the measure done by the stack analyzer. In order to prevent overflow, the stack
20657 is not entirely analyzed, and it's not possible to know exactly how
20658 much has actually been used. The report thus contains the theoretical stack usage
20659 (Value) and the possible variation (Variation) around this value.
20664 The environment task stack, e.g., the stack that contains the main unit, is
20665 only processed when the environment variable GNAT_STACK_LIMIT is set.
20668 @c *********************************
20670 @c *********************************
20671 @node Verifying Properties Using gnatcheck
20672 @chapter Verifying Properties Using @command{gnatcheck}
20674 @cindex @command{gnatcheck}
20677 The @command{gnatcheck} tool is an ASIS-based utility that checks properties
20678 of Ada source files according to a given set of semantic rules.
20681 In order to check compliance with a given rule, @command{gnatcheck} has to
20682 semantically analyze the Ada sources.
20683 Therefore, checks can only be performed on
20684 legal Ada units. Moreover, when a unit depends semantically upon units located
20685 outside the current directory, the source search path has to be provided when
20686 calling @command{gnatcheck}, either through a specified project file or
20687 through @command{gnatcheck} switches as described below.
20689 A number of rules are predefined in @command{gnatcheck} and are described
20690 later in this chapter.
20691 You can also add new rules, by modifying the @command{gnatcheck} code and
20692 rebuilding the tool. In order to add a simple rule making some local checks,
20693 a small amount of straightforward ASIS-based programming is usually needed.
20695 Project support for @command{gnatcheck} is provided by the GNAT
20696 driver (see @ref{The GNAT Driver and Project Files}).
20698 Invoking @command{gnatcheck} on the command line has the form:
20701 $ gnatcheck @ovar{switches} @{@var{filename}@}
20702 @r{[}^-files^/FILES^=@{@var{arg_list_filename}@}@r{]}
20703 @r{[}-cargs @var{gcc_switches}@r{]} -rules @var{rule_options}
20710 @var{switches} specify the general tool options
20713 Each @var{filename} is the name (including the extension) of a source
20714 file to process. ``Wildcards'' are allowed, and
20715 the file name may contain path information.
20718 Each @var{arg_list_filename} is the name (including the extension) of a text
20719 file containing the names of the source files to process, separated by spaces
20723 @var{gcc_switches} is a list of switches for
20724 @command{gcc}. They will be passed on to all compiler invocations made by
20725 @command{gnatcheck} to generate the ASIS trees. Here you can provide
20726 @option{^-I^/INCLUDE_DIRS=^} switches to form the source search path,
20727 and use the @option{-gnatec} switch to set the configuration file.
20730 @var{rule_options} is a list of options for controlling a set of
20731 rules to be checked by @command{gnatcheck} (@pxref{gnatcheck Rule Options}).
20735 Either a @file{@var{filename}} or an @file{@var{arg_list_filename}} must be
20739 * Format of the Report File::
20740 * General gnatcheck Switches::
20741 * gnatcheck Rule Options::
20742 * Adding the Results of Compiler Checks to gnatcheck Output::
20743 * Project-Wide Checks::
20745 * Predefined Rules::
20746 * Example of gnatcheck Usage::
20749 @node Format of the Report File
20750 @section Format of the Report File
20751 @cindex Report file (for @code{gnatcheck})
20754 The @command{gnatcheck} tool outputs on @file{stdout} all messages concerning
20756 It also creates a text file that
20757 contains the complete report of the last gnatcheck run. By default this file
20758 is named named @file{^gnatcheck.out^GNATCHECK.OUT^} and it is located in the
20759 current directory; the @option{^-o^/OUTPUT^} option can be used to change the
20760 name and/or location of the report file. This report contains:
20762 @item date and time of @command{gnatcheck} run, the version of
20763 the tool that has generated this report and the full parameters
20764 of the @command{gnatcheck} invocation;
20765 @item list of enabled rules;
20766 @item total number of detected violations;
20767 @item list of source files where rule violations have been detected;
20768 @item list of source files where no violations have been detected.
20771 @node General gnatcheck Switches
20772 @section General @command{gnatcheck} Switches
20775 The following switches control the general @command{gnatcheck} behavior
20779 @cindex @option{^-a^/ALL^} (@command{gnatcheck})
20781 Process all units including those with read-only ALI files such as
20782 those from the GNAT Run-Time library.
20786 @cindex @option{-d} (@command{gnatcheck})
20791 @cindex @option{-dd} (@command{gnatcheck})
20793 Progress indicator mode (for use in GPS).
20796 @cindex @option{^-h^/HELP^} (@command{gnatcheck})
20798 List the predefined and user-defined rules. For more details see
20799 @ref{Predefined Rules}.
20801 @cindex @option{^-l^/LOCS^} (@command{gnatcheck})
20803 Use full source locations references in the report file. For a construct from
20804 a generic instantiation a full source location is a chain from the location
20805 of this construct in the generic unit to the place where this unit is
20808 @cindex @option{^-log^/LOG^} (@command{gnatcheck})
20810 Duplicate all the output sent to @file{stderr} into a log file. The log file
20811 is named @file{gnatcheck.log} and is located in the current directory.
20813 @cindex @option{^-m^/DIAGNOSTIC_LIMIT^} (@command{gnatcheck})
20814 @item ^-m@i{nnnn}^/DIAGNOSTIC_LIMIT=@i{nnnn}^
20815 Maximum number of diagnostics to be sent to @file{stdout}, where @i{nnnn} is in
20816 the range 0@dots{}1000;
20817 the default value is 500. Zero means that there is no limitation on
20818 the number of diagnostic messages to be output.
20820 @cindex @option{^-q^/QUIET^} (@command{gnatcheck})
20822 Quiet mode. All the diagnostics about rule violations are placed in the
20823 @command{gnatcheck} report file only, without duplication on @file{stdout}.
20825 @cindex @option{^-s^/SHORT^} (@command{gnatcheck})
20827 Short format of the report file (no version information, no list of applied
20828 rules, no list of checked sources is included)
20830 @cindex @option{^--include-file=@var{file}^/INCLUDE_FILE=@var{file}^} (@command{gnatcheck})
20831 @item ^--include-file^/INCLUDE_FILE^
20832 Append the content of the specified text file to the report file
20834 @cindex @option{^-t^/TIME^} (@command{gnatcheck})
20836 Print out execution time.
20838 @cindex @option{^-v^/VERBOSE^} (@command{gnatcheck})
20839 @item ^-v^/VERBOSE^
20840 Verbose mode; @command{gnatcheck} generates version information and then
20841 a trace of sources being processed.
20843 @cindex @option{^-o ^/OUTPUT^} (@command{gnatcheck})
20844 @item ^-o ^/OUTPUT=^@var{report_file}
20845 Set name of report file file to @var{report_file} .
20850 Note that if any of the options @option{^-s1^/COMPILER_STYLE^},
20851 @option{^-s2^/BY_RULES^} or
20852 @option{^-s3^/BY_FILES_BY_RULES^} is specified,
20853 then the @command{gnatcheck} report file will only contain sections
20854 explicitly denoted by these options.
20856 @node gnatcheck Rule Options
20857 @section @command{gnatcheck} Rule Options
20860 The following options control the processing performed by
20861 @command{gnatcheck}.
20864 @cindex @option{+ALL} (@command{gnatcheck})
20866 Turn all the rule checks ON.
20868 @cindex @option{-ALL} (@command{gnatcheck})
20870 Turn all the rule checks OFF.
20872 @cindex @option{+R} (@command{gnatcheck})
20873 @item +R@var{rule_id}@r{[}:@var{param}@r{]}
20874 Turn on the check for a specified rule with the specified parameter, if any.
20875 @var{rule_id} must be the identifier of one of the currently implemented rules
20876 (use @option{^-h^/HELP^} for the list of implemented rules). Rule identifiers
20877 are not case-sensitive. The @var{param} item must
20878 be a string representing a valid parameter(s) for the specified rule.
20879 If it contains any space characters then this string must be enclosed in
20882 @cindex @option{-R} (@command{gnatcheck})
20883 @item -R@var{rule_id}@r{[}:@var{param}@r{]}
20884 Turn off the check for a specified rule with the specified parameter, if any.
20886 @cindex @option{-from} (@command{gnatcheck})
20887 @item -from=@var{rule_option_filename}
20888 Read the rule options from the text file @var{rule_option_filename}, referred
20889 to as a ``coding standard file'' below.
20894 The default behavior is that all the rule checks are disabled.
20896 A coding standard file is a text file that contains a set of rule options
20898 @cindex Coding standard file (for @code{gnatcheck})
20899 The file may contain empty lines and Ada-style comments (comment
20900 lines and end-of-line comments). There can be several rule options on a
20901 single line (separated by a space).
20903 A coding standard file may reference other coding standard files by including
20904 more @option{-from=@var{rule_option_filename}}
20905 options, each such option being replaced with the content of the
20906 corresponding coding standard file during processing. In case a
20907 cycle is detected (that is, @file{@var{rule_file_1}} reads rule options
20908 from @file{@var{rule_file_2}}, and @file{@var{rule_file_2}} reads
20909 (directly or indirectly) rule options from @file{@var{rule_file_1}}),
20910 processing fails with an error message.
20913 @node Adding the Results of Compiler Checks to gnatcheck Output
20914 @section Adding the Results of Compiler Checks to @command{gnatcheck} Output
20917 The @command{gnatcheck} tool can include in the generated diagnostic messages
20919 the report file the results of the checks performed by the compiler. Though
20920 disabled by default, this effect may be obtained by using @option{+R} with
20921 the following rule identifiers and parameters:
20925 To record restrictions violations (which are performed by the compiler if the
20926 pragma @code{Restrictions} or @code{Restriction_Warnings} are given),
20927 use the @code{Restrictions} rule
20928 with the same parameters as pragma
20929 @code{Restrictions} or @code{Restriction_Warnings}.
20932 To record compiler style checks (@pxref{Style Checking}), use the
20933 @code{Style_Checks} rule.
20934 This rule takes a parameter in one of the following forms:
20938 which enables the standard style checks corresponding to the @option{-gnatyy}
20939 GNAT style check option, or
20942 a string with the same
20943 structure and semantics as the @code{string_LITERAL} parameter of the
20944 GNAT pragma @code{Style_Checks}
20945 (for further information about this pragma,
20946 @pxref{Pragma Style_Checks,,, gnat_rm, GNAT Reference Manual}).
20951 @code{+RStyle_Checks:O} rule option activates
20952 the compiler style check that corresponds to
20953 @code{-gnatyO} style check option.
20956 To record compiler warnings (@pxref{Warning Message Control}), use the
20957 @code{Warnings} rule with a parameter that is a valid
20958 @i{static_string_expression} argument of the GNAT pragma @code{Warnings}
20959 (for further information about this pragma,
20960 @pxref{Pragma Warnings,,,gnat_rm, GNAT Reference Manual}).
20961 Note that in case of gnatcheck
20962 's' parameter, that corresponds to the GNAT @option{-gnatws} option, disables
20963 all the specific warnings, but not suppresses the warning mode,
20964 and 'e' parameter, corresponding to @option{-gnatwe} that means
20965 "treat warnings as errors", does not have any effect.
20969 To disable a specific restriction check, use @code{-RStyle_Checks} gnatcheck
20970 option with the corresponding restriction name as a parameter. @code{-R} is
20971 not available for @code{Style_Checks} and @code{Warnings} options, to disable
20972 warnings and style checks, use the corresponding warning and style options.
20974 @node Project-Wide Checks
20975 @section Project-Wide Checks
20976 @cindex Project-wide checks (for @command{gnatcheck})
20979 In order to perform checks on all units of a given project, you can use
20980 the GNAT driver along with the @option{-P} option:
20982 gnat check -Pproj -rules -from=my_rules
20986 If the project @code{proj} depends upon other projects, you can perform
20987 checks on the project closure using the @option{-U} option:
20989 gnat check -Pproj -U -rules -from=my_rules
20993 Finally, if not all the units are relevant to a particular main
20994 program in the project closure, you can perform checks for the set
20995 of units needed to create a given main program (unit closure) using
20996 the @option{-U} option followed by the name of the main unit:
20998 gnat check -Pproj -U main -rules -from=my_rules
21002 @node Rule exemption
21003 @section Rule exemption
21004 @cindex Rule exemption (for @command{gnatcheck})
21007 One of the most useful applications of @command{gnatcheck} is to
21008 automate the enforcement of project-specific coding standards,
21009 for example in safety-critical systems where particular features
21010 must be restricted in order to simplify the certification effort.
21011 However, it may sometimes be appropriate to violate a coding standard rule,
21012 and in such cases the rationale for the violation should be provided
21013 in the source program itself so that the individuals
21014 reviewing or maintaining the program can immediately understand the intent.
21016 The @command{gnatcheck} tool supports this practice with the notion of
21017 a ``rule exemption'' covering a specific source code section. Normally
21018 rule violation messages are issued both on @file{stderr}
21019 and in a report file. In contrast, exempted violations are not listed on
21020 @file{stderr}; thus users invoking @command{gnatcheck} interactively
21021 (e.g. in its GPS interface) do not need to pay attention to known and
21022 justified violations. However, exempted violations along with their
21023 justification are documented in a special section of the report file that
21024 @command{gnatcheck} generates.
21027 * Using pragma Annotate to Control Rule Exemption::
21028 * gnatcheck Annotations Rules::
21031 @node Using pragma Annotate to Control Rule Exemption
21032 @subsection Using pragma @code{Annotate} to Control Rule Exemption
21033 @cindex Using pragma Annotate to control rule exemption
21036 Rule exemption is controlled by pragma @code{Annotate} when its first
21037 argument is ``gnatcheck''. The syntax of @command{gnatcheck}'s
21038 exemption control annotations is as follows:
21040 @smallexample @c ada
21042 pragma Annotate (gnatcheck, @i{exemption_control}, @i{Rule_Name}, [@i{justification}]);
21044 @i{exemption_control} ::= Exempt_On | Exempt_Off
21046 @i{Rule_Name} ::= string_literal
21048 @i{justification} ::= string_literal
21053 When a @command{gnatcheck} annotation has more then four arguments,
21054 @command{gnatcheck} issues a warning and ignores the additional arguments.
21055 If the additional arguments do not follow the syntax above,
21056 @command{gnatcheck} emits a warning and ignores the annotation.
21058 The @i{@code{Rule_Name}} argument should be the name of some existing
21059 @command{gnatcheck} rule.
21060 Otherwise a warning message is generated and the pragma is
21061 ignored. If @code{Rule_Name} denotes a rule that is not activated by the given
21062 @command{gnatcheck} call, the pragma is ignored and no warning is issued.
21064 A source code section where an exemption is active for a given rule is
21065 delimited by an @code{exempt_on} and @code{exempt_off} annotation pair:
21067 @smallexample @c ada
21068 pragma Annotate (gnatcheck, Exempt_On, Rule_Name, "justification");
21069 -- source code section
21070 pragma Annotate (gnatcheck, Exempt_Off, Rule_Name);
21074 @node gnatcheck Annotations Rules
21075 @subsection @command{gnatcheck} Annotations Rules
21076 @cindex @command{gnatcheck} annotations rules
21081 An ``Exempt_Off'' annotation can only appear after a corresponding
21082 ``Exempt_On'' annotation.
21085 Exempted source code sections are only based on the source location of the
21086 annotations. Any source construct between the two
21087 annotations is part of the exempted source code section.
21090 Exempted source code sections for different rules are independent. They can
21091 be nested or intersect with one another without limitation.
21092 Creating nested or intersecting source code sections for the same rule is
21096 Malformed exempted source code sections are reported by a warning, and
21097 the corresponding rule exemptions are ignored.
21100 When an exempted source code section does not contain at least one violation
21101 of the exempted rule, a warning is emitted on @file{stderr}.
21104 If an ``Exempt_On'' annotation pragma does not have a matching
21105 ``Exempt_Off'' annotation pragma in the same compilation unit, then the
21106 exemption for the given rule is ignored and a warning is issued.
21110 @node Predefined Rules
21111 @section Predefined Rules
21112 @cindex Predefined rules (for @command{gnatcheck})
21115 @c (Jan 2007) Since the global rules are still under development and are not
21116 @c documented, there is no point in explaining the difference between
21117 @c global and local rules
21119 A rule in @command{gnatcheck} is either local or global.
21120 A @emph{local rule} is a rule that applies to a well-defined section
21121 of a program and that can be checked by analyzing only this section.
21122 A @emph{global rule} requires analysis of some global properties of the
21123 whole program (mostly related to the program call graph).
21124 As of @value{NOW}, the implementation of global rules should be
21125 considered to be at a preliminary stage. You can use the
21126 @option{+GLOBAL} option to enable all the global rules, and the
21127 @option{-GLOBAL} rule option to disable all the global rules.
21129 All the global rules in the list below are
21130 so indicated by marking them ``GLOBAL''.
21131 This +GLOBAL and -GLOBAL options are not
21132 included in the list of gnatcheck options above, because at the moment they
21133 are considered as a temporary debug options.
21135 @command{gnatcheck} performs rule checks for generic
21136 instances only for global rules. This limitation may be relaxed in a later
21141 The following subsections document the rules implemented in
21142 @command{gnatcheck}.
21143 The subsection title is the same as the rule identifier, which may be
21144 used as a parameter of the @option{+R} or @option{-R} options.
21148 * Abstract_Type_Declarations::
21149 * Anonymous_Arrays::
21150 * Anonymous_Subtypes::
21152 * Boolean_Relational_Operators::
21154 * Ceiling_Violations::
21156 * Complex_Inlined_Subprograms::
21157 * Controlled_Type_Declarations::
21158 * Declarations_In_Blocks::
21159 * Deep_Inheritance_Hierarchies::
21160 * Deeply_Nested_Generics::
21161 * Deeply_Nested_Inlining::
21163 * Deeply_Nested_Local_Inlining::
21165 * Default_Parameters::
21166 * Direct_Calls_To_Primitives::
21167 * Discriminated_Records::
21168 * Enumeration_Ranges_In_CASE_Statements::
21169 * Exceptions_As_Control_Flow::
21170 * Exits_From_Conditional_Loops::
21171 * EXIT_Statements_With_No_Loop_Name::
21172 * Expanded_Loop_Exit_Names::
21173 * Explicit_Full_Discrete_Ranges::
21174 * Float_Equality_Checks::
21175 * Forbidden_Attributes::
21176 * Forbidden_Pragmas::
21177 * Function_Style_Procedures::
21178 * Generics_In_Subprograms::
21179 * GOTO_Statements::
21180 * Implicit_IN_Mode_Parameters::
21181 * Implicit_SMALL_For_Fixed_Point_Types::
21182 * Improperly_Located_Instantiations::
21183 * Improper_Returns::
21184 * Library_Level_Subprograms::
21187 * Improperly_Called_Protected_Entries::
21190 * Misnamed_Controlling_Parameters::
21191 * Misnamed_Identifiers::
21192 * Multiple_Entries_In_Protected_Definitions::
21194 * Non_Qualified_Aggregates::
21195 * Non_Short_Circuit_Operators::
21196 * Non_SPARK_Attributes::
21197 * Non_Tagged_Derived_Types::
21198 * Non_Visible_Exceptions::
21199 * Numeric_Literals::
21200 * OTHERS_In_Aggregates::
21201 * OTHERS_In_CASE_Statements::
21202 * OTHERS_In_Exception_Handlers::
21203 * Outer_Loop_Exits::
21204 * Overloaded_Operators::
21205 * Overly_Nested_Control_Structures::
21206 * Parameters_Out_Of_Order::
21207 * Positional_Actuals_For_Defaulted_Generic_Parameters::
21208 * Positional_Actuals_For_Defaulted_Parameters::
21209 * Positional_Components::
21210 * Positional_Generic_Parameters::
21211 * Positional_Parameters::
21212 * Predefined_Numeric_Types::
21213 * Raising_External_Exceptions::
21214 * Raising_Predefined_Exceptions::
21215 * Separate_Numeric_Error_Handlers::
21218 * Side_Effect_Functions::
21221 * Too_Many_Parents::
21222 * Unassigned_OUT_Parameters::
21223 * Uncommented_BEGIN_In_Package_Bodies::
21224 * Unconditional_Exits::
21225 * Unconstrained_Array_Returns::
21226 * Universal_Ranges::
21227 * Unnamed_Blocks_And_Loops::
21229 * Unused_Subprograms::
21231 * USE_PACKAGE_Clauses::
21232 * Visible_Components::
21233 * Volatile_Objects_Without_Address_Clauses::
21237 @node Abstract_Type_Declarations
21238 @subsection @code{Abstract_Type_Declarations}
21239 @cindex @code{Abstract_Type_Declarations} rule (for @command{gnatcheck})
21242 Flag all declarations of abstract types. For an abstract private
21243 type, both the private and full type declarations are flagged.
21245 This rule has no parameters.
21248 @node Anonymous_Arrays
21249 @subsection @code{Anonymous_Arrays}
21250 @cindex @code{Anonymous_Arrays} rule (for @command{gnatcheck})
21253 Flag all anonymous array type definitions (by Ada semantics these can only
21254 occur in object declarations).
21256 This rule has no parameters.
21258 @node Anonymous_Subtypes
21259 @subsection @code{Anonymous_Subtypes}
21260 @cindex @code{Anonymous_Subtypes} rule (for @command{gnatcheck})
21263 Flag all uses of anonymous subtypes (except cases when subtype indication
21264 is a part of a record component definition, and this subtype indication
21265 depends on a discriminant). A use of an anonymous subtype is
21266 any instance of a subtype indication with a constraint, other than one
21267 that occurs immediately within a subtype declaration. Any use of a range
21268 other than as a constraint used immediately within a subtype declaration
21269 is considered as an anonymous subtype.
21271 An effect of this rule is that @code{for} loops such as the following are
21272 flagged (since @code{1..N} is formally a ``range''):
21274 @smallexample @c ada
21275 for I in 1 .. N loop
21281 Declaring an explicit subtype solves the problem:
21283 @smallexample @c ada
21284 subtype S is Integer range 1..N;
21292 This rule has no parameters.
21295 @subsection @code{Blocks}
21296 @cindex @code{Blocks} rule (for @command{gnatcheck})
21299 Flag each block statement.
21301 This rule has no parameters.
21303 @node Boolean_Relational_Operators
21304 @subsection @code{Boolean_Relational_Operators}
21305 @cindex @code{Boolean_Relational_Operators} rule (for @command{gnatcheck})
21308 Flag each call to a predefined relational operator (``<'', ``>'', ``<='',
21309 ``>='', ``='' and ``/='') for the predefined Boolean type.
21310 (This rule is useful in enforcing the SPARK language restrictions.)
21312 Calls to predefined relational operators of any type derived from
21313 @code{Standard.Boolean} are not detected. Calls to user-defined functions
21314 with these designators, and uses of operators that are renamings
21315 of the predefined relational operators for @code{Standard.Boolean},
21316 are likewise not detected.
21318 This rule has no parameters.
21321 @node Ceiling_Violations
21322 @subsection @code{Ceiling5_Violations} (under construction, GLOBAL)
21323 @cindex @code{Ceiling_Violations} rule (for @command{gnatcheck})
21326 Flag invocations of a protected operation by a task whose priority exceeds
21327 the protected object's ceiling.
21329 As of @value{NOW}, this rule has the following limitations:
21334 We consider only pragmas Priority and Interrupt_Priority as means to define
21335 a task/protected operation priority. We do not consider the effect of using
21336 Ada.Dynamic_Priorities.Set_Priority procedure;
21339 We consider only base task priorities, and no priority inheritance. That is,
21340 we do not make a difference between calls issued during task activation and
21341 execution of the sequence of statements from task body;
21344 Any situation when the priority of protected operation caller is set by a
21345 dynamic expression (that is, the corresponding Priority or
21346 Interrupt_Priority pragma has a non-static expression as an argument) we
21347 treat as a priority inconsistency (and, therefore, detect this situation).
21351 At the moment the notion of the main subprogram is not implemented in
21352 gnatcheck, so any pragma Priority in a library level subprogram body (in case
21353 if this subprogram can be a main subprogram of a partition) changes the
21354 priority of an environment task. So if we have more then one such pragma in
21355 the set of processed sources, the pragma that is processed last, defines the
21356 priority of an environment task.
21358 This rule has no parameters.
21361 @node Controlled_Type_Declarations
21362 @subsection @code{Controlled_Type_Declarations}
21363 @cindex @code{Controlled_Type_Declarations} rule (for @command{gnatcheck})
21366 Flag all declarations of controlled types. A declaration of a private type
21367 is flagged if its full declaration declares a controlled type. A declaration
21368 of a derived type is flagged if its ancestor type is controlled. Subtype
21369 declarations are not checked. A declaration of a type that itself is not a
21370 descendant of a type declared in @code{Ada.Finalization} but has a controlled
21371 component is not checked.
21373 This rule has no parameters.
21376 @node Complex_Inlined_Subprograms
21377 @subsection @code{Complex_Inlined_Subprograms}
21378 @cindex @code{Complex_Inlined_Subprograms} rule (for @command{gnatcheck})
21381 Flags a subprogram (or generic subprogram) if
21382 pragma Inline is applied to the subprogram and at least one of the following
21387 it contains at least one complex declaration such as a subprogram body,
21388 package, task, protected declaration, or a generic instantiation
21389 (except instantiation of @code{Ada.Unchecked_Conversion});
21392 it contains at least one complex statement such as a loop, a case
21393 or a if statement, or a short circuit control form;
21396 the number of statements exceeds
21397 a value specified by the @option{N} rule parameter;
21401 This rule has the following (mandatory) parameter for the @option{+R} option:
21405 Positive integer specifying the maximum allowed total number of statements
21406 in the subprogram body.
21410 @node Declarations_In_Blocks
21411 @subsection @code{Declarations_In_Blocks}
21412 @cindex @code{Declarations_In_Blocks} rule (for @command{gnatcheck})
21415 Flag all block statements containing local declarations. A @code{declare}
21416 block with an empty @i{declarative_part} or with a @i{declarative part}
21417 containing only pragmas and/or @code{use} clauses is not flagged.
21419 This rule has no parameters.
21422 @node Deep_Inheritance_Hierarchies
21423 @subsection @code{Deep_Inheritance_Hierarchies}
21424 @cindex @code{Deep_Inheritance_Hierarchies} rule (for @command{gnatcheck})
21427 Flags a tagged derived type declaration or an interface type declaration if
21428 its depth (in its inheritance
21429 hierarchy) exceeds the value specified by the @option{N} rule parameter.
21431 The inheritance depth of a tagged type or interface type is defined as 0 for
21432 a type with no parent and no progenitor, and otherwise as 1 + max of the
21433 depths of the immediate parent and immediate progenitors.
21435 This rule does not flag private extension
21436 declarations. In the case of a private extension, the corresponding full
21437 declaration is checked.
21439 This rule has the following (mandatory) parameter for the @option{+R} option:
21443 Integer not less than -1 specifying the maximal allowed depth of any inheritance
21444 hierarchy. If the rule parameter is set to -1, the rule flags all the declarations
21445 of tagged and interface types.
21449 @node Deeply_Nested_Generics
21450 @subsection @code{Deeply_Nested_Generics}
21451 @cindex @code{Deeply_Nested_Generics} rule (for @command{gnatcheck})
21454 Flags a generic declaration nested in another generic declaration if
21455 the nesting level of the inner generic exceeds
21456 a value specified by the @option{N} rule parameter.
21457 The nesting level is the number of generic declaratons that enclose the given
21458 (generic) declaration. Formal packages are not flagged by this rule.
21460 This rule has the following (mandatory) parameters for the @option{+R} option:
21464 Positive integer specifying the maximal allowed nesting level
21465 for a generic declaration.
21468 @node Deeply_Nested_Inlining
21469 @subsection @code{Deeply_Nested_Inlining}
21470 @cindex @code{Deeply_Nested_Inlining} rule (for @command{gnatcheck})
21473 Flags a subprogram (or generic subprogram) if
21474 pragma Inline has been applied to the subprogram but the subprogram
21475 calls to another inlined subprogram that results in nested inlining
21476 with nesting depth exceeding the value specified by the
21477 @option{N} rule parameter.
21479 This rule requires the global analysis of all the compilation units that
21480 are @command{gnatcheck} arguments; such analysis may affect the tool's
21483 This rule has the following (mandatory) parameter for the @option{+R} option:
21487 Positive integer specifying the maximal allowed level of nested inlining.
21492 @node Deeply_Nested_Local_Inlining
21493 @subsection @code{Deeply_Nested_Local_Inlining}
21494 @cindex @code{Deeply_Nested_Local_Inlining} rule (for @command{gnatcheck})
21497 Flags a subprogram body if a pragma @code{Inline} is applied to the
21498 corresponding subprogram (or generic subprogram) and the body contains a call
21499 to another inlined subprogram that results in nested inlining with nesting
21500 depth more then a value specified by the @option{N} rule parameter.
21501 This rule is similar to @code{Deeply_Nested_Inlining} rule, but it
21502 assumes that calls to subprograms in
21503 with'ed units are not inlided, so all the analysis of the depth of inlining is
21504 limited by the compilation unit where the subprogram body is located and the
21505 units it depends semantically upon. Such analysis may be usefull for the case
21506 when neiter @option{-gnatn} nor @option{-gnatN} option is used when building
21509 This rule has the following (mandatory) parameters for the @option{+R} option:
21513 Positive integer specifying the maximal allowed level of nested inlining.
21518 @node Default_Parameters
21519 @subsection @code{Default_Parameters}
21520 @cindex @code{Default_Parameters} rule (for @command{gnatcheck})
21523 Flag all default expressions for subprogram parameters. Parameter
21524 declarations of formal and generic subprograms are also checked.
21526 This rule has no parameters.
21529 @node Direct_Calls_To_Primitives
21530 @subsection @code{Direct_Calls_To_Primitives}
21531 @cindex @code{Direct_Calls_To_Primitives} rule (for @command{gnatcheck})
21534 Flags any non-dispatching call to a dispatching primitive operation, except
21535 for the common idiom where a primitive subprogram for a tagged type
21536 directly calls the same primitive subprogram of the type's immediate ancestor.
21538 This rule has no parameters.
21541 @node Discriminated_Records
21542 @subsection @code{Discriminated_Records}
21543 @cindex @code{Discriminated_Records} rule (for @command{gnatcheck})
21546 Flag all declarations of record types with discriminants. Only the
21547 declarations of record and record extension types are checked. Incomplete,
21548 formal, private, derived and private extension type declarations are not
21549 checked. Task and protected type declarations also are not checked.
21551 This rule has no parameters.
21554 @node Enumeration_Ranges_In_CASE_Statements
21555 @subsection @code{Enumeration_Ranges_In_CASE_Statements}
21556 @cindex @code{Enumeration_Ranges_In_CASE_Statements} (for @command{gnatcheck})
21559 Flag each use of a range of enumeration literals as a choice in a
21560 @code{case} statement.
21561 All forms for specifying a range (explicit ranges
21562 such as @code{A .. B}, subtype marks and @code{'Range} attributes) are flagged.
21563 An enumeration range is
21564 flagged even if contains exactly one enumeration value or no values at all. A
21565 type derived from an enumeration type is considered as an enumeration type.
21567 This rule helps prevent maintenance problems arising from adding an
21568 enumeration value to a type and having it implicitly handled by an existing
21569 @code{case} statement with an enumeration range that includes the new literal.
21571 This rule has no parameters.
21574 @node Exceptions_As_Control_Flow
21575 @subsection @code{Exceptions_As_Control_Flow}
21576 @cindex @code{Exceptions_As_Control_Flow} (for @command{gnatcheck})
21579 Flag each place where an exception is explicitly raised and handled in the
21580 same subprogram body. A @code{raise} statement in an exception handler,
21581 package body, task body or entry body is not flagged.
21583 The rule has no parameters.
21585 @node Exits_From_Conditional_Loops
21586 @subsection @code{Exits_From_Conditional_Loops}
21587 @cindex @code{Exits_From_Conditional_Loops} (for @command{gnatcheck})
21590 Flag any exit statement if it transfers the control out of a @code{for} loop
21591 or a @code{while} loop. This includes cases when the @code{exit} statement
21592 applies to a @code{FOR} or @code{while} loop, and cases when it is enclosed
21593 in some @code{for} or @code{while} loop, but transfers the control from some
21594 outer (inconditional) @code{loop} statement.
21596 The rule has no parameters.
21599 @node EXIT_Statements_With_No_Loop_Name
21600 @subsection @code{EXIT_Statements_With_No_Loop_Name}
21601 @cindex @code{EXIT_Statements_With_No_Loop_Name} (for @command{gnatcheck})
21604 Flag each @code{exit} statement that does not specify the name of the loop
21607 The rule has no parameters.
21610 @node Expanded_Loop_Exit_Names
21611 @subsection @code{Expanded_Loop_Exit_Names}
21612 @cindex @code{Expanded_Loop_Exit_Names} rule (for @command{gnatcheck})
21615 Flag all expanded loop names in @code{exit} statements.
21617 This rule has no parameters.
21619 @node Explicit_Full_Discrete_Ranges
21620 @subsection @code{Explicit_Full_Discrete_Ranges}
21621 @cindex @code{Explicit_Full_Discrete_Ranges} rule (for @command{gnatcheck})
21624 Flag each discrete range that has the form @code{A'First .. A'Last}.
21626 This rule has no parameters.
21628 @node Float_Equality_Checks
21629 @subsection @code{Float_Equality_Checks}
21630 @cindex @code{Float_Equality_Checks} rule (for @command{gnatcheck})
21633 Flag all calls to the predefined equality operations for floating-point types.
21634 Both ``@code{=}'' and ``@code{/=}'' operations are checked.
21635 User-defined equality operations are not flagged, nor are ``@code{=}''
21636 and ``@code{/=}'' operations for fixed-point types.
21638 This rule has no parameters.
21641 @node Forbidden_Attributes
21642 @subsection @code{Forbidden_Attributes}
21643 @cindex @code{Forbidden_Attributes} rule (for @command{gnatcheck})
21646 Flag each use of the specified attributes. The attributes to be detected are
21647 named in the rule's parameters.
21649 This rule has the following parameters:
21652 @item For the @option{+R} option
21655 @item @emph{Attribute_Designator}
21656 Adds the specified attribute to the set of attributes to be detected and sets
21657 the detection checks for all the specified attributes ON.
21658 If @emph{Attribute_Designator}
21659 does not denote any attribute defined in the Ada standard
21661 @ref{Implementation Defined Attributes,,, gnat_rm, GNAT Reference
21662 Manual}, it is treated as the name of unknown attribute.
21665 All the GNAT-specific attributes are detected; this sets
21666 the detection checks for all the specified attributes ON.
21669 All attributes are detected; this sets the rule ON.
21672 @item For the @option{-R} option
21674 @item @emph{Attribute_Designator}
21675 Removes the specified attribute from the set of attributes to be
21676 detected without affecting detection checks for
21677 other attributes. If @emph{Attribute_Designator} does not correspond to any
21678 attribute defined in the Ada standard or in
21679 @ref{Implementation Defined Attributes,,, gnat_rm, GNAT Reference Manual},
21680 this option is treated as turning OFF detection of all unknown attributes.
21683 Turn OFF detection of all GNAT-specific attributes
21686 Clear the list of the attributes to be detected and
21692 Parameters are not case sensitive. If @emph{Attribute_Designator} does not
21693 have the syntax of an Ada identifier and therefore can not be considered as a
21694 (part of an) attribute designator, a diagnostic message is generated and the
21695 corresponding parameter is ignored. (If an attribute allows a static
21696 expression to be a part of the attribute designator, this expression is
21697 ignored by this rule.)
21699 When more then one parameter is given in the same rule option, the parameters
21700 must be separated by commas.
21702 If more then one option for this rule is specified for the gnatcheck call, a
21703 new option overrides the previous one(s).
21705 The @option{+R} option with no parameters turns the rule ON, with the set of
21706 attributes to be detected defined by the previous rule options.
21707 (By default this set is empty, so if the only option specified for the rule is
21708 @option{+RForbidden_Attributes} (with
21709 no parameter), then the rule is enabled, but it does not detect anything).
21710 The @option{-R} option with no parameter turns the rule OFF, but it does not
21711 affect the set of attributes to be detected.
21714 @node Forbidden_Pragmas
21715 @subsection @code{Forbidden_Pragmas}
21716 @cindex @code{Forbidden_Pragmas} rule (for @command{gnatcheck})
21719 Flag each use of the specified pragmas. The pragmas to be detected
21720 are named in the rule's parameters.
21722 This rule has the following parameters:
21725 @item For the @option{+R} option
21728 @item @emph{Pragma_Name}
21729 Adds the specified pragma to the set of pragmas to be
21730 checked and sets the checks for all the specified pragmas
21731 ON. @emph{Pragma_Name} is treated as a name of a pragma. If it
21732 does not correspond to any pragma name defined in the Ada
21733 standard or to the name of a GNAT-specific pragma defined
21734 in @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference
21735 Manual}, it is treated as the name of unknown pragma.
21738 All the GNAT-specific pragmas are detected; this sets
21739 the checks for all the specified pragmas ON.
21742 All pragmas are detected; this sets the rule ON.
21745 @item For the @option{-R} option
21747 @item @emph{Pragma_Name}
21748 Removes the specified pragma from the set of pragmas to be
21749 checked without affecting checks for
21750 other pragmas. @emph{Pragma_Name} is treated as a name
21751 of a pragma. If it does not correspond to any pragma
21752 defined in the Ada standard or to any name defined in
21753 @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
21754 this option is treated as turning OFF detection of all unknown pragmas.
21757 Turn OFF detection of all GNAT-specific pragmas
21760 Clear the list of the pragmas to be detected and
21766 Parameters are not case sensitive. If @emph{Pragma_Name} does not have
21767 the syntax of an Ada identifier and therefore can not be considered
21768 as a pragma name, a diagnostic message is generated and the corresponding
21769 parameter is ignored.
21771 When more then one parameter is given in the same rule option, the parameters
21772 must be separated by a comma.
21774 If more then one option for this rule is specified for the @command{gnatcheck}
21775 call, a new option overrides the previous one(s).
21777 The @option{+R} option with no parameters turns the rule ON with the set of
21778 pragmas to be detected defined by the previous rule options.
21779 (By default this set is empty, so if the only option specified for the rule is
21780 @option{+RForbidden_Pragmas} (with
21781 no parameter), then the rule is enabled, but it does not detect anything).
21782 The @option{-R} option with no parameter turns the rule OFF, but it does not
21783 affect the set of pragmas to be detected.
21788 @node Function_Style_Procedures
21789 @subsection @code{Function_Style_Procedures}
21790 @cindex @code{Function_Style_Procedures} rule (for @command{gnatcheck})
21793 Flag each procedure that can be rewritten as a function. A procedure can be
21794 converted into a function if it has exactly one parameter of mode @code{out}
21795 and no parameters of mode @code{in out}. Procedure declarations,
21796 formal procedure declarations, and generic procedure declarations are always
21798 bodies and body stubs are flagged only if they do not have corresponding
21799 separate declarations. Procedure renamings and procedure instantiations are
21802 If a procedure can be rewritten as a function, but its @code{out} parameter is
21803 of a limited type, it is not flagged.
21805 Protected procedures are not flagged. Null procedures also are not flagged.
21807 This rule has no parameters.
21810 @node Generics_In_Subprograms
21811 @subsection @code{Generics_In_Subprograms}
21812 @cindex @code{Generics_In_Subprograms} rule (for @command{gnatcheck})
21815 Flag each declaration of a generic unit in a subprogram. Generic
21816 declarations in the bodies of generic subprograms are also flagged.
21817 A generic unit nested in another generic unit is not flagged.
21818 If a generic unit is
21819 declared in a local package that is declared in a subprogram body, the
21820 generic unit is flagged.
21822 This rule has no parameters.
21825 @node GOTO_Statements
21826 @subsection @code{GOTO_Statements}
21827 @cindex @code{GOTO_Statements} rule (for @command{gnatcheck})
21830 Flag each occurrence of a @code{goto} statement.
21832 This rule has no parameters.
21835 @node Implicit_IN_Mode_Parameters
21836 @subsection @code{Implicit_IN_Mode_Parameters}
21837 @cindex @code{Implicit_IN_Mode_Parameters} rule (for @command{gnatcheck})
21840 Flag each occurrence of a formal parameter with an implicit @code{in} mode.
21841 Note that @code{access} parameters, although they technically behave
21842 like @code{in} parameters, are not flagged.
21844 This rule has no parameters.
21847 @node Implicit_SMALL_For_Fixed_Point_Types
21848 @subsection @code{Implicit_SMALL_For_Fixed_Point_Types}
21849 @cindex @code{Implicit_SMALL_For_Fixed_Point_Types} rule (for @command{gnatcheck})
21852 Flag each fixed point type declaration that lacks an explicit
21853 representation clause to define its @code{'Small} value.
21854 Since @code{'Small} can be defined only for ordinary fixed point types,
21855 decimal fixed point type declarations are not checked.
21857 This rule has no parameters.
21860 @node Improperly_Located_Instantiations
21861 @subsection @code{Improperly_Located_Instantiations}
21862 @cindex @code{Improperly_Located_Instantiations} rule (for @command{gnatcheck})
21865 Flag all generic instantiations in library-level package specs
21866 (including library generic packages) and in all subprogram bodies.
21868 Instantiations in task and entry bodies are not flagged. Instantiations in the
21869 bodies of protected subprograms are flagged.
21871 This rule has no parameters.
21875 @node Improper_Returns
21876 @subsection @code{Improper_Returns}
21877 @cindex @code{Improper_Returns} rule (for @command{gnatcheck})
21880 Flag each explicit @code{return} statement in procedures, and
21881 multiple @code{return} statements in functions.
21882 Diagnostic messages are generated for all @code{return} statements
21883 in a procedure (thus each procedure must be written so that it
21884 returns implicitly at the end of its statement part),
21885 and for all @code{return} statements in a function after the first one.
21886 This rule supports the stylistic convention that each subprogram
21887 should have no more than one point of normal return.
21889 This rule has no parameters.
21892 @node Library_Level_Subprograms
21893 @subsection @code{Library_Level_Subprograms}
21894 @cindex @code{Library_Level_Subprograms} rule (for @command{gnatcheck})
21897 Flag all library-level subprograms (including generic subprogram instantiations).
21899 This rule has no parameters.
21902 @node Local_Packages
21903 @subsection @code{Local_Packages}
21904 @cindex @code{Local_Packages} rule (for @command{gnatcheck})
21907 Flag all local packages declared in package and generic package
21909 Local packages in bodies are not flagged.
21911 This rule has no parameters.
21914 @node Improperly_Called_Protected_Entries
21915 @subsection @code{Improperly_Called_Protected_Entries} (under construction, GLOBAL)
21916 @cindex @code{Improperly_Called_Protected_Entries} rule (for @command{gnatcheck})
21919 Flag each protected entry that can be called from more than one task.
21921 This rule has no parameters.
21925 @subsection @code{Metrics}
21926 @cindex @code{Metrics} rule (for @command{gnatcheck})
21929 There is a set of checks based on computing a metric value and comparing the
21930 result with the specified upper (or lower, depending on a specific metric)
21931 value specified for a given metric. A construct is flagged if a given metric
21932 is applicable (can be computed) for it and the computed value is greater
21933 then (lover then) the specified upper (lower) bound.
21935 The name of any metric-based rule consists of the prefix @code{Metrics_}
21936 followed by the name of the corresponding metric (see the table below).
21937 For @option{+R} option, each metric-based rule has a numeric parameter
21938 specifying the bound (integer or real, depending on a metric), @option{-R}
21939 option for metric rules does not have a parameter.
21941 The following table shows the metric names for that the corresponding
21942 metrics-based checks are supported by gnatcheck, including the
21943 constraint that must be satisfied by the bound that is specified for the check
21944 and what bound - upper (U) or lower (L) - should be specified.
21946 @multitable {@code{Cyclomatic_Complexity}}{Cyclomatic complexity}{Positive integer}
21948 @headitem Check Name @tab Description @tab Bounds Value
21951 @item @b{Check Name} @tab @b{Description} @tab @b{Bounds Value}
21953 @c Above conditional code is workaround to bug in texi2html (Feb 2008)
21954 @item @code{Essential_Complexity} @tab Essential complexity @tab Positive integer (U)
21955 @item @code{Cyclomatic_Complexity} @tab Cyclomatic complexity @tab Positive integer (U)
21956 @item @code{LSLOC} @tab Logical Source Lines of Code @tab Positive integer (U)
21960 The meaning and the computed values for all these metrics are exactly
21961 the same as for the corresponding metrics in @command{gnatmetric}.
21963 @emph{Example:} the rule
21965 +RMetrics_Cyclomatic_Complexity : 7
21968 means that all bodies with cyclomatic complexity exceeding 7 will be flagged.
21970 To turn OFF the check for cyclomatic complexity metric, use the following option:
21972 -RMetrics_Cyclomatic_Complexity
21976 @node Misnamed_Controlling_Parameters
21977 @subsection @code{Misnamed_Controlling_Parameters}
21978 @cindex @code{Misnamed_Controlling_Parameters} rule (for @command{gnatcheck})
21981 Flags a declaration of a dispatching operation, if the first parameter is
21982 not a controlling one and its name is not @code{This} (the check for
21983 parameter name is not case-sensitive). Declarations of dispatching functions
21984 with controlling result and no controlling parameter are never flagged.
21986 A subprogram body declaration, subprogram renaming declaration or subprogram
21987 body stub is flagged only if it is not a completion of a prior subprogram
21990 This rule has no parameters.
21994 @node Misnamed_Identifiers
21995 @subsection @code{Misnamed_Identifiers}
21996 @cindex @code{Misnamed_Identifiers} rule (for @command{gnatcheck})
21999 Flag the declaration of each identifier that does not have a suffix
22000 corresponding to the kind of entity being declared.
22001 The following declarations are checked:
22008 subtype declarations
22011 constant declarations (but not number declarations)
22014 package renaming declarations (but not generic package renaming
22019 This rule may have parameters. When used without parameters, the rule enforces
22020 the following checks:
22024 type-defining names end with @code{_T}, unless the type is an access type,
22025 in which case the suffix must be @code{_A}
22027 constant names end with @code{_C}
22029 names defining package renamings end with @code{_R}
22033 Defining identifiers from incomplete type declarations are never flagged.
22035 For a private type declaration (including private extensions), the defining
22036 identifier from the private type declaration is checked against the type
22037 suffix (even if the corresponding full declaration is an access type
22038 declaration), and the defining identifier from the corresponding full type
22039 declaration is not checked.
22042 For a deferred constant, the defining name in the corresponding full constant
22043 declaration is not checked.
22045 Defining names of formal types are not checked.
22047 The rule may have the following parameters:
22051 For the @option{+R} option:
22054 Sets the default listed above for all the names to be checked.
22056 @item Type_Suffix=@emph{string}
22057 Specifies the suffix for a type name.
22059 @item Access_Suffix=@emph{string}
22060 Specifies the suffix for an access type name. If
22061 this parameter is set, it overrides for access
22062 types the suffix set by the @code{Type_Suffix} parameter.
22063 For access types, @emph{string} may have the following format:
22064 @emph{suffix1(suffix2)}. That means that an access type name
22065 should have the @emph{suffix1} suffix except for the case when
22066 the designated type is also an access type, in this case the
22067 type name should have the @emph{suffix1 & suffix2} suffix.
22069 @item Class_Access_Suffix=@emph{string}
22070 Specifies the suffix for the name of an access type that points to some class-wide
22071 type. If this parameter is set, it overrides for such access
22072 types the suffix set by the @code{Type_Suffix} or @code{Access_Suffix}
22075 @item Class_Subtype_Suffix=@emph{string}
22076 Specifies the suffix for the name of a subtype that denotes a class-wide type.
22078 @item Constant_Suffix=@emph{string}
22079 Specifies the suffix for a constant name.
22081 @item Renaming_Suffix=@emph{string}
22082 Specifies the suffix for a package renaming name.
22086 For the @option{-R} option:
22089 Remove all the suffixes specified for the
22090 identifier suffix checks, whether by default or
22091 as specified by other rule parameters. All the
22092 checks for this rule are disabled as a result.
22095 Removes the suffix specified for types. This
22096 disables checks for types but does not disable
22097 any other checks for this rule (including the
22098 check for access type names if @code{Access_Suffix} is
22101 @item Access_Suffix
22102 Removes the suffix specified for access types.
22103 This disables checks for access type names but
22104 does not disable any other checks for this rule.
22105 If @code{Type_Suffix} is set, access type names are
22106 checked as ordinary type names.
22108 @item Class_Access_Suffix
22109 Removes the suffix specified for access types pointing to class-wide
22110 type. This disables specific checks for names of access types pointing to
22111 class-wide types but does not disable any other checks for this rule.
22112 If @code{Type_Suffix} is set, access type names are
22113 checked as ordinary type names. If @code{Access_Suffix} is set, these
22114 access types are checked as any other access type name.
22116 @item Class_Subtype_Suffix=@emph{string}
22117 Removes the suffix specified for subtype names.
22118 This disables checks for subtype names but
22119 does not disable any other checks for this rule.
22121 @item Constant_Suffix
22122 Removes the suffix specified for constants. This
22123 disables checks for constant names but does not
22124 disable any other checks for this rule.
22126 @item Renaming_Suffix
22127 Removes the suffix specified for package
22128 renamings. This disables checks for package
22129 renamings but does not disable any other checks
22135 If more than one parameter is used, parameters must be separated by commas.
22137 If more than one option is specified for the @command{gnatcheck} invocation,
22138 a new option overrides the previous one(s).
22140 The @option{+RMisnamed_Identifiers} option (with no parameter) enables
22142 name suffixes specified by previous options used for this rule.
22144 The @option{-RMisnamed_Identifiers} option (with no parameter) disables
22145 all the checks but keeps
22146 all the suffixes specified by previous options used for this rule.
22148 The @emph{string} value must be a valid suffix for an Ada identifier (after
22149 trimming all the leading and trailing space characters, if any).
22150 Parameters are not case sensitive, except the @emph{string} part.
22152 If any error is detected in a rule parameter, the parameter is ignored.
22153 In such a case the options that are set for the rule are not
22158 @node Multiple_Entries_In_Protected_Definitions
22159 @subsection @code{Multiple_Entries_In_Protected_Definitions}
22160 @cindex @code{Multiple_Entries_In_Protected_Definitions} rule (for @command{gnatcheck})
22163 Flag each protected definition (i.e., each protected object/type declaration)
22164 that defines more than one entry.
22165 Diagnostic messages are generated for all the entry declarations
22166 except the first one. An entry family is counted as one entry. Entries from
22167 the private part of the protected definition are also checked.
22169 This rule has no parameters.
22172 @subsection @code{Name_Clashes}
22173 @cindex @code{Name_Clashes} rule (for @command{gnatcheck})
22176 Check that certain names are not used as defining identifiers. To activate
22177 this rule, you need to supply a reference to the dictionary file(s) as a rule
22178 parameter(s) (more then one dictionary file can be specified). If no
22179 dictionary file is set, this rule will not cause anything to be flagged.
22180 Only defining occurrences, not references, are checked.
22181 The check is not case-sensitive.
22183 This rule is enabled by default, but without setting any corresponding
22184 dictionary file(s); thus the default effect is to do no checks.
22186 A dictionary file is a plain text file. The maximum line length for this file
22187 is 1024 characters. If the line is longer then this limit, extra characters
22190 Each line can be either an empty line, a comment line, or a line containing
22191 a list of identifiers separated by space or HT characters.
22192 A comment is an Ada-style comment (from @code{--} to end-of-line).
22193 Identifiers must follow the Ada syntax for identifiers.
22194 A line containing one or more identifiers may end with a comment.
22196 @node Non_Qualified_Aggregates
22197 @subsection @code{Non_Qualified_Aggregates}
22198 @cindex @code{Non_Qualified_Aggregates} rule (for @command{gnatcheck})
22201 Flag each non-qualified aggregate.
22202 A non-qualified aggregate is an
22203 aggregate that is not the expression of a qualified expression. A
22204 string literal is not considered an aggregate, but an array
22205 aggregate of a string type is considered as a normal aggregate.
22206 Aggregates of anonymous array types are not flagged.
22208 This rule has no parameters.
22211 @node Non_Short_Circuit_Operators
22212 @subsection @code{Non_Short_Circuit_Operators}
22213 @cindex @code{Non_Short_Circuit_Operators} rule (for @command{gnatcheck})
22216 Flag all calls to predefined @code{and} and @code{or} operators for
22217 any boolean type. Calls to
22218 user-defined @code{and} and @code{or} and to operators defined by renaming
22219 declarations are not flagged. Calls to predefined @code{and} and @code{or}
22220 operators for modular types or boolean array types are not flagged.
22222 This rule has no parameters.
22226 @node Non_SPARK_Attributes
22227 @subsection @code{Non_SPARK_Attributes}
22228 @cindex @code{Non_SPARK_Attributes} rule (for @command{gnatcheck})
22231 The SPARK language defines the following subset of Ada 95 attribute
22232 designators as those that can be used in SPARK programs. The use of
22233 any other attribute is flagged.
22236 @item @code{'Adjacent}
22239 @item @code{'Ceiling}
22240 @item @code{'Component_Size}
22241 @item @code{'Compose}
22242 @item @code{'Copy_Sign}
22243 @item @code{'Delta}
22244 @item @code{'Denorm}
22245 @item @code{'Digits}
22246 @item @code{'Exponent}
22247 @item @code{'First}
22248 @item @code{'Floor}
22250 @item @code{'Fraction}
22252 @item @code{'Leading_Part}
22253 @item @code{'Length}
22254 @item @code{'Machine}
22255 @item @code{'Machine_Emax}
22256 @item @code{'Machine_Emin}
22257 @item @code{'Machine_Mantissa}
22258 @item @code{'Machine_Overflows}
22259 @item @code{'Machine_Radix}
22260 @item @code{'Machine_Rounds}
22263 @item @code{'Model}
22264 @item @code{'Model_Emin}
22265 @item @code{'Model_Epsilon}
22266 @item @code{'Model_Mantissa}
22267 @item @code{'Model_Small}
22268 @item @code{'Modulus}
22271 @item @code{'Range}
22272 @item @code{'Remainder}
22273 @item @code{'Rounding}
22274 @item @code{'Safe_First}
22275 @item @code{'Safe_Last}
22276 @item @code{'Scaling}
22277 @item @code{'Signed_Zeros}
22279 @item @code{'Small}
22281 @item @code{'Truncation}
22282 @item @code{'Unbiased_Rounding}
22284 @item @code{'Valid}
22288 This rule has no parameters.
22291 @node Non_Tagged_Derived_Types
22292 @subsection @code{Non_Tagged_Derived_Types}
22293 @cindex @code{Non_Tagged_Derived_Types} rule (for @command{gnatcheck})
22296 Flag all derived type declarations that do not have a record extension part.
22298 This rule has no parameters.
22302 @node Non_Visible_Exceptions
22303 @subsection @code{Non_Visible_Exceptions}
22304 @cindex @code{Non_Visible_Exceptions} rule (for @command{gnatcheck})
22307 Flag constructs leading to the possibility of propagating an exception
22308 out of the scope in which the exception is declared.
22309 Two cases are detected:
22313 An exception declaration in a subprogram body, task body or block
22314 statement is flagged if the body or statement does not contain a handler for
22315 that exception or a handler with an @code{others} choice.
22318 A @code{raise} statement in an exception handler of a subprogram body,
22319 task body or block statement is flagged if it (re)raises a locally
22320 declared exception. This may occur under the following circumstances:
22323 it explicitly raises a locally declared exception, or
22325 it does not specify an exception name (i.e., it is simply @code{raise;})
22326 and the enclosing handler contains a locally declared exception in its
22332 Renamings of local exceptions are not flagged.
22334 This rule has no parameters.
22337 @node Numeric_Literals
22338 @subsection @code{Numeric_Literals}
22339 @cindex @code{Numeric_Literals} rule (for @command{gnatcheck})
22342 Flag each use of a numeric literal in an index expression, and in any
22343 circumstance except for the following:
22347 a literal occurring in the initialization expression for a constant
22348 declaration or a named number declaration, or
22351 an integer literal that is less than or equal to a value
22352 specified by the @option{N} rule parameter.
22356 This rule may have the following parameters for the @option{+R} option:
22360 @emph{N} is an integer literal used as the maximal value that is not flagged
22361 (i.e., integer literals not exceeding this value are allowed)
22364 All integer literals are flagged
22368 If no parameters are set, the maximum unflagged value is 1.
22370 The last specified check limit (or the fact that there is no limit at
22371 all) is used when multiple @option{+R} options appear.
22373 The @option{-R} option for this rule has no parameters.
22374 It disables the rule but retains the last specified maximum unflagged value.
22375 If the @option{+R} option subsequently appears, this value is used as the
22376 threshold for the check.
22379 @node OTHERS_In_Aggregates
22380 @subsection @code{OTHERS_In_Aggregates}
22381 @cindex @code{OTHERS_In_Aggregates} rule (for @command{gnatcheck})
22384 Flag each use of an @code{others} choice in extension aggregates.
22385 In record and array aggregates, an @code{others} choice is flagged unless
22386 it is used to refer to all components, or to all but one component.
22388 If, in case of a named array aggregate, there are two associations, one
22389 with an @code{others} choice and another with a discrete range, the
22390 @code{others} choice is flagged even if the discrete range specifies
22391 exactly one component; for example, @code{(1..1 => 0, others => 1)}.
22393 This rule has no parameters.
22395 @node OTHERS_In_CASE_Statements
22396 @subsection @code{OTHERS_In_CASE_Statements}
22397 @cindex @code{OTHERS_In_CASE_Statements} rule (for @command{gnatcheck})
22400 Flag any use of an @code{others} choice in a @code{case} statement.
22402 This rule has no parameters.
22404 @node OTHERS_In_Exception_Handlers
22405 @subsection @code{OTHERS_In_Exception_Handlers}
22406 @cindex @code{OTHERS_In_Exception_Handlers} rule (for @command{gnatcheck})
22409 Flag any use of an @code{others} choice in an exception handler.
22411 This rule has no parameters.
22414 @node Outer_Loop_Exits
22415 @subsection @code{Outer_Loop_Exits}
22416 @cindex @code{Outer_Loop_Exits} rule (for @command{gnatcheck})
22419 Flag each @code{exit} statement containing a loop name that is not the name
22420 of the immediately enclosing @code{loop} statement.
22422 This rule has no parameters.
22425 @node Overloaded_Operators
22426 @subsection @code{Overloaded_Operators}
22427 @cindex @code{Overloaded_Operators} rule (for @command{gnatcheck})
22430 Flag each function declaration that overloads an operator symbol.
22431 A function body is checked only if the body does not have a
22432 separate spec. Formal functions are also checked. For a
22433 renaming declaration, only renaming-as-declaration is checked
22435 This rule has no parameters.
22438 @node Overly_Nested_Control_Structures
22439 @subsection @code{Overly_Nested_Control_Structures}
22440 @cindex @code{Overly_Nested_Control_Structures} rule (for @command{gnatcheck})
22443 Flag each control structure whose nesting level exceeds the value provided
22444 in the rule parameter.
22446 The control structures checked are the following:
22449 @item @code{if} statement
22450 @item @code{case} statement
22451 @item @code{loop} statement
22452 @item Selective accept statement
22453 @item Timed entry call statement
22454 @item Conditional entry call
22455 @item Asynchronous select statement
22459 The rule has the following parameter for the @option{+R} option:
22463 Positive integer specifying the maximal control structure nesting
22464 level that is not flagged
22468 If the parameter for the @option{+R} option is not specified or
22469 if it is not a positive integer, @option{+R} option is ignored.
22471 If more then one option is specified for the gnatcheck call, the later option and
22472 new parameter override the previous one(s).
22475 @node Parameters_Out_Of_Order
22476 @subsection @code{Parameters_Out_Of_Order}
22477 @cindex @code{Parameters_Out_Of_Order} rule (for @command{gnatcheck})
22480 Flag each subprogram and entry declaration whose formal parameters are not
22481 ordered according to the following scheme:
22485 @item @code{in} and @code{access} parameters first,
22486 then @code{in out} parameters,
22487 and then @code{out} parameters;
22489 @item for @code{in} mode, parameters with default initialization expressions
22494 Only the first violation of the described order is flagged.
22496 The following constructs are checked:
22499 @item subprogram declarations (including null procedures);
22500 @item generic subprogram declarations;
22501 @item formal subprogram declarations;
22502 @item entry declarations;
22503 @item subprogram bodies and subprogram body stubs that do not
22504 have separate specifications
22508 Subprogram renamings are not checked.
22510 This rule has no parameters.
22513 @node Positional_Actuals_For_Defaulted_Generic_Parameters
22514 @subsection @code{Positional_Actuals_For_Defaulted_Generic_Parameters}
22515 @cindex @code{Positional_Actuals_For_Defaulted_Generic_Parameters} rule (for @command{gnatcheck})
22518 Flag each generic actual parameter corresponding to a generic formal
22519 parameter with a default initialization, if positional notation is used.
22521 This rule has no parameters.
22523 @node Positional_Actuals_For_Defaulted_Parameters
22524 @subsection @code{Positional_Actuals_For_Defaulted_Parameters}
22525 @cindex @code{Positional_Actuals_For_Defaulted_Parameters} rule (for @command{gnatcheck})
22528 Flag each actual parameter to a subprogram or entry call where the
22529 corresponding formal parameter has a default expression, if positional
22532 This rule has no parameters.
22534 @node Positional_Components
22535 @subsection @code{Positional_Components}
22536 @cindex @code{Positional_Components} rule (for @command{gnatcheck})
22539 Flag each array, record and extension aggregate that includes positional
22542 This rule has no parameters.
22545 @node Positional_Generic_Parameters
22546 @subsection @code{Positional_Generic_Parameters}
22547 @cindex @code{Positional_Generic_Parameters} rule (for @command{gnatcheck})
22550 Flag each positional actual generic parameter except for the case when
22551 the generic unit being iinstantiated has exactly one generic formal
22554 This rule has no parameters.
22557 @node Positional_Parameters
22558 @subsection @code{Positional_Parameters}
22559 @cindex @code{Positional_Parameters} rule (for @command{gnatcheck})
22562 Flag each positional parameter notation in a subprogram or entry call,
22563 except for the following:
22567 Parameters of calls to of prefix or infix operators are not flagged
22569 If the called subprogram or entry has only one formal parameter,
22570 the parameter of the call is not flagged;
22572 If a subprogram call uses the @emph{Object.Operation} notation, then
22575 the first parameter (that is, @emph{Object}) is not flagged;
22577 if the called subprogram has only two parameters, the second parameter
22578 of the call is not flagged;
22583 This rule has no parameters.
22588 @node Predefined_Numeric_Types
22589 @subsection @code{Predefined_Numeric_Types}
22590 @cindex @code{Predefined_Numeric_Types} rule (for @command{gnatcheck})
22593 Flag each explicit use of the name of any numeric type or subtype defined
22594 in package @code{Standard}.
22596 The rationale for this rule is to detect when the
22597 program may depend on platform-specific characteristics of the implementation
22598 of the predefined numeric types. Note that this rule is over-pessimistic;
22599 for example, a program that uses @code{String} indexing
22600 likely needs a variable of type @code{Integer}.
22601 Another example is the flagging of predefined numeric types with explicit
22604 @smallexample @c ada
22605 subtype My_Integer is Integer range Left .. Right;
22606 Vy_Var : My_Integer;
22610 This rule detects only numeric types and subtypes defined in
22611 @code{Standard}. The use of numeric types and subtypes defined in other
22612 predefined packages (such as @code{System.Any_Priority} or
22613 @code{Ada.Text_IO.Count}) is not flagged
22615 This rule has no parameters.
22619 @node Raising_External_Exceptions
22620 @subsection @code{Raising_External_Exceptions}
22621 @cindex @code{Raising_External_Exceptions} rule (for @command{gnatcheck})
22624 Flag any @code{raise} statement, in a program unit declared in a library
22625 package or in a generic library package, for an exception that is
22626 neither a predefined exception nor an exception that is also declared (or
22627 renamed) in the visible part of the package.
22629 This rule has no parameters.
22633 @node Raising_Predefined_Exceptions
22634 @subsection @code{Raising_Predefined_Exceptions}
22635 @cindex @code{Raising_Predefined_Exceptions} rule (for @command{gnatcheck})
22638 Flag each @code{raise} statement that raises a predefined exception
22639 (i.e., one of the exceptions @code{Constraint_Error}, @code{Numeric_Error},
22640 @code{Program_Error}, @code{Storage_Error}, or @code{Tasking_Error}).
22642 This rule has no parameters.
22644 @node Separate_Numeric_Error_Handlers
22645 @subsection @code{Separate_Numeric_Error_Handlers}
22646 @cindex @code{Separate_Numeric_Error_Handlers} rule (for @command{gnatcheck})
22649 Flags each exception handler that contains a choice for
22650 the predefined @code{Constraint_Error} exception, but does not contain
22651 the choice for the predefined @code{Numeric_Error} exception, or
22652 that contains the choice for @code{Numeric_Error}, but does not contain the
22653 choice for @code{Constraint_Error}.
22655 This rule has no parameters.
22659 @subsection @code{Recursion} (under construction, GLOBAL)
22660 @cindex @code{Recursion} rule (for @command{gnatcheck})
22663 Flag recursive subprograms (cycles in the call graph). Declarations, and not
22664 calls, of recursive subprograms are detected.
22666 This rule has no parameters.
22670 @node Side_Effect_Functions
22671 @subsection @code{Side_Effect_Functions} (under construction, GLOBAL)
22672 @cindex @code{Side_Effect_Functions} rule (for @command{gnatcheck})
22675 Flag functions with side effects.
22677 We define a side effect as changing any data object that is not local for the
22678 body of this function.
22680 At the moment, we do NOT consider a side effect any input-output operations
22681 (changing a state or a content of any file).
22683 We do not consider protected functions for this rule (???)
22685 There are the following sources of side effect:
22688 @item Explicit (or direct) side-effect:
22692 direct assignment to a non-local variable;
22695 direct call to an entity that is known to change some data object that is
22696 not local for the body of this function (Note, that if F1 calls F2 and F2
22697 does have a side effect, this does not automatically mean that F1 also
22698 have a side effect, because it may be the case that F2 is declared in
22699 F1's body and it changes some data object that is global for F2, but
22703 @item Indirect side-effect:
22706 Subprogram calls implicitly issued by:
22709 computing initialization expressions from type declarations as a part
22710 of object elaboration or allocator evaluation;
22712 computing implicit parameters of subprogram or entry calls or generic
22717 activation of a task that change some non-local data object (directly or
22721 elaboration code of a package that is a result of a package instantiation;
22724 controlled objects;
22727 @item Situations when we can suspect a side-effect, but the full static check
22728 is either impossible or too hard:
22731 assignment to access variables or to the objects pointed by access
22735 call to a subprogram pointed by access-to-subprogram value
22743 This rule has no parameters.
22747 @subsection @code{Slices}
22748 @cindex @code{Slices} rule (for @command{gnatcheck})
22751 Flag all uses of array slicing
22753 This rule has no parameters.
22756 @node Too_Many_Parents
22757 @subsection @code{Too_Many_Parents}
22758 @cindex @code{Too_Many_Parents} rule (for @command{gnatcheck})
22761 Flags any type declaration, single task declaration or single protected
22762 declaration that has more then @option{N} parents, @option{N} is a parameter
22764 A parent here is either a (sub)type denoted by the subtype mark from the
22765 parent_subtype_indication (in case of a derived type declaration), or
22766 any of the progenitors from the interface list, if any.
22768 This rule has the following (mandatory) parameters for the @option{+R} option:
22772 Positive integer specifying the maximal allowed number of parents.
22776 @node Unassigned_OUT_Parameters
22777 @subsection @code{Unassigned_OUT_Parameters}
22778 @cindex @code{Unassigned_OUT_Parameters} rule (for @command{gnatcheck})
22781 Flags procedures' @code{out} parameters that are not assigned, and
22782 identifies the contexts in which the assignments are missing.
22784 An @code{out} parameter is flagged in the statements in the procedure
22785 body's handled sequence of statements (before the procedure body's
22786 @code{exception} part, if any) if this sequence of statements contains
22787 no assignments to the parameter.
22789 An @code{out} parameter is flagged in an exception handler in the exception
22790 part of the procedure body's handled sequence of statements if the handler
22791 contains no assignment to the parameter.
22793 Bodies of generic procedures are also considered.
22795 The following are treated as assignments to an @code{out} parameter:
22799 an assignment statement, with the parameter or some component as the target;
22802 passing the parameter (or one of its components) as an @code{out} or
22803 @code{in out} parameter.
22807 This rule does not have any parameters.
22811 @node Uncommented_BEGIN_In_Package_Bodies
22812 @subsection @code{Uncommented_BEGIN_In_Package_Bodies}
22813 @cindex @code{Uncommented_BEGIN_In_Package_Bodies} rule (for @command{gnatcheck})
22816 Flags each package body with declarations and a statement part that does not
22817 include a trailing comment on the line containing the @code{begin} keyword;
22818 this trailing comment needs to specify the package name and nothing else.
22819 The @code{begin} is not flagged if the package body does not
22820 contain any declarations.
22822 If the @code{begin} keyword is placed on the
22823 same line as the last declaration or the first statement, it is flagged
22824 independently of whether the line contains a trailing comment. The
22825 diagnostic message is attached to the line containing the first statement.
22827 This rule has no parameters.
22829 @node Unconditional_Exits
22830 @subsection @code{Unconditional_Exits}
22831 @cindex @code{Unconditional_Exits} rule (for @command{gnatcheck})
22834 Flag unconditional @code{exit} statements.
22836 This rule has no parameters.
22838 @node Unconstrained_Array_Returns
22839 @subsection @code{Unconstrained_Array_Returns}
22840 @cindex @code{Unconstrained_Array_Returns} rule (for @command{gnatcheck})
22843 Flag each function returning an unconstrained array. Function declarations,
22844 function bodies (and body stubs) having no separate specifications,
22845 and generic function instantiations are checked.
22846 Function calls and function renamings are
22849 Generic function declarations, and function declarations in generic
22850 packages are not checked, instead this rule checks the results of
22851 generic instantiations (that is, expanded specification and expanded
22852 body corresponding to an instantiation).
22854 This rule has no parameters.
22856 @node Universal_Ranges
22857 @subsection @code{Universal_Ranges}
22858 @cindex @code{Universal_Ranges} rule (for @command{gnatcheck})
22861 Flag discrete ranges that are a part of an index constraint, constrained
22862 array definition, or @code{for}-loop parameter specification, and whose bounds
22863 are both of type @i{universal_integer}. Ranges that have at least one
22864 bound of a specific type (such as @code{1 .. N}, where @code{N} is a variable
22865 or an expression of non-universal type) are not flagged.
22867 This rule has no parameters.
22870 @node Unnamed_Blocks_And_Loops
22871 @subsection @code{Unnamed_Blocks_And_Loops}
22872 @cindex @code{Unnamed_Blocks_And_Loops} rule (for @command{gnatcheck})
22875 Flag each unnamed block statement and loop statement.
22877 The rule has no parameters.
22882 @node Unused_Subprograms
22883 @subsection @code{Unused_Subprograms} (under construction, GLOBAL)
22884 @cindex @code{Unused_Subprograms} rule (for @command{gnatcheck})
22887 Flag all unused subprograms.
22889 This rule has no parameters.
22895 @node USE_PACKAGE_Clauses
22896 @subsection @code{USE_PACKAGE_Clauses}
22897 @cindex @code{USE_PACKAGE_Clauses} rule (for @command{gnatcheck})
22900 Flag all @code{use} clauses for packages; @code{use type} clauses are
22903 This rule has no parameters.
22906 @node Visible_Components
22907 @subsection @code{Visible_Components}
22908 @cindex @code{Visible_Components} rule (for @command{gnatcheck})
22911 Flags all the type declarations located in the visible part of a library
22912 package or a library generic package that can declare a visible component. A
22913 type is considered as declaring a visible component if it contains a record
22914 definition by its own or as a part of a record extension. Type declaration is
22915 flagged even if it contains a record definition that defines no components.
22917 Declarations located in private parts of local (generic) packages are not
22918 flagged. Declarations in private packages are not flagged.
22920 This rule has no parameters.
22923 @node Volatile_Objects_Without_Address_Clauses
22924 @subsection @code{Volatile_Objects_Without_Address_Clauses}
22925 @cindex @code{Volatile_Objects_Without_Address_Clauses} rule (for @command{gnatcheck})
22928 Flag each volatile object that does not have an address clause.
22930 The following check is made: if the pragma @code{Volatile} is applied to a
22931 data object or to its type, then an address clause must
22932 be supplied for this object.
22934 This rule does not check the components of data objects,
22935 array components that are volatile as a result of the pragma
22936 @code{Volatile_Components}, or objects that are volatile because
22937 they are atomic as a result of pragmas @code{Atomic} or
22938 @code{Atomic_Components}.
22940 Only variable declarations, and not constant declarations, are checked.
22942 This rule has no parameters.
22944 @node Example of gnatcheck Usage
22945 @section Example of @command{gnatcheck} Usage
22948 Here is a simple example. Suppose that in the current directory we have a
22949 project file named @file{gnatcheck_example.gpr} with the following content:
22951 @smallexample @c projectfile
22952 project Gnatcheck_Example is
22954 for Source_Dirs use ("src");
22955 for Object_Dir use "obj";
22956 for Main use ("main.adb");
22959 for Default_Switches ("ada") use ("-rules", "-from=coding_standard");
22962 end Gnatcheck_Example;
22966 And the file named @file{coding_standard} is also located in the current
22967 directory and has the following content:
22970 -----------------------------------------------------
22971 -- This is a sample gnatcheck coding standard file --
22972 -----------------------------------------------------
22974 -- First, turning on rules, that are directly implemented in gnatcheck
22975 +RAbstract_Type_Declarations
22978 +RFloat_Equality_Checks
22979 +REXIT_Statements_With_No_Loop_Name
22981 -- Then, activating compiler checks of interest:
22983 -- This style check checks if a unit name is present on END keyword that
22984 -- is the end of the unit declaration
22988 And the subdirectory @file{src} contains the following Ada sources:
22992 @smallexample @c ada
22994 type T is abstract tagged private;
22995 procedure P (X : T) is abstract;
22998 type My_Float is digits 8;
22999 function Is_Equal (L, R : My_Float) return Boolean;
23002 type T is abstract tagged null record;
23009 @smallexample @c ada
23010 package body Pack is
23011 package body Inner is
23012 function Is_Equal (L, R : My_Float) return Boolean is
23021 and @file{main.adb}
23023 @smallexample @c ada
23024 with Pack; use Pack;
23028 (gnatcheck, Exempt_On, "Anonymous_Arrays", "this one is fine");
23029 Float_Array : array (1 .. 10) of Inner.My_Float;
23030 pragma Annotate (gnatcheck, Exempt_Off, "Anonymous_Arrays");
23032 Another_Float_Array : array (1 .. 10) of Inner.My_Float;
23036 B : Boolean := False;
23039 for J in Float_Array'Range loop
23040 if Is_Equal (Float_Array (J), Another_Float_Array (J)) then
23049 And suppose we call @command{gnatcheck} from the current directory using
23050 the @command{gnat} driver:
23053 gnat check -Pgnatcheck_example.gpr
23057 As a result, @command{gnatcheck} is called to check all the files from the
23058 project @file{gnatcheck_example.gpr} using the coding standard defined by
23059 the file @file{coding_standard}. As the result, the @command{gnatcheck}
23060 report file named @file{gnatcheck.out} will be created in the current
23061 directory, and it will have the following content:
23064 RULE CHECKING REPORT
23068 Date and time of execution: 2009.10.28 14:17
23069 Tool version: GNATCHECK (built with ASIS 2.0.R for GNAT Pro 6.3.0w (20091016))
23072 gnatcheck -files=.../GNAT-TEMP-000004.TMP -cargs -gnatec=.../GNAT-TEMP-000003.TMP -rules -from=coding_standard
23074 Coding standard (applied rules):
23075 Abstract_Type_Declarations
23077 EXIT_Statements_With_No_Loop_Name
23078 Float_Equality_Checks
23081 Compiler style checks: -gnatye
23083 Number of coding standard violations: 6
23084 Number of exempted coding standard violations: 1
23086 2. DETECTED RULE VIOLATIONS
23088 2.1. NON-EXEMPTED VIOLATIONS
23090 Source files with non-exempted violations
23095 List of violations grouped by files, and ordered by increasing source location:
23097 pack.ads:2:4: declaration of abstract type
23098 pack.ads:5:4: declaration of local package
23099 pack.ads:10:30: declaration of abstract type
23100 pack.ads:11:1: (style) "end Pack" required
23101 pack.adb:5:19: use of equality operation for float values
23102 pack.adb:6:7: (style) "end Is_Equal" required
23103 main.adb:9:26: anonymous array type
23104 main.adb:19:10: exit statement with no loop name
23106 2.2. EXEMPTED VIOLATIONS
23108 Source files with exempted violations
23111 List of violations grouped by files, and ordered by increasing source location:
23113 main.adb:6:18: anonymous array type
23116 2.3. SOURCE FILES WITH NO VIOLATION
23118 No files without violations
23124 @c *********************************
23125 @node Creating Sample Bodies Using gnatstub
23126 @chapter Creating Sample Bodies Using @command{gnatstub}
23130 @command{gnatstub} creates body stubs, that is, empty but compilable bodies
23131 for library unit declarations.
23133 Note: to invoke @code{gnatstub} with a project file, use the @code{gnat}
23134 driver (see @ref{The GNAT Driver and Project Files}).
23136 To create a body stub, @command{gnatstub} has to compile the library
23137 unit declaration. Therefore, bodies can be created only for legal
23138 library units. Moreover, if a library unit depends semantically upon
23139 units located outside the current directory, you have to provide
23140 the source search path when calling @command{gnatstub}, see the description
23141 of @command{gnatstub} switches below.
23143 By default, all the program unit body stubs generated by @code{gnatstub}
23144 raise the predefined @code{Program_Error} exception, which will catch
23145 accidental calls of generated stubs. This behavior can be changed with
23146 option @option{^--no-exception^/NO_EXCEPTION^} (see below).
23149 * Running gnatstub::
23150 * Switches for gnatstub::
23153 @node Running gnatstub
23154 @section Running @command{gnatstub}
23157 @command{gnatstub} has the command-line interface of the form
23160 $ gnatstub @ovar{switches} @var{filename} @ovar{directory}
23167 is the name of the source file that contains a library unit declaration
23168 for which a body must be created. The file name may contain the path
23170 The file name does not have to follow the GNAT file name conventions. If the
23172 does not follow GNAT file naming conventions, the name of the body file must
23174 explicitly as the value of the @option{^-o^/BODY=^@var{body-name}} option.
23175 If the file name follows the GNAT file naming
23176 conventions and the name of the body file is not provided,
23179 of the body file from the argument file name by replacing the @file{.ads}
23181 with the @file{.adb} suffix.
23184 indicates the directory in which the body stub is to be placed (the default
23189 is an optional sequence of switches as described in the next section
23192 @node Switches for gnatstub
23193 @section Switches for @command{gnatstub}
23199 @cindex @option{^-f^/FULL^} (@command{gnatstub})
23200 If the destination directory already contains a file with the name of the
23202 for the argument spec file, replace it with the generated body stub.
23204 @item ^-hs^/HEADER=SPEC^
23205 @cindex @option{^-hs^/HEADER=SPEC^} (@command{gnatstub})
23206 Put the comment header (i.e., all the comments preceding the
23207 compilation unit) from the source of the library unit declaration
23208 into the body stub.
23210 @item ^-hg^/HEADER=GENERAL^
23211 @cindex @option{^-hg^/HEADER=GENERAL^} (@command{gnatstub})
23212 Put a sample comment header into the body stub.
23214 @item ^--header-file=@var{filename}^/FROM_HEADER_FILE=@var{filename}^
23215 @cindex @option{^--header-file^/FROM_HEADER_FILE=^} (@command{gnatstub})
23216 Use the content of the file as the comment header for a generated body stub.
23220 @cindex @option{-IDIR} (@command{gnatstub})
23222 @cindex @option{-I-} (@command{gnatstub})
23225 @item /NOCURRENT_DIRECTORY
23226 @cindex @option{/NOCURRENT_DIRECTORY} (@command{gnatstub})
23228 ^These switches have ^This switch has^ the same meaning as in calls to
23230 ^They define ^It defines ^ the source search path in the call to
23231 @command{gcc} issued
23232 by @command{gnatstub} to compile an argument source file.
23234 @item ^-gnatec^/CONFIGURATION_PRAGMAS_FILE=^@var{PATH}
23235 @cindex @option{^-gnatec^/CONFIGURATION_PRAGMAS_FILE^} (@command{gnatstub})
23236 This switch has the same meaning as in calls to @command{gcc}.
23237 It defines the additional configuration file to be passed to the call to
23238 @command{gcc} issued
23239 by @command{gnatstub} to compile an argument source file.
23241 @item ^-gnatyM^/MAX_LINE_LENGTH=^@var{n}
23242 @cindex @option{^-gnatyM^/MAX_LINE_LENGTH^} (@command{gnatstub})
23243 (@var{n} is a non-negative integer). Set the maximum line length in the
23244 body stub to @var{n}; the default is 79. The maximum value that can be
23245 specified is 32767. Note that in the special case of configuration
23246 pragma files, the maximum is always 32767 regardless of whether or
23247 not this switch appears.
23249 @item ^-gnaty^/STYLE_CHECKS=^@var{n}
23250 @cindex @option{^-gnaty^/STYLE_CHECKS=^} (@command{gnatstub})
23251 (@var{n} is a non-negative integer from 1 to 9). Set the indentation level in
23252 the generated body sample to @var{n}.
23253 The default indentation is 3.
23255 @item ^-gnatyo^/ORDERED_SUBPROGRAMS^
23256 @cindex @option{^-gnato^/ORDERED_SUBPROGRAMS^} (@command{gnatstub})
23257 Order local bodies alphabetically. (By default local bodies are ordered
23258 in the same way as the corresponding local specs in the argument spec file.)
23260 @item ^-i^/INDENTATION=^@var{n}
23261 @cindex @option{^-i^/INDENTATION^} (@command{gnatstub})
23262 Same as @option{^-gnaty^/STYLE_CHECKS=^@var{n}}
23264 @item ^-k^/TREE_FILE=SAVE^
23265 @cindex @option{^-k^/TREE_FILE=SAVE^} (@command{gnatstub})
23266 Do not remove the tree file (i.e., the snapshot of the compiler internal
23267 structures used by @command{gnatstub}) after creating the body stub.
23269 @item ^-l^/LINE_LENGTH=^@var{n}
23270 @cindex @option{^-l^/LINE_LENGTH^} (@command{gnatstub})
23271 Same as @option{^-gnatyM^/MAX_LINE_LENGTH=^@var{n}}
23273 @item ^--no-exception^/NO_EXCEPTION^
23274 @cindex @option{^--no-exception^/NO_EXCEPTION^} (@command{gnatstub})
23275 Avoind raising PROGRAM_ERROR in the generated bodies of program unit stubs.
23276 This is not always possible for function stubs.
23278 @item ^--no-local-header^/NO_LOCAL_HEADER^
23279 @cindex @option{^--no-local-header^/NO_LOCAL_HEADER^} (@command{gnatstub})
23280 Do not place local comment header with unit name before body stub for a
23283 @item ^-o ^/BODY=^@var{body-name}
23284 @cindex @option{^-o^/BODY^} (@command{gnatstub})
23285 Body file name. This should be set if the argument file name does not
23287 the GNAT file naming
23288 conventions. If this switch is omitted the default name for the body will be
23290 from the argument file name according to the GNAT file naming conventions.
23293 @cindex @option{^-q^/QUIET^} (@command{gnatstub})
23294 Quiet mode: do not generate a confirmation when a body is
23295 successfully created, and do not generate a message when a body is not
23299 @item ^-r^/TREE_FILE=REUSE^
23300 @cindex @option{^-r^/TREE_FILE=REUSE^} (@command{gnatstub})
23301 Reuse the tree file (if it exists) instead of creating it. Instead of
23302 creating the tree file for the library unit declaration, @command{gnatstub}
23303 tries to find it in the current directory and use it for creating
23304 a body. If the tree file is not found, no body is created. This option
23305 also implies @option{^-k^/SAVE^}, whether or not
23306 the latter is set explicitly.
23308 @item ^-t^/TREE_FILE=OVERWRITE^
23309 @cindex @option{^-t^/TREE_FILE=OVERWRITE^} (@command{gnatstub})
23310 Overwrite the existing tree file. If the current directory already
23311 contains the file which, according to the GNAT file naming rules should
23312 be considered as a tree file for the argument source file,
23314 will refuse to create the tree file needed to create a sample body
23315 unless this option is set.
23317 @item ^-v^/VERBOSE^
23318 @cindex @option{^-v^/VERBOSE^} (@command{gnatstub})
23319 Verbose mode: generate version information.
23323 @c *********************************
23324 @node Generating Ada Bindings for C and C++ headers
23325 @chapter Generating Ada Bindings for C and C++ headers
23329 GNAT now comes with a binding generator for C and C++ headers which is
23330 intended to do 95% of the tedious work of generating Ada specs from C
23331 or C++ header files.
23333 Note that this capability is not intended to generate 100% correct Ada specs,
23334 and will is some cases require manual adjustments, although it can often
23335 be used out of the box in practice.
23337 Some of the known limitations include:
23340 @item only very simple character constant macros are translated into Ada
23341 constants. Function macros (macros with arguments) are partially translated
23342 as comments, to be completed manually if needed.
23343 @item some extensions (e.g. vector types) are not supported
23344 @item pointers to pointers or complex structures are mapped to System.Address
23347 The code generated is using the Ada 2005 syntax, which makes it
23348 easier to interface with other languages than previous versions of Ada.
23351 * Running the binding generator::
23352 * Generating bindings for C++ headers::
23356 @node Running the binding generator
23357 @section Running the binding generator
23360 The binding generator is part of the @command{gcc} compiler and can be
23361 invoked via the @option{-fdump-ada-spec} switch, which will generate Ada
23362 spec files for the header files specified on the command line, and all
23363 header files needed by these files transitivitely. For example:
23366 $ g++ -c -fdump-ada-spec -C /usr/include/time.h
23367 $ gcc -c -gnat05 *.ads
23370 will generate, under GNU/Linux, the following files: @file{time_h.ads},
23371 @file{bits_time_h.ads}, @file{stddef_h.ads}, @file{bits_types_h.ads} which
23372 correspond to the files @file{/usr/include/time.h},
23373 @file{/usr/include/bits/time.h}, etc@dots{}, and will then compile in Ada 2005
23374 mode these Ada specs.
23376 The @code{-C} switch tells @command{gcc} to extract comments from headers,
23377 and will attempt to generate corresponding Ada comments.
23379 If you want to generate a single Ada file and not the transitive closure, you
23380 can use instead the @option{-fdump-ada-spec-slim} switch.
23382 Note that we recommend when possible to use the @command{g++} driver to
23383 generate bindings, even for most C headers, since this will in general
23384 generate better Ada specs. For generating bindings for C++ headers, it is
23385 mandatory to use the @command{g++} command, or @command{gcc -x c++} which
23386 is equivalent in this case. If @command{g++} cannot work on your C headers
23387 because of incompatibilities between C and C++, then you can fallback to
23388 @command{gcc} instead.
23390 For an example of better bindings generated from the C++ front-end,
23391 the name of the parameters (when available) are actually ignored by the C
23392 front-end. Consider the following C header:
23395 extern void foo (int variable);
23398 with the C front-end, @code{variable} is ignored, and the above is handled as:
23401 extern void foo (int);
23404 generating a generic:
23407 procedure foo (param1 : int);
23410 with the C++ front-end, the name is available, and we generate:
23413 procedure foo (variable : int);
23416 In some cases, the generated bindings will be more complete or more meaningful
23417 when defining some macros, which you can do via the @option{-D} switch. This
23418 is for example the case with @file{Xlib.h} under GNU/Linux:
23421 g++ -c -fdump-ada-spec -DXLIB_ILLEGAL_ACCESS -C /usr/include/X11/Xlib.h
23424 The above will generate more complete bindings than a straight call without
23425 the @option{-DXLIB_ILLEGAL_ACCESS} switch.
23427 In other cases, it is not possible to parse a header file in a stand alone
23428 manner, because other include files need to be included first. In this
23429 case, the solution is to create a small header file including the needed
23430 @code{#include} and possible @code{#define} directives. For example, to
23431 generate Ada bindings for @file{readline/readline.h}, you need to first
23432 include @file{stdio.h}, so you can create a file with the following two
23433 lines in e.g. @file{readline1.h}:
23437 #include <readline/readline.h>
23440 and then generate Ada bindings from this file:
23443 $ g++ -c -fdump-ada-spec readline1.h
23446 @node Generating bindings for C++ headers
23447 @section Generating bindings for C++ headers
23450 Generating bindings for C++ headers is done using the same options, always
23451 with the @command{g++} compiler.
23453 In this mode, C++ classes will be mapped to Ada tagged types, constructors
23454 will be mapped using the @code{CPP_Constructor} pragma, and when possible,
23455 multiple inheritance of abstract classes will be mapped to Ada interfaces
23456 (@xref{Interfacing to C++,,,gnat_rm, GNAT Reference Manual}, for additional
23457 information on interfacing to C++).
23459 For example, given the following C++ header file:
23466 virtual int Number_Of_Teeth () = 0;
23471 virtual void Set_Owner (char* Name) = 0;
23477 virtual void Set_Age (int New_Age);
23480 class Dog : Animal, Carnivore, Domestic @{
23485 virtual int Number_Of_Teeth ();
23486 virtual void Set_Owner (char* Name);
23494 The corresponding Ada code is generated:
23496 @smallexample @c ada
23499 package Class_Carnivore is
23500 type Carnivore is limited interface;
23501 pragma Import (CPP, Carnivore);
23503 function Number_Of_Teeth (this : access Carnivore) return int is abstract;
23505 use Class_Carnivore;
23507 package Class_Domestic is
23508 type Domestic is limited interface;
23509 pragma Import (CPP, Domestic);
23511 procedure Set_Owner
23512 (this : access Domestic;
23513 Name : Interfaces.C.Strings.chars_ptr) is abstract;
23515 use Class_Domestic;
23517 package Class_Animal is
23518 type Animal is tagged limited record
23519 Age_Count : aliased int;
23521 pragma Import (CPP, Animal);
23523 procedure Set_Age (this : access Animal; New_Age : int);
23524 pragma Import (CPP, Set_Age, "_ZN6Animal7Set_AgeEi");
23528 package Class_Dog is
23529 type Dog is new Animal and Carnivore and Domestic with record
23530 Tooth_Count : aliased int;
23531 Owner : Interfaces.C.Strings.chars_ptr;
23533 pragma Import (CPP, Dog);
23535 function Number_Of_Teeth (this : access Dog) return int;
23536 pragma Import (CPP, Number_Of_Teeth, "_ZN3Dog15Number_Of_TeethEv");
23538 procedure Set_Owner
23539 (this : access Dog; Name : Interfaces.C.Strings.chars_ptr);
23540 pragma Import (CPP, Set_Owner, "_ZN3Dog9Set_OwnerEPc");
23542 function New_Dog return Dog;
23543 pragma CPP_Constructor (New_Dog);
23544 pragma Import (CPP, New_Dog, "_ZN3DogC1Ev");
23555 @item -fdump-ada-spec
23556 @cindex @option{-fdump-ada-spec} (@command{gcc})
23557 Generate Ada spec files for the given header files transitively (including
23558 all header files that these headers depend upon).
23560 @item -fdump-ada-spec-slim
23561 @cindex @option{-fdump-ada-spec-slim} (@command{gcc})
23562 Generate Ada spec files for the header files specified on the command line
23566 @cindex @option{-C} (@command{gcc})
23567 Extract comments from headers and generate Ada comments in the Ada spec files.
23570 @node Other Utility Programs
23571 @chapter Other Utility Programs
23574 This chapter discusses some other utility programs available in the Ada
23578 * Using Other Utility Programs with GNAT::
23579 * The External Symbol Naming Scheme of GNAT::
23580 * Converting Ada Files to html with gnathtml::
23581 * Installing gnathtml::
23588 @node Using Other Utility Programs with GNAT
23589 @section Using Other Utility Programs with GNAT
23592 The object files generated by GNAT are in standard system format and in
23593 particular the debugging information uses this format. This means
23594 programs generated by GNAT can be used with existing utilities that
23595 depend on these formats.
23598 In general, any utility program that works with C will also often work with
23599 Ada programs generated by GNAT. This includes software utilities such as
23600 gprof (a profiling program), @code{gdb} (the FSF debugger), and utilities such
23604 @node The External Symbol Naming Scheme of GNAT
23605 @section The External Symbol Naming Scheme of GNAT
23608 In order to interpret the output from GNAT, when using tools that are
23609 originally intended for use with other languages, it is useful to
23610 understand the conventions used to generate link names from the Ada
23613 All link names are in all lowercase letters. With the exception of library
23614 procedure names, the mechanism used is simply to use the full expanded
23615 Ada name with dots replaced by double underscores. For example, suppose
23616 we have the following package spec:
23618 @smallexample @c ada
23629 The variable @code{MN} has a full expanded Ada name of @code{QRS.MN}, so
23630 the corresponding link name is @code{qrs__mn}.
23632 Of course if a @code{pragma Export} is used this may be overridden:
23634 @smallexample @c ada
23639 pragma Export (Var1, C, External_Name => "var1_name");
23641 pragma Export (Var2, C, Link_Name => "var2_link_name");
23648 In this case, the link name for @var{Var1} is whatever link name the
23649 C compiler would assign for the C function @var{var1_name}. This typically
23650 would be either @var{var1_name} or @var{_var1_name}, depending on operating
23651 system conventions, but other possibilities exist. The link name for
23652 @var{Var2} is @var{var2_link_name}, and this is not operating system
23656 One exception occurs for library level procedures. A potential ambiguity
23657 arises between the required name @code{_main} for the C main program,
23658 and the name we would otherwise assign to an Ada library level procedure
23659 called @code{Main} (which might well not be the main program).
23661 To avoid this ambiguity, we attach the prefix @code{_ada_} to such
23662 names. So if we have a library level procedure such as
23664 @smallexample @c ada
23667 procedure Hello (S : String);
23673 the external name of this procedure will be @var{_ada_hello}.
23676 @node Converting Ada Files to html with gnathtml
23677 @section Converting Ada Files to HTML with @code{gnathtml}
23680 This @code{Perl} script allows Ada source files to be browsed using
23681 standard Web browsers. For installation procedure, see the section
23682 @xref{Installing gnathtml}.
23684 Ada reserved keywords are highlighted in a bold font and Ada comments in
23685 a blue font. Unless your program was compiled with the gcc @option{-gnatx}
23686 switch to suppress the generation of cross-referencing information, user
23687 defined variables and types will appear in a different color; you will
23688 be able to click on any identifier and go to its declaration.
23690 The command line is as follow:
23692 $ perl gnathtml.pl @ovar{^switches^options^} @var{ada-files}
23696 You can pass it as many Ada files as you want. @code{gnathtml} will generate
23697 an html file for every ada file, and a global file called @file{index.htm}.
23698 This file is an index of every identifier defined in the files.
23700 The available ^switches^options^ are the following ones:
23704 @cindex @option{-83} (@code{gnathtml})
23705 Only the Ada 83 subset of keywords will be highlighted.
23707 @item -cc @var{color}
23708 @cindex @option{-cc} (@code{gnathtml})
23709 This option allows you to change the color used for comments. The default
23710 value is green. The color argument can be any name accepted by html.
23713 @cindex @option{-d} (@code{gnathtml})
23714 If the Ada files depend on some other files (for instance through
23715 @code{with} clauses, the latter files will also be converted to html.
23716 Only the files in the user project will be converted to html, not the files
23717 in the run-time library itself.
23720 @cindex @option{-D} (@code{gnathtml})
23721 This command is the same as @option{-d} above, but @command{gnathtml} will
23722 also look for files in the run-time library, and generate html files for them.
23724 @item -ext @var{extension}
23725 @cindex @option{-ext} (@code{gnathtml})
23726 This option allows you to change the extension of the generated HTML files.
23727 If you do not specify an extension, it will default to @file{htm}.
23730 @cindex @option{-f} (@code{gnathtml})
23731 By default, gnathtml will generate html links only for global entities
23732 ('with'ed units, global variables and types,@dots{}). If you specify
23733 @option{-f} on the command line, then links will be generated for local
23736 @item -l @var{number}
23737 @cindex @option{-l} (@code{gnathtml})
23738 If this ^switch^option^ is provided and @var{number} is not 0, then
23739 @code{gnathtml} will number the html files every @var{number} line.
23742 @cindex @option{-I} (@code{gnathtml})
23743 Specify a directory to search for library files (@file{.ALI} files) and
23744 source files. You can provide several -I switches on the command line,
23745 and the directories will be parsed in the order of the command line.
23748 @cindex @option{-o} (@code{gnathtml})
23749 Specify the output directory for html files. By default, gnathtml will
23750 saved the generated html files in a subdirectory named @file{html/}.
23752 @item -p @var{file}
23753 @cindex @option{-p} (@code{gnathtml})
23754 If you are using Emacs and the most recent Emacs Ada mode, which provides
23755 a full Integrated Development Environment for compiling, checking,
23756 running and debugging applications, you may use @file{.gpr} files
23757 to give the directories where Emacs can find sources and object files.
23759 Using this ^switch^option^, you can tell gnathtml to use these files.
23760 This allows you to get an html version of your application, even if it
23761 is spread over multiple directories.
23763 @item -sc @var{color}
23764 @cindex @option{-sc} (@code{gnathtml})
23765 This ^switch^option^ allows you to change the color used for symbol
23767 The default value is red. The color argument can be any name accepted by html.
23769 @item -t @var{file}
23770 @cindex @option{-t} (@code{gnathtml})
23771 This ^switch^option^ provides the name of a file. This file contains a list of
23772 file names to be converted, and the effect is exactly as though they had
23773 appeared explicitly on the command line. This
23774 is the recommended way to work around the command line length limit on some
23779 @node Installing gnathtml
23780 @section Installing @code{gnathtml}
23783 @code{Perl} needs to be installed on your machine to run this script.
23784 @code{Perl} is freely available for almost every architecture and
23785 Operating System via the Internet.
23787 On Unix systems, you may want to modify the first line of the script
23788 @code{gnathtml}, to explicitly tell the Operating system where Perl
23789 is. The syntax of this line is:
23791 #!full_path_name_to_perl
23795 Alternatively, you may run the script using the following command line:
23798 $ perl gnathtml.pl @ovar{switches} @var{files}
23807 The GNAT distribution provides an Ada 95 template for the HP Language
23808 Sensitive Editor (LSE), a component of DECset. In order to
23809 access it, invoke LSE with the qualifier /ENVIRONMENT=GNU:[LIB]ADA95.ENV.
23816 GNAT supports The HP Performance Coverage Analyzer (PCA), a component
23817 of DECset. To use it proceed as outlined under ``HELP PCA'', except for running
23818 the collection phase with the /DEBUG qualifier.
23821 $ GNAT MAKE /DEBUG <PROGRAM_NAME>
23822 $ DEFINE LIB$DEBUG PCA$COLLECTOR
23823 $ RUN/DEBUG <PROGRAM_NAME>
23829 @c ******************************
23830 @node Code Coverage and Profiling
23831 @chapter Code Coverage and Profiling
23832 @cindex Code Coverage
23836 This chapter describes how to use @code{gcov} - coverage testing tool - and
23837 @code{gprof} - profiler tool - on your Ada programs.
23840 * Code Coverage of Ada Programs using gcov::
23841 * Profiling an Ada Program using gprof::
23844 @node Code Coverage of Ada Programs using gcov
23845 @section Code Coverage of Ada Programs using gcov
23847 @cindex -fprofile-arcs
23848 @cindex -ftest-coverage
23850 @cindex Code Coverage
23853 @code{gcov} is a test coverage program: it analyzes the execution of a given
23854 program on selected tests, to help you determine the portions of the program
23855 that are still untested.
23857 @code{gcov} is part of the GCC suite, and is described in detail in the GCC
23858 User's Guide. You can refer to this documentation for a more complete
23861 This chapter provides a quick startup guide, and
23862 details some Gnat-specific features.
23865 * Quick startup guide::
23869 @node Quick startup guide
23870 @subsection Quick startup guide
23872 In order to perform coverage analysis of a program using @code{gcov}, 3
23877 Code instrumentation during the compilation process
23879 Execution of the instrumented program
23881 Execution of the @code{gcov} tool to generate the result.
23884 The code instrumentation needed by gcov is created at the object level:
23885 The source code is not modified in any way, because the instrumentation code is
23886 inserted by gcc during the compilation process. To compile your code with code
23887 coverage activated, you need to recompile your whole project using the
23889 @code{-fprofile-arcs} and @code{-ftest-coverage}, and link it using
23890 @code{-fprofile-arcs}.
23893 $ gnatmake -P my_project.gpr -f -cargs -fprofile-arcs -ftest-coverage \
23894 -largs -fprofile-arcs
23897 This compilation process will create @file{.gcno} files together with
23898 the usual object files.
23900 Once the program is compiled with coverage instrumentation, you can
23901 run it as many times as needed - on portions of a test suite for
23902 example. The first execution will produce @file{.gcda} files at the
23903 same location as the @file{.gcno} files. The following executions
23904 will update those files, so that a cumulative result of the covered
23905 portions of the program is generated.
23907 Finally, you need to call the @code{gcov} tool. The different options of
23908 @code{gcov} are available in the GCC User's Guide, section 'Invoking gcov'.
23910 This will create annotated source files with a @file{.gcov} extension:
23911 @file{my_main.adb} file will be analysed in @file{my_main.adb.gcov}.
23913 @node Gnat specifics
23914 @subsection Gnat specifics
23916 Because Ada semantics, portions of the source code may be shared among
23917 several object files. This is the case for example when generics are
23918 involved, when inlining is active or when declarations generate initialisation
23919 calls. In order to take
23920 into account this shared code, you need to call @code{gcov} on all
23921 source files of the tested program at once.
23923 The list of source files might exceed the system's maximum command line
23924 length. In order to bypass this limitation, a new mechanism has been
23925 implemented in @code{gcov}: you can now list all your project's files into a
23926 text file, and provide this file to gcov as a parameter, preceded by a @@
23927 (e.g. @samp{gcov @@mysrclist.txt}).
23929 Note that on AIX compiling a static library with @code{-fprofile-arcs} is
23930 not supported as there can be unresolved symbols during the final link.
23932 @node Profiling an Ada Program using gprof
23933 @section Profiling an Ada Program using gprof
23939 This section is not meant to be an exhaustive documentation of @code{gprof}.
23940 Full documentation for it can be found in the GNU Profiler User's Guide
23941 documentation that is part of this GNAT distribution.
23943 Profiling a program helps determine the parts of a program that are executed
23944 most often, and are therefore the most time-consuming.
23946 @code{gprof} is the standard GNU profiling tool; it has been enhanced to
23947 better handle Ada programs and multitasking.
23948 It is currently supported on the following platforms
23953 solaris sparc/sparc64/x86
23959 In order to profile a program using @code{gprof}, 3 steps are needed:
23963 Code instrumentation, requiring a full recompilation of the project with the
23966 Execution of the program under the analysis conditions, i.e. with the desired
23969 Analysis of the results using the @code{gprof} tool.
23973 The following sections detail the different steps, and indicate how
23974 to interpret the results:
23976 * Compilation for profiling::
23977 * Program execution::
23979 * Interpretation of profiling results::
23982 @node Compilation for profiling
23983 @subsection Compilation for profiling
23987 In order to profile a program the first step is to tell the compiler
23988 to generate the necessary profiling information. The compiler switch to be used
23989 is @code{-pg}, which must be added to other compilation switches. This
23990 switch needs to be specified both during compilation and link stages, and can
23991 be specified once when using gnatmake:
23994 gnatmake -f -pg -P my_project
23998 Note that only the objects that were compiled with the @samp{-pg} switch will be
23999 profiled; if you need to profile your whole project, use the
24000 @samp{-f} gnatmake switch to force full recompilation.
24002 @node Program execution
24003 @subsection Program execution
24006 Once the program has been compiled for profiling, you can run it as usual.
24008 The only constraint imposed by profiling is that the program must terminate
24009 normally. An interrupted program (via a Ctrl-C, kill, etc.) will not be
24012 Once the program completes execution, a data file called @file{gmon.out} is
24013 generated in the directory where the program was launched from. If this file
24014 already exists, it will be overwritten.
24016 @node Running gprof
24017 @subsection Running gprof
24020 The @code{gprof} tool is called as follow:
24023 gprof my_prog gmon.out
24034 The complete form of the gprof command line is the following:
24037 gprof [^switches^options^] [executable [data-file]]
24041 @code{gprof} supports numerous ^switch^options^. The order of these
24042 ^switch^options^ does not matter. The full list of options can be found in
24043 the GNU Profiler User's Guide documentation that comes with this documentation.
24045 The following is the subset of those switches that is most relevant:
24049 @item --demangle[=@var{style}]
24050 @itemx --no-demangle
24051 @cindex @option{--demangle} (@code{gprof})
24052 These options control whether symbol names should be demangled when
24053 printing output. The default is to demangle C++ symbols. The
24054 @code{--no-demangle} option may be used to turn off demangling. Different
24055 compilers have different mangling styles. The optional demangling style
24056 argument can be used to choose an appropriate demangling style for your
24057 compiler, in particular Ada symbols generated by GNAT can be demangled using
24058 @code{--demangle=gnat}.
24060 @item -e @var{function_name}
24061 @cindex @option{-e} (@code{gprof})
24062 The @samp{-e @var{function}} option tells @code{gprof} not to print
24063 information about the function @var{function_name} (and its
24064 children@dots{}) in the call graph. The function will still be listed
24065 as a child of any functions that call it, but its index number will be
24066 shown as @samp{[not printed]}. More than one @samp{-e} option may be
24067 given; only one @var{function_name} may be indicated with each @samp{-e}
24070 @item -E @var{function_name}
24071 @cindex @option{-E} (@code{gprof})
24072 The @code{-E @var{function}} option works like the @code{-e} option, but
24073 execution time spent in the function (and children who were not called from
24074 anywhere else), will not be used to compute the percentages-of-time for
24075 the call graph. More than one @samp{-E} option may be given; only one
24076 @var{function_name} may be indicated with each @samp{-E} option.
24078 @item -f @var{function_name}
24079 @cindex @option{-f} (@code{gprof})
24080 The @samp{-f @var{function}} option causes @code{gprof} to limit the
24081 call graph to the function @var{function_name} and its children (and
24082 their children@dots{}). More than one @samp{-f} option may be given;
24083 only one @var{function_name} may be indicated with each @samp{-f}
24086 @item -F @var{function_name}
24087 @cindex @option{-F} (@code{gprof})
24088 The @samp{-F @var{function}} option works like the @code{-f} option, but
24089 only time spent in the function and its children (and their
24090 children@dots{}) will be used to determine total-time and
24091 percentages-of-time for the call graph. More than one @samp{-F} option
24092 may be given; only one @var{function_name} may be indicated with each
24093 @samp{-F} option. The @samp{-F} option overrides the @samp{-E} option.
24097 @node Interpretation of profiling results
24098 @subsection Interpretation of profiling results
24102 The results of the profiling analysis are represented by two arrays: the
24103 'flat profile' and the 'call graph'. Full documentation of those outputs
24104 can be found in the GNU Profiler User's Guide.
24106 The flat profile shows the time spent in each function of the program, and how
24107 many time it has been called. This allows you to locate easily the most
24108 time-consuming functions.
24110 The call graph shows, for each subprogram, the subprograms that call it,
24111 and the subprograms that it calls. It also provides an estimate of the time
24112 spent in each of those callers/called subprograms.
24115 @c ******************************
24116 @node Running and Debugging Ada Programs
24117 @chapter Running and Debugging Ada Programs
24121 This chapter discusses how to debug Ada programs.
24123 It applies to GNAT on the Alpha OpenVMS platform;
24124 for I64 OpenVMS please refer to the @cite{OpenVMS Debugger Manual},
24125 since HP has implemented Ada support in the OpenVMS debugger on I64.
24128 An incorrect Ada program may be handled in three ways by the GNAT compiler:
24132 The illegality may be a violation of the static semantics of Ada. In
24133 that case GNAT diagnoses the constructs in the program that are illegal.
24134 It is then a straightforward matter for the user to modify those parts of
24138 The illegality may be a violation of the dynamic semantics of Ada. In
24139 that case the program compiles and executes, but may generate incorrect
24140 results, or may terminate abnormally with some exception.
24143 When presented with a program that contains convoluted errors, GNAT
24144 itself may terminate abnormally without providing full diagnostics on
24145 the incorrect user program.
24149 * The GNAT Debugger GDB::
24151 * Introduction to GDB Commands::
24152 * Using Ada Expressions::
24153 * Calling User-Defined Subprograms::
24154 * Using the Next Command in a Function::
24157 * Debugging Generic Units::
24158 * GNAT Abnormal Termination or Failure to Terminate::
24159 * Naming Conventions for GNAT Source Files::
24160 * Getting Internal Debugging Information::
24161 * Stack Traceback::
24167 @node The GNAT Debugger GDB
24168 @section The GNAT Debugger GDB
24171 @code{GDB} is a general purpose, platform-independent debugger that
24172 can be used to debug mixed-language programs compiled with @command{gcc},
24173 and in particular is capable of debugging Ada programs compiled with
24174 GNAT. The latest versions of @code{GDB} are Ada-aware and can handle
24175 complex Ada data structures.
24177 @xref{Top,, Debugging with GDB, gdb, Debugging with GDB},
24179 located in the GNU:[DOCS] directory,
24181 for full details on the usage of @code{GDB}, including a section on
24182 its usage on programs. This manual should be consulted for full
24183 details. The section that follows is a brief introduction to the
24184 philosophy and use of @code{GDB}.
24186 When GNAT programs are compiled, the compiler optionally writes debugging
24187 information into the generated object file, including information on
24188 line numbers, and on declared types and variables. This information is
24189 separate from the generated code. It makes the object files considerably
24190 larger, but it does not add to the size of the actual executable that
24191 will be loaded into memory, and has no impact on run-time performance. The
24192 generation of debug information is triggered by the use of the
24193 ^-g^/DEBUG^ switch in the @command{gcc} or @command{gnatmake} command
24194 used to carry out the compilations. It is important to emphasize that
24195 the use of these options does not change the generated code.
24197 The debugging information is written in standard system formats that
24198 are used by many tools, including debuggers and profilers. The format
24199 of the information is typically designed to describe C types and
24200 semantics, but GNAT implements a translation scheme which allows full
24201 details about Ada types and variables to be encoded into these
24202 standard C formats. Details of this encoding scheme may be found in
24203 the file exp_dbug.ads in the GNAT source distribution. However, the
24204 details of this encoding are, in general, of no interest to a user,
24205 since @code{GDB} automatically performs the necessary decoding.
24207 When a program is bound and linked, the debugging information is
24208 collected from the object files, and stored in the executable image of
24209 the program. Again, this process significantly increases the size of
24210 the generated executable file, but it does not increase the size of
24211 the executable program itself. Furthermore, if this program is run in
24212 the normal manner, it runs exactly as if the debug information were
24213 not present, and takes no more actual memory.
24215 However, if the program is run under control of @code{GDB}, the
24216 debugger is activated. The image of the program is loaded, at which
24217 point it is ready to run. If a run command is given, then the program
24218 will run exactly as it would have if @code{GDB} were not present. This
24219 is a crucial part of the @code{GDB} design philosophy. @code{GDB} is
24220 entirely non-intrusive until a breakpoint is encountered. If no
24221 breakpoint is ever hit, the program will run exactly as it would if no
24222 debugger were present. When a breakpoint is hit, @code{GDB} accesses
24223 the debugging information and can respond to user commands to inspect
24224 variables, and more generally to report on the state of execution.
24228 @section Running GDB
24231 This section describes how to initiate the debugger.
24232 @c The above sentence is really just filler, but it was otherwise
24233 @c clumsy to get the first paragraph nonindented given the conditional
24234 @c nature of the description
24237 The debugger can be launched from a @code{GPS} menu or
24238 directly from the command line. The description below covers the latter use.
24239 All the commands shown can be used in the @code{GPS} debug console window,
24240 but there are usually more GUI-based ways to achieve the same effect.
24243 The command to run @code{GDB} is
24246 $ ^gdb program^GDB PROGRAM^
24250 where @code{^program^PROGRAM^} is the name of the executable file. This
24251 activates the debugger and results in a prompt for debugger commands.
24252 The simplest command is simply @code{run}, which causes the program to run
24253 exactly as if the debugger were not present. The following section
24254 describes some of the additional commands that can be given to @code{GDB}.
24256 @c *******************************
24257 @node Introduction to GDB Commands
24258 @section Introduction to GDB Commands
24261 @code{GDB} contains a large repertoire of commands. @xref{Top,,
24262 Debugging with GDB, gdb, Debugging with GDB},
24264 located in the GNU:[DOCS] directory,
24266 for extensive documentation on the use
24267 of these commands, together with examples of their use. Furthermore,
24268 the command @command{help} invoked from within GDB activates a simple help
24269 facility which summarizes the available commands and their options.
24270 In this section we summarize a few of the most commonly
24271 used commands to give an idea of what @code{GDB} is about. You should create
24272 a simple program with debugging information and experiment with the use of
24273 these @code{GDB} commands on the program as you read through the
24277 @item set args @var{arguments}
24278 The @var{arguments} list above is a list of arguments to be passed to
24279 the program on a subsequent run command, just as though the arguments
24280 had been entered on a normal invocation of the program. The @code{set args}
24281 command is not needed if the program does not require arguments.
24284 The @code{run} command causes execution of the program to start from
24285 the beginning. If the program is already running, that is to say if
24286 you are currently positioned at a breakpoint, then a prompt will ask
24287 for confirmation that you want to abandon the current execution and
24290 @item breakpoint @var{location}
24291 The breakpoint command sets a breakpoint, that is to say a point at which
24292 execution will halt and @code{GDB} will await further
24293 commands. @var{location} is
24294 either a line number within a file, given in the format @code{file:linenumber},
24295 or it is the name of a subprogram. If you request that a breakpoint be set on
24296 a subprogram that is overloaded, a prompt will ask you to specify on which of
24297 those subprograms you want to breakpoint. You can also
24298 specify that all of them should be breakpointed. If the program is run
24299 and execution encounters the breakpoint, then the program
24300 stops and @code{GDB} signals that the breakpoint was encountered by
24301 printing the line of code before which the program is halted.
24303 @item breakpoint exception @var{name}
24304 A special form of the breakpoint command which breakpoints whenever
24305 exception @var{name} is raised.
24306 If @var{name} is omitted,
24307 then a breakpoint will occur when any exception is raised.
24309 @item print @var{expression}
24310 This will print the value of the given expression. Most simple
24311 Ada expression formats are properly handled by @code{GDB}, so the expression
24312 can contain function calls, variables, operators, and attribute references.
24315 Continues execution following a breakpoint, until the next breakpoint or the
24316 termination of the program.
24319 Executes a single line after a breakpoint. If the next statement
24320 is a subprogram call, execution continues into (the first statement of)
24321 the called subprogram.
24324 Executes a single line. If this line is a subprogram call, executes and
24325 returns from the call.
24328 Lists a few lines around the current source location. In practice, it
24329 is usually more convenient to have a separate edit window open with the
24330 relevant source file displayed. Successive applications of this command
24331 print subsequent lines. The command can be given an argument which is a
24332 line number, in which case it displays a few lines around the specified one.
24335 Displays a backtrace of the call chain. This command is typically
24336 used after a breakpoint has occurred, to examine the sequence of calls that
24337 leads to the current breakpoint. The display includes one line for each
24338 activation record (frame) corresponding to an active subprogram.
24341 At a breakpoint, @code{GDB} can display the values of variables local
24342 to the current frame. The command @code{up} can be used to
24343 examine the contents of other active frames, by moving the focus up
24344 the stack, that is to say from callee to caller, one frame at a time.
24347 Moves the focus of @code{GDB} down from the frame currently being
24348 examined to the frame of its callee (the reverse of the previous command),
24350 @item frame @var{n}
24351 Inspect the frame with the given number. The value 0 denotes the frame
24352 of the current breakpoint, that is to say the top of the call stack.
24357 The above list is a very short introduction to the commands that
24358 @code{GDB} provides. Important additional capabilities, including conditional
24359 breakpoints, the ability to execute command sequences on a breakpoint,
24360 the ability to debug at the machine instruction level and many other
24361 features are described in detail in @ref{Top,, Debugging with GDB, gdb,
24362 Debugging with GDB}. Note that most commands can be abbreviated
24363 (for example, c for continue, bt for backtrace).
24365 @node Using Ada Expressions
24366 @section Using Ada Expressions
24367 @cindex Ada expressions
24370 @code{GDB} supports a fairly large subset of Ada expression syntax, with some
24371 extensions. The philosophy behind the design of this subset is
24375 That @code{GDB} should provide basic literals and access to operations for
24376 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
24377 leaving more sophisticated computations to subprograms written into the
24378 program (which therefore may be called from @code{GDB}).
24381 That type safety and strict adherence to Ada language restrictions
24382 are not particularly important to the @code{GDB} user.
24385 That brevity is important to the @code{GDB} user.
24389 Thus, for brevity, the debugger acts as if there were
24390 implicit @code{with} and @code{use} clauses in effect for all user-written
24391 packages, thus making it unnecessary to fully qualify most names with
24392 their packages, regardless of context. Where this causes ambiguity,
24393 @code{GDB} asks the user's intent.
24395 For details on the supported Ada syntax, see @ref{Top,, Debugging with
24396 GDB, gdb, Debugging with GDB}.
24398 @node Calling User-Defined Subprograms
24399 @section Calling User-Defined Subprograms
24402 An important capability of @code{GDB} is the ability to call user-defined
24403 subprograms while debugging. This is achieved simply by entering
24404 a subprogram call statement in the form:
24407 call subprogram-name (parameters)
24411 The keyword @code{call} can be omitted in the normal case where the
24412 @code{subprogram-name} does not coincide with any of the predefined
24413 @code{GDB} commands.
24415 The effect is to invoke the given subprogram, passing it the
24416 list of parameters that is supplied. The parameters can be expressions and
24417 can include variables from the program being debugged. The
24418 subprogram must be defined
24419 at the library level within your program, and @code{GDB} will call the
24420 subprogram within the environment of your program execution (which
24421 means that the subprogram is free to access or even modify variables
24422 within your program).
24424 The most important use of this facility is in allowing the inclusion of
24425 debugging routines that are tailored to particular data structures
24426 in your program. Such debugging routines can be written to provide a suitably
24427 high-level description of an abstract type, rather than a low-level dump
24428 of its physical layout. After all, the standard
24429 @code{GDB print} command only knows the physical layout of your
24430 types, not their abstract meaning. Debugging routines can provide information
24431 at the desired semantic level and are thus enormously useful.
24433 For example, when debugging GNAT itself, it is crucial to have access to
24434 the contents of the tree nodes used to represent the program internally.
24435 But tree nodes are represented simply by an integer value (which in turn
24436 is an index into a table of nodes).
24437 Using the @code{print} command on a tree node would simply print this integer
24438 value, which is not very useful. But the PN routine (defined in file
24439 treepr.adb in the GNAT sources) takes a tree node as input, and displays
24440 a useful high level representation of the tree node, which includes the
24441 syntactic category of the node, its position in the source, the integers
24442 that denote descendant nodes and parent node, as well as varied
24443 semantic information. To study this example in more detail, you might want to
24444 look at the body of the PN procedure in the stated file.
24446 @node Using the Next Command in a Function
24447 @section Using the Next Command in a Function
24450 When you use the @code{next} command in a function, the current source
24451 location will advance to the next statement as usual. A special case
24452 arises in the case of a @code{return} statement.
24454 Part of the code for a return statement is the ``epilog'' of the function.
24455 This is the code that returns to the caller. There is only one copy of
24456 this epilog code, and it is typically associated with the last return
24457 statement in the function if there is more than one return. In some
24458 implementations, this epilog is associated with the first statement
24461 The result is that if you use the @code{next} command from a return
24462 statement that is not the last return statement of the function you
24463 may see a strange apparent jump to the last return statement or to
24464 the start of the function. You should simply ignore this odd jump.
24465 The value returned is always that from the first return statement
24466 that was stepped through.
24468 @node Ada Exceptions
24469 @section Breaking on Ada Exceptions
24473 You can set breakpoints that trip when your program raises
24474 selected exceptions.
24477 @item break exception
24478 Set a breakpoint that trips whenever (any task in the) program raises
24481 @item break exception @var{name}
24482 Set a breakpoint that trips whenever (any task in the) program raises
24483 the exception @var{name}.
24485 @item break exception unhandled
24486 Set a breakpoint that trips whenever (any task in the) program raises an
24487 exception for which there is no handler.
24489 @item info exceptions
24490 @itemx info exceptions @var{regexp}
24491 The @code{info exceptions} command permits the user to examine all defined
24492 exceptions within Ada programs. With a regular expression, @var{regexp}, as
24493 argument, prints out only those exceptions whose name matches @var{regexp}.
24501 @code{GDB} allows the following task-related commands:
24505 This command shows a list of current Ada tasks, as in the following example:
24512 ID TID P-ID Thread Pri State Name
24513 1 8088000 0 807e000 15 Child Activation Wait main_task
24514 2 80a4000 1 80ae000 15 Accept/Select Wait b
24515 3 809a800 1 80a4800 15 Child Activation Wait a
24516 * 4 80ae800 3 80b8000 15 Running c
24520 In this listing, the asterisk before the first task indicates it to be the
24521 currently running task. The first column lists the task ID that is used
24522 to refer to tasks in the following commands.
24524 @item break @var{linespec} task @var{taskid}
24525 @itemx break @var{linespec} task @var{taskid} if @dots{}
24526 @cindex Breakpoints and tasks
24527 These commands are like the @code{break @dots{} thread @dots{}}.
24528 @var{linespec} specifies source lines.
24530 Use the qualifier @samp{task @var{taskid}} with a breakpoint command
24531 to specify that you only want @code{GDB} to stop the program when a
24532 particular Ada task reaches this breakpoint. @var{taskid} is one of the
24533 numeric task identifiers assigned by @code{GDB}, shown in the first
24534 column of the @samp{info tasks} display.
24536 If you do not specify @samp{task @var{taskid}} when you set a
24537 breakpoint, the breakpoint applies to @emph{all} tasks of your
24540 You can use the @code{task} qualifier on conditional breakpoints as
24541 well; in this case, place @samp{task @var{taskid}} before the
24542 breakpoint condition (before the @code{if}).
24544 @item task @var{taskno}
24545 @cindex Task switching
24547 This command allows to switch to the task referred by @var{taskno}. In
24548 particular, This allows to browse the backtrace of the specified
24549 task. It is advised to switch back to the original task before
24550 continuing execution otherwise the scheduling of the program may be
24555 For more detailed information on the tasking support,
24556 see @ref{Top,, Debugging with GDB, gdb, Debugging with GDB}.
24558 @node Debugging Generic Units
24559 @section Debugging Generic Units
24560 @cindex Debugging Generic Units
24564 GNAT always uses code expansion for generic instantiation. This means that
24565 each time an instantiation occurs, a complete copy of the original code is
24566 made, with appropriate substitutions of formals by actuals.
24568 It is not possible to refer to the original generic entities in
24569 @code{GDB}, but it is always possible to debug a particular instance of
24570 a generic, by using the appropriate expanded names. For example, if we have
24572 @smallexample @c ada
24577 generic package k is
24578 procedure kp (v1 : in out integer);
24582 procedure kp (v1 : in out integer) is
24588 package k1 is new k;
24589 package k2 is new k;
24591 var : integer := 1;
24604 Then to break on a call to procedure kp in the k2 instance, simply
24608 (gdb) break g.k2.kp
24612 When the breakpoint occurs, you can step through the code of the
24613 instance in the normal manner and examine the values of local variables, as for
24616 @node GNAT Abnormal Termination or Failure to Terminate
24617 @section GNAT Abnormal Termination or Failure to Terminate
24618 @cindex GNAT Abnormal Termination or Failure to Terminate
24621 When presented with programs that contain serious errors in syntax
24623 GNAT may on rare occasions experience problems in operation, such
24625 segmentation fault or illegal memory access, raising an internal
24626 exception, terminating abnormally, or failing to terminate at all.
24627 In such cases, you can activate
24628 various features of GNAT that can help you pinpoint the construct in your
24629 program that is the likely source of the problem.
24631 The following strategies are presented in increasing order of
24632 difficulty, corresponding to your experience in using GNAT and your
24633 familiarity with compiler internals.
24637 Run @command{gcc} with the @option{-gnatf}. This first
24638 switch causes all errors on a given line to be reported. In its absence,
24639 only the first error on a line is displayed.
24641 The @option{-gnatdO} switch causes errors to be displayed as soon as they
24642 are encountered, rather than after compilation is terminated. If GNAT
24643 terminates prematurely or goes into an infinite loop, the last error
24644 message displayed may help to pinpoint the culprit.
24647 Run @command{gcc} with the @option{^-v (verbose)^/VERBOSE^} switch. In this
24648 mode, @command{gcc} produces ongoing information about the progress of the
24649 compilation and provides the name of each procedure as code is
24650 generated. This switch allows you to find which Ada procedure was being
24651 compiled when it encountered a code generation problem.
24654 @cindex @option{-gnatdc} switch
24655 Run @command{gcc} with the @option{-gnatdc} switch. This is a GNAT specific
24656 switch that does for the front-end what @option{^-v^VERBOSE^} does
24657 for the back end. The system prints the name of each unit,
24658 either a compilation unit or nested unit, as it is being analyzed.
24660 Finally, you can start
24661 @code{gdb} directly on the @code{gnat1} executable. @code{gnat1} is the
24662 front-end of GNAT, and can be run independently (normally it is just
24663 called from @command{gcc}). You can use @code{gdb} on @code{gnat1} as you
24664 would on a C program (but @pxref{The GNAT Debugger GDB} for caveats). The
24665 @code{where} command is the first line of attack; the variable
24666 @code{lineno} (seen by @code{print lineno}), used by the second phase of
24667 @code{gnat1} and by the @command{gcc} backend, indicates the source line at
24668 which the execution stopped, and @code{input_file name} indicates the name of
24672 @node Naming Conventions for GNAT Source Files
24673 @section Naming Conventions for GNAT Source Files
24676 In order to examine the workings of the GNAT system, the following
24677 brief description of its organization may be helpful:
24681 Files with prefix @file{^sc^SC^} contain the lexical scanner.
24684 All files prefixed with @file{^par^PAR^} are components of the parser. The
24685 numbers correspond to chapters of the Ada Reference Manual. For example,
24686 parsing of select statements can be found in @file{par-ch9.adb}.
24689 All files prefixed with @file{^sem^SEM^} perform semantic analysis. The
24690 numbers correspond to chapters of the Ada standard. For example, all
24691 issues involving context clauses can be found in @file{sem_ch10.adb}. In
24692 addition, some features of the language require sufficient special processing
24693 to justify their own semantic files: sem_aggr for aggregates, sem_disp for
24694 dynamic dispatching, etc.
24697 All files prefixed with @file{^exp^EXP^} perform normalization and
24698 expansion of the intermediate representation (abstract syntax tree, or AST).
24699 these files use the same numbering scheme as the parser and semantics files.
24700 For example, the construction of record initialization procedures is done in
24701 @file{exp_ch3.adb}.
24704 The files prefixed with @file{^bind^BIND^} implement the binder, which
24705 verifies the consistency of the compilation, determines an order of
24706 elaboration, and generates the bind file.
24709 The files @file{atree.ads} and @file{atree.adb} detail the low-level
24710 data structures used by the front-end.
24713 The files @file{sinfo.ads} and @file{sinfo.adb} detail the structure of
24714 the abstract syntax tree as produced by the parser.
24717 The files @file{einfo.ads} and @file{einfo.adb} detail the attributes of
24718 all entities, computed during semantic analysis.
24721 Library management issues are dealt with in files with prefix
24727 Ada files with the prefix @file{^a-^A-^} are children of @code{Ada}, as
24728 defined in Annex A.
24733 Files with prefix @file{^i-^I-^} are children of @code{Interfaces}, as
24734 defined in Annex B.
24738 Files with prefix @file{^s-^S-^} are children of @code{System}. This includes
24739 both language-defined children and GNAT run-time routines.
24743 Files with prefix @file{^g-^G-^} are children of @code{GNAT}. These are useful
24744 general-purpose packages, fully documented in their specs. All
24745 the other @file{.c} files are modifications of common @command{gcc} files.
24748 @node Getting Internal Debugging Information
24749 @section Getting Internal Debugging Information
24752 Most compilers have internal debugging switches and modes. GNAT
24753 does also, except GNAT internal debugging switches and modes are not
24754 secret. A summary and full description of all the compiler and binder
24755 debug flags are in the file @file{debug.adb}. You must obtain the
24756 sources of the compiler to see the full detailed effects of these flags.
24758 The switches that print the source of the program (reconstructed from
24759 the internal tree) are of general interest for user programs, as are the
24761 the full internal tree, and the entity table (the symbol table
24762 information). The reconstructed source provides a readable version of the
24763 program after the front-end has completed analysis and expansion,
24764 and is useful when studying the performance of specific constructs.
24765 For example, constraint checks are indicated, complex aggregates
24766 are replaced with loops and assignments, and tasking primitives
24767 are replaced with run-time calls.
24769 @node Stack Traceback
24770 @section Stack Traceback
24772 @cindex stack traceback
24773 @cindex stack unwinding
24776 Traceback is a mechanism to display the sequence of subprogram calls that
24777 leads to a specified execution point in a program. Often (but not always)
24778 the execution point is an instruction at which an exception has been raised.
24779 This mechanism is also known as @i{stack unwinding} because it obtains
24780 its information by scanning the run-time stack and recovering the activation
24781 records of all active subprograms. Stack unwinding is one of the most
24782 important tools for program debugging.
24784 The first entry stored in traceback corresponds to the deepest calling level,
24785 that is to say the subprogram currently executing the instruction
24786 from which we want to obtain the traceback.
24788 Note that there is no runtime performance penalty when stack traceback
24789 is enabled, and no exception is raised during program execution.
24792 * Non-Symbolic Traceback::
24793 * Symbolic Traceback::
24796 @node Non-Symbolic Traceback
24797 @subsection Non-Symbolic Traceback
24798 @cindex traceback, non-symbolic
24801 Note: this feature is not supported on all platforms. See
24802 @file{GNAT.Traceback spec in g-traceb.ads} for a complete list of supported
24806 * Tracebacks From an Unhandled Exception::
24807 * Tracebacks From Exception Occurrences (non-symbolic)::
24808 * Tracebacks From Anywhere in a Program (non-symbolic)::
24811 @node Tracebacks From an Unhandled Exception
24812 @subsubsection Tracebacks From an Unhandled Exception
24815 A runtime non-symbolic traceback is a list of addresses of call instructions.
24816 To enable this feature you must use the @option{-E}
24817 @code{gnatbind}'s option. With this option a stack traceback is stored as part
24818 of exception information. You can retrieve this information using the
24819 @code{addr2line} tool.
24821 Here is a simple example:
24823 @smallexample @c ada
24829 raise Constraint_Error;
24844 $ gnatmake stb -bargs -E
24847 Execution terminated by unhandled exception
24848 Exception name: CONSTRAINT_ERROR
24850 Call stack traceback locations:
24851 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
24855 As we see the traceback lists a sequence of addresses for the unhandled
24856 exception @code{CONSTRAINT_ERROR} raised in procedure P1. It is easy to
24857 guess that this exception come from procedure P1. To translate these
24858 addresses into the source lines where the calls appear, the
24859 @code{addr2line} tool, described below, is invaluable. The use of this tool
24860 requires the program to be compiled with debug information.
24863 $ gnatmake -g stb -bargs -E
24866 Execution terminated by unhandled exception
24867 Exception name: CONSTRAINT_ERROR
24869 Call stack traceback locations:
24870 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
24872 $ addr2line --exe=stb 0x401373 0x40138b 0x40139c 0x401335 0x4011c4
24873 0x4011f1 0x77e892a4
24875 00401373 at d:/stb/stb.adb:5
24876 0040138B at d:/stb/stb.adb:10
24877 0040139C at d:/stb/stb.adb:14
24878 00401335 at d:/stb/b~stb.adb:104
24879 004011C4 at /build/@dots{}/crt1.c:200
24880 004011F1 at /build/@dots{}/crt1.c:222
24881 77E892A4 in ?? at ??:0
24885 The @code{addr2line} tool has several other useful options:
24889 to get the function name corresponding to any location
24891 @item --demangle=gnat
24892 to use the gnat decoding mode for the function names. Note that
24893 for binutils version 2.9.x the option is simply @option{--demangle}.
24897 $ addr2line --exe=stb --functions --demangle=gnat 0x401373 0x40138b
24898 0x40139c 0x401335 0x4011c4 0x4011f1
24900 00401373 in stb.p1 at d:/stb/stb.adb:5
24901 0040138B in stb.p2 at d:/stb/stb.adb:10
24902 0040139C in stb at d:/stb/stb.adb:14
24903 00401335 in main at d:/stb/b~stb.adb:104
24904 004011C4 in <__mingw_CRTStartup> at /build/@dots{}/crt1.c:200
24905 004011F1 in <mainCRTStartup> at /build/@dots{}/crt1.c:222
24909 From this traceback we can see that the exception was raised in
24910 @file{stb.adb} at line 5, which was reached from a procedure call in
24911 @file{stb.adb} at line 10, and so on. The @file{b~std.adb} is the binder file,
24912 which contains the call to the main program.
24913 @xref{Running gnatbind}. The remaining entries are assorted runtime routines,
24914 and the output will vary from platform to platform.
24916 It is also possible to use @code{GDB} with these traceback addresses to debug
24917 the program. For example, we can break at a given code location, as reported
24918 in the stack traceback:
24924 Furthermore, this feature is not implemented inside Windows DLL. Only
24925 the non-symbolic traceback is reported in this case.
24928 (gdb) break *0x401373
24929 Breakpoint 1 at 0x401373: file stb.adb, line 5.
24933 It is important to note that the stack traceback addresses
24934 do not change when debug information is included. This is particularly useful
24935 because it makes it possible to release software without debug information (to
24936 minimize object size), get a field report that includes a stack traceback
24937 whenever an internal bug occurs, and then be able to retrieve the sequence
24938 of calls with the same program compiled with debug information.
24940 @node Tracebacks From Exception Occurrences (non-symbolic)
24941 @subsubsection Tracebacks From Exception Occurrences
24944 Non-symbolic tracebacks are obtained by using the @option{-E} binder argument.
24945 The stack traceback is attached to the exception information string, and can
24946 be retrieved in an exception handler within the Ada program, by means of the
24947 Ada facilities defined in @code{Ada.Exceptions}. Here is a simple example:
24949 @smallexample @c ada
24951 with Ada.Exceptions;
24956 use Ada.Exceptions;
24964 Text_IO.Put_Line (Exception_Information (E));
24978 This program will output:
24983 Exception name: CONSTRAINT_ERROR
24984 Message: stb.adb:12
24985 Call stack traceback locations:
24986 0x4015e4 0x401633 0x401644 0x401461 0x4011c4 0x4011f1 0x77e892a4
24989 @node Tracebacks From Anywhere in a Program (non-symbolic)
24990 @subsubsection Tracebacks From Anywhere in a Program
24993 It is also possible to retrieve a stack traceback from anywhere in a
24994 program. For this you need to
24995 use the @code{GNAT.Traceback} API. This package includes a procedure called
24996 @code{Call_Chain} that computes a complete stack traceback, as well as useful
24997 display procedures described below. It is not necessary to use the
24998 @option{-E gnatbind} option in this case, because the stack traceback mechanism
24999 is invoked explicitly.
25002 In the following example we compute a traceback at a specific location in
25003 the program, and we display it using @code{GNAT.Debug_Utilities.Image} to
25004 convert addresses to strings:
25006 @smallexample @c ada
25008 with GNAT.Traceback;
25009 with GNAT.Debug_Utilities;
25015 use GNAT.Traceback;
25018 TB : Tracebacks_Array (1 .. 10);
25019 -- We are asking for a maximum of 10 stack frames.
25021 -- Len will receive the actual number of stack frames returned.
25023 Call_Chain (TB, Len);
25025 Text_IO.Put ("In STB.P1 : ");
25027 for K in 1 .. Len loop
25028 Text_IO.Put (Debug_Utilities.Image (TB (K)));
25049 In STB.P1 : 16#0040_F1E4# 16#0040_14F2# 16#0040_170B# 16#0040_171C#
25050 16#0040_1461# 16#0040_11C4# 16#0040_11F1# 16#77E8_92A4#
25054 You can then get further information by invoking the @code{addr2line}
25055 tool as described earlier (note that the hexadecimal addresses
25056 need to be specified in C format, with a leading ``0x'').
25058 @node Symbolic Traceback
25059 @subsection Symbolic Traceback
25060 @cindex traceback, symbolic
25063 A symbolic traceback is a stack traceback in which procedure names are
25064 associated with each code location.
25067 Note that this feature is not supported on all platforms. See
25068 @file{GNAT.Traceback.Symbolic spec in g-trasym.ads} for a complete
25069 list of currently supported platforms.
25072 Note that the symbolic traceback requires that the program be compiled
25073 with debug information. If it is not compiled with debug information
25074 only the non-symbolic information will be valid.
25077 * Tracebacks From Exception Occurrences (symbolic)::
25078 * Tracebacks From Anywhere in a Program (symbolic)::
25081 @node Tracebacks From Exception Occurrences (symbolic)
25082 @subsubsection Tracebacks From Exception Occurrences
25084 @smallexample @c ada
25086 with GNAT.Traceback.Symbolic;
25092 raise Constraint_Error;
25109 Ada.Text_IO.Put_Line (GNAT.Traceback.Symbolic.Symbolic_Traceback (E));
25114 $ gnatmake -g .\stb -bargs -E -largs -lgnat -laddr2line -lintl
25117 0040149F in stb.p1 at stb.adb:8
25118 004014B7 in stb.p2 at stb.adb:13
25119 004014CF in stb.p3 at stb.adb:18
25120 004015DD in ada.stb at stb.adb:22
25121 00401461 in main at b~stb.adb:168
25122 004011C4 in __mingw_CRTStartup at crt1.c:200
25123 004011F1 in mainCRTStartup at crt1.c:222
25124 77E892A4 in ?? at ??:0
25128 In the above example the ``.\'' syntax in the @command{gnatmake} command
25129 is currently required by @command{addr2line} for files that are in
25130 the current working directory.
25131 Moreover, the exact sequence of linker options may vary from platform
25133 The above @option{-largs} section is for Windows platforms. By contrast,
25134 under Unix there is no need for the @option{-largs} section.
25135 Differences across platforms are due to details of linker implementation.
25137 @node Tracebacks From Anywhere in a Program (symbolic)
25138 @subsubsection Tracebacks From Anywhere in a Program
25141 It is possible to get a symbolic stack traceback
25142 from anywhere in a program, just as for non-symbolic tracebacks.
25143 The first step is to obtain a non-symbolic
25144 traceback, and then call @code{Symbolic_Traceback} to compute the symbolic
25145 information. Here is an example:
25147 @smallexample @c ada
25149 with GNAT.Traceback;
25150 with GNAT.Traceback.Symbolic;
25155 use GNAT.Traceback;
25156 use GNAT.Traceback.Symbolic;
25159 TB : Tracebacks_Array (1 .. 10);
25160 -- We are asking for a maximum of 10 stack frames.
25162 -- Len will receive the actual number of stack frames returned.
25164 Call_Chain (TB, Len);
25165 Text_IO.Put_Line (Symbolic_Traceback (TB (1 .. Len)));
25178 @c ******************************
25180 @node Compatibility with HP Ada
25181 @chapter Compatibility with HP Ada
25182 @cindex Compatibility
25187 @cindex Compatibility between GNAT and HP Ada
25188 This chapter compares HP Ada (formerly known as ``DEC Ada'')
25189 for OpenVMS Alpha and GNAT for OpenVMS for Alpha and for I64.
25190 GNAT is highly compatible
25191 with HP Ada, and it should generally be straightforward to port code
25192 from the HP Ada environment to GNAT. However, there are a few language
25193 and implementation differences of which the user must be aware. These
25194 differences are discussed in this chapter. In
25195 addition, the operating environment and command structure for the
25196 compiler are different, and these differences are also discussed.
25198 For further details on these and other compatibility issues,
25199 see Appendix E of the HP publication
25200 @cite{HP Ada, Technical Overview and Comparison on HP Platforms}.
25202 Except where otherwise indicated, the description of GNAT for OpenVMS
25203 applies to both the Alpha and I64 platforms.
25205 For information on porting Ada code from GNAT on Alpha OpenVMS to GNAT on
25206 I64 OpenVMS, see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
25208 The discussion in this chapter addresses specifically the implementation
25209 of Ada 83 for HP OpenVMS Alpha Systems. In cases where the implementation
25210 of HP Ada differs between OpenVMS Alpha Systems and OpenVMS VAX Systems,
25211 GNAT always follows the Alpha implementation.
25213 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
25214 attributes are recognized, although only a subset of them can sensibly
25215 be implemented. The description of pragmas in
25216 @xref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference Manual},
25217 indicates whether or not they are applicable to non-VMS systems.
25220 * Ada Language Compatibility::
25221 * Differences in the Definition of Package System::
25222 * Language-Related Features::
25223 * The Package STANDARD::
25224 * The Package SYSTEM::
25225 * Tasking and Task-Related Features::
25226 * Pragmas and Pragma-Related Features::
25227 * Library of Predefined Units::
25229 * Main Program Definition::
25230 * Implementation-Defined Attributes::
25231 * Compiler and Run-Time Interfacing::
25232 * Program Compilation and Library Management::
25234 * Implementation Limits::
25235 * Tools and Utilities::
25238 @node Ada Language Compatibility
25239 @section Ada Language Compatibility
25242 GNAT handles Ada 95 and Ada 2005 as well as Ada 83, whereas HP Ada is only
25243 for Ada 83. Ada 95 and Ada 2005 are almost completely upwards compatible
25244 with Ada 83, and therefore Ada 83 programs will compile
25245 and run under GNAT with
25246 no changes or only minor changes. The @cite{Annotated Ada Reference Manual}
25247 provides details on specific incompatibilities.
25249 GNAT provides the switch @option{/83} on the @command{GNAT COMPILE} command,
25250 as well as the pragma @code{ADA_83}, to force the compiler to
25251 operate in Ada 83 mode. This mode does not guarantee complete
25252 conformance to Ada 83, but in practice is sufficient to
25253 eliminate most sources of incompatibilities.
25254 In particular, it eliminates the recognition of the
25255 additional Ada 95 and Ada 2005 keywords, so that their use as identifiers
25256 in Ada 83 programs is legal, and handles the cases of packages
25257 with optional bodies, and generics that instantiate unconstrained
25258 types without the use of @code{(<>)}.
25260 @node Differences in the Definition of Package System
25261 @section Differences in the Definition of Package @code{System}
25264 An Ada compiler is allowed to add
25265 implementation-dependent declarations to package @code{System}.
25267 GNAT does not take advantage of this permission, and the version of
25268 @code{System} provided by GNAT exactly matches that defined in the Ada
25271 However, HP Ada adds an extensive set of declarations to package
25273 as fully documented in the HP Ada manuals. To minimize changes required
25274 for programs that make use of these extensions, GNAT provides the pragma
25275 @code{Extend_System} for extending the definition of package System. By using:
25276 @cindex pragma @code{Extend_System}
25277 @cindex @code{Extend_System} pragma
25279 @smallexample @c ada
25282 pragma Extend_System (Aux_DEC);
25288 the set of definitions in @code{System} is extended to include those in
25289 package @code{System.Aux_DEC}.
25290 @cindex @code{System.Aux_DEC} package
25291 @cindex @code{Aux_DEC} package (child of @code{System})
25292 These definitions are incorporated directly into package @code{System},
25293 as though they had been declared there. For a
25294 list of the declarations added, see the spec of this package,
25295 which can be found in the file @file{s-auxdec.ads} in the GNAT library.
25296 @cindex @file{s-auxdec.ads} file
25297 The pragma @code{Extend_System} is a configuration pragma, which means that
25298 it can be placed in the file @file{gnat.adc}, so that it will automatically
25299 apply to all subsequent compilations. See @ref{Configuration Pragmas},
25300 for further details.
25302 An alternative approach that avoids the use of the non-standard
25303 @code{Extend_System} pragma is to add a context clause to the unit that
25304 references these facilities:
25306 @smallexample @c ada
25308 with System.Aux_DEC;
25309 use System.Aux_DEC;
25314 The effect is not quite semantically identical to incorporating
25315 the declarations directly into package @code{System},
25316 but most programs will not notice a difference
25317 unless they use prefix notation (e.g.@: @code{System.Integer_8})
25318 to reference the entities directly in package @code{System}.
25319 For units containing such references,
25320 the prefixes must either be removed, or the pragma @code{Extend_System}
25323 @node Language-Related Features
25324 @section Language-Related Features
25327 The following sections highlight differences in types,
25328 representations of types, operations, alignment, and
25332 * Integer Types and Representations::
25333 * Floating-Point Types and Representations::
25334 * Pragmas Float_Representation and Long_Float::
25335 * Fixed-Point Types and Representations::
25336 * Record and Array Component Alignment::
25337 * Address Clauses::
25338 * Other Representation Clauses::
25341 @node Integer Types and Representations
25342 @subsection Integer Types and Representations
25345 The set of predefined integer types is identical in HP Ada and GNAT.
25346 Furthermore the representation of these integer types is also identical,
25347 including the capability of size clauses forcing biased representation.
25350 HP Ada for OpenVMS Alpha systems has defined the
25351 following additional integer types in package @code{System}:
25368 @code{LARGEST_INTEGER}
25372 In GNAT, the first four of these types may be obtained from the
25373 standard Ada package @code{Interfaces}.
25374 Alternatively, by use of the pragma @code{Extend_System}, identical
25375 declarations can be referenced directly in package @code{System}.
25376 On both GNAT and HP Ada, the maximum integer size is 64 bits.
25378 @node Floating-Point Types and Representations
25379 @subsection Floating-Point Types and Representations
25380 @cindex Floating-Point types
25383 The set of predefined floating-point types is identical in HP Ada and GNAT.
25384 Furthermore the representation of these floating-point
25385 types is also identical. One important difference is that the default
25386 representation for HP Ada is @code{VAX_Float}, but the default representation
25389 Specific types may be declared to be @code{VAX_Float} or IEEE, using the
25390 pragma @code{Float_Representation} as described in the HP Ada
25392 For example, the declarations:
25394 @smallexample @c ada
25396 type F_Float is digits 6;
25397 pragma Float_Representation (VAX_Float, F_Float);
25402 declares a type @code{F_Float} that will be represented in @code{VAX_Float}
25404 This set of declarations actually appears in @code{System.Aux_DEC},
25406 the full set of additional floating-point declarations provided in
25407 the HP Ada version of package @code{System}.
25408 This and similar declarations may be accessed in a user program
25409 by using pragma @code{Extend_System}. The use of this
25410 pragma, and the related pragma @code{Long_Float} is described in further
25411 detail in the following section.
25413 @node Pragmas Float_Representation and Long_Float
25414 @subsection Pragmas @code{Float_Representation} and @code{Long_Float}
25417 HP Ada provides the pragma @code{Float_Representation}, which
25418 acts as a program library switch to allow control over
25419 the internal representation chosen for the predefined
25420 floating-point types declared in the package @code{Standard}.
25421 The format of this pragma is as follows:
25423 @smallexample @c ada
25425 pragma Float_Representation(VAX_Float | IEEE_Float);
25430 This pragma controls the representation of floating-point
25435 @code{VAX_Float} specifies that floating-point
25436 types are represented by default with the VAX system hardware types
25437 @code{F-floating}, @code{D-floating}, @code{G-floating}.
25438 Note that the @code{H-floating}
25439 type was available only on VAX systems, and is not available
25440 in either HP Ada or GNAT.
25443 @code{IEEE_Float} specifies that floating-point
25444 types are represented by default with the IEEE single and
25445 double floating-point types.
25449 GNAT provides an identical implementation of the pragma
25450 @code{Float_Representation}, except that it functions as a
25451 configuration pragma. Note that the
25452 notion of configuration pragma corresponds closely to the
25453 HP Ada notion of a program library switch.
25455 When no pragma is used in GNAT, the default is @code{IEEE_Float},
25457 from HP Ada 83, where the default is @code{VAX_Float}. In addition, the
25458 predefined libraries in GNAT are built using @code{IEEE_Float}, so it is not
25459 advisable to change the format of numbers passed to standard library
25460 routines, and if necessary explicit type conversions may be needed.
25462 The use of @code{IEEE_Float} is recommended in GNAT since it is more
25463 efficient, and (given that it conforms to an international standard)
25464 potentially more portable.
25465 The situation in which @code{VAX_Float} may be useful is in interfacing
25466 to existing code and data that expect the use of @code{VAX_Float}.
25467 In such a situation use the predefined @code{VAX_Float}
25468 types in package @code{System}, as extended by
25469 @code{Extend_System}. For example, use @code{System.F_Float}
25470 to specify the 32-bit @code{F-Float} format.
25473 On OpenVMS systems, HP Ada provides the pragma @code{Long_Float}
25474 to allow control over the internal representation chosen
25475 for the predefined type @code{Long_Float} and for floating-point
25476 type declarations with digits specified in the range 7 .. 15.
25477 The format of this pragma is as follows:
25479 @smallexample @c ada
25481 pragma Long_Float (D_FLOAT | G_FLOAT);
25485 @node Fixed-Point Types and Representations
25486 @subsection Fixed-Point Types and Representations
25489 On HP Ada for OpenVMS Alpha systems, rounding is
25490 away from zero for both positive and negative numbers.
25491 Therefore, @code{+0.5} rounds to @code{1},
25492 and @code{-0.5} rounds to @code{-1}.
25494 On GNAT the results of operations
25495 on fixed-point types are in accordance with the Ada
25496 rules. In particular, results of operations on decimal
25497 fixed-point types are truncated.
25499 @node Record and Array Component Alignment
25500 @subsection Record and Array Component Alignment
25503 On HP Ada for OpenVMS Alpha, all non-composite components
25504 are aligned on natural boundaries. For example, 1-byte
25505 components are aligned on byte boundaries, 2-byte
25506 components on 2-byte boundaries, 4-byte components on 4-byte
25507 byte boundaries, and so on. The OpenVMS Alpha hardware
25508 runs more efficiently with naturally aligned data.
25510 On GNAT, alignment rules are compatible
25511 with HP Ada for OpenVMS Alpha.
25513 @node Address Clauses
25514 @subsection Address Clauses
25517 In HP Ada and GNAT, address clauses are supported for
25518 objects and imported subprograms.
25519 The predefined type @code{System.Address} is a private type
25520 in both compilers on Alpha OpenVMS, with the same representation
25521 (it is simply a machine pointer). Addition, subtraction, and comparison
25522 operations are available in the standard Ada package
25523 @code{System.Storage_Elements}, or in package @code{System}
25524 if it is extended to include @code{System.Aux_DEC} using a
25525 pragma @code{Extend_System} as previously described.
25527 Note that code that @code{with}'s both this extended package @code{System}
25528 and the package @code{System.Storage_Elements} should not @code{use}
25529 both packages, or ambiguities will result. In general it is better
25530 not to mix these two sets of facilities. The Ada package was
25531 designed specifically to provide the kind of features that HP Ada
25532 adds directly to package @code{System}.
25534 The type @code{System.Address} is a 64-bit integer type in GNAT for
25535 I64 OpenVMS. For more information,
25536 see @ref{Transitioning to 64-Bit GNAT for OpenVMS}.
25538 GNAT is compatible with HP Ada in its handling of address
25539 clauses, except for some limitations in
25540 the form of address clauses for composite objects with
25541 initialization. Such address clauses are easily replaced
25542 by the use of an explicitly-defined constant as described
25543 in the Ada Reference Manual (13.1(22)). For example, the sequence
25546 @smallexample @c ada
25548 X, Y : Integer := Init_Func;
25549 Q : String (X .. Y) := "abc";
25551 for Q'Address use Compute_Address;
25556 will be rejected by GNAT, since the address cannot be computed at the time
25557 that @code{Q} is declared. To achieve the intended effect, write instead:
25559 @smallexample @c ada
25562 X, Y : Integer := Init_Func;
25563 Q_Address : constant Address := Compute_Address;
25564 Q : String (X .. Y) := "abc";
25566 for Q'Address use Q_Address;
25572 which will be accepted by GNAT (and other Ada compilers), and is also
25573 compatible with Ada 83. A fuller description of the restrictions
25574 on address specifications is found in @ref{Top, GNAT Reference Manual,
25575 About This Guide, gnat_rm, GNAT Reference Manual}.
25577 @node Other Representation Clauses
25578 @subsection Other Representation Clauses
25581 GNAT implements in a compatible manner all the representation
25582 clauses supported by HP Ada. In addition, GNAT
25583 implements the representation clause forms that were introduced in Ada 95,
25584 including @code{COMPONENT_SIZE} and @code{SIZE} clauses for objects.
25586 @node The Package STANDARD
25587 @section The Package @code{STANDARD}
25590 The package @code{STANDARD}, as implemented by HP Ada, is fully
25591 described in the @cite{Ada Reference Manual} and in the
25592 @cite{HP Ada Language Reference Manual}. As implemented by GNAT, the
25593 package @code{STANDARD} is described in the @cite{Ada Reference Manual}.
25595 In addition, HP Ada supports the Latin-1 character set in
25596 the type @code{CHARACTER}. GNAT supports the Latin-1 character set
25597 in the type @code{CHARACTER} and also Unicode (ISO 10646 BMP) in
25598 the type @code{WIDE_CHARACTER}.
25600 The floating-point types supported by GNAT are those
25601 supported by HP Ada, but the defaults are different, and are controlled by
25602 pragmas. See @ref{Floating-Point Types and Representations}, for details.
25604 @node The Package SYSTEM
25605 @section The Package @code{SYSTEM}
25608 HP Ada provides a specific version of the package
25609 @code{SYSTEM} for each platform on which the language is implemented.
25610 For the complete spec of the package @code{SYSTEM}, see
25611 Appendix F of the @cite{HP Ada Language Reference Manual}.
25613 On HP Ada, the package @code{SYSTEM} includes the following conversion
25616 @item @code{TO_ADDRESS(INTEGER)}
25618 @item @code{TO_ADDRESS(UNSIGNED_LONGWORD)}
25620 @item @code{TO_ADDRESS(}@i{universal_integer}@code{)}
25622 @item @code{TO_INTEGER(ADDRESS)}
25624 @item @code{TO_UNSIGNED_LONGWORD(ADDRESS)}
25626 @item Function @code{IMPORT_VALUE return UNSIGNED_LONGWORD} and the
25627 functions @code{IMPORT_ADDRESS} and @code{IMPORT_LARGEST_VALUE}
25631 By default, GNAT supplies a version of @code{SYSTEM} that matches
25632 the definition given in the @cite{Ada Reference Manual}.
25634 is a subset of the HP system definitions, which is as
25635 close as possible to the original definitions. The only difference
25636 is that the definition of @code{SYSTEM_NAME} is different:
25638 @smallexample @c ada
25640 type Name is (SYSTEM_NAME_GNAT);
25641 System_Name : constant Name := SYSTEM_NAME_GNAT;
25646 Also, GNAT adds the Ada declarations for
25647 @code{BIT_ORDER} and @code{DEFAULT_BIT_ORDER}.
25649 However, the use of the following pragma causes GNAT
25650 to extend the definition of package @code{SYSTEM} so that it
25651 encompasses the full set of HP-specific extensions,
25652 including the functions listed above:
25654 @smallexample @c ada
25656 pragma Extend_System (Aux_DEC);
25661 The pragma @code{Extend_System} is a configuration pragma that
25662 is most conveniently placed in the @file{gnat.adc} file. @xref{Pragma
25663 Extend_System,,, gnat_rm, GNAT Reference Manual}, for further details.
25665 HP Ada does not allow the recompilation of the package
25666 @code{SYSTEM}. Instead HP Ada provides several pragmas
25667 (@code{SYSTEM_NAME}, @code{STORAGE_UNIT}, and @code{MEMORY_SIZE})
25668 to modify values in the package @code{SYSTEM}.
25669 On OpenVMS Alpha systems, the pragma
25670 @code{SYSTEM_NAME} takes the enumeration literal @code{OPENVMS_AXP} as
25671 its single argument.
25673 GNAT does permit the recompilation of package @code{SYSTEM} using
25674 the special switch @option{-gnatg}, and this switch can be used if
25675 it is necessary to modify the definitions in @code{SYSTEM}. GNAT does
25676 not permit the specification of @code{SYSTEM_NAME}, @code{STORAGE_UNIT}
25677 or @code{MEMORY_SIZE} by any other means.
25679 On GNAT systems, the pragma @code{SYSTEM_NAME} takes the
25680 enumeration literal @code{SYSTEM_NAME_GNAT}.
25682 The definitions provided by the use of
25684 @smallexample @c ada
25685 pragma Extend_System (AUX_Dec);
25689 are virtually identical to those provided by the HP Ada 83 package
25690 @code{SYSTEM}. One important difference is that the name of the
25692 function for type @code{UNSIGNED_LONGWORD} is changed to
25693 @code{TO_ADDRESS_LONG}.
25694 @xref{Address Clauses,,, gnat_rm, GNAT Reference Manual}, for a
25695 discussion of why this change was necessary.
25698 The version of @code{TO_ADDRESS} taking a @i{universal_integer} argument
25700 an extension to Ada 83 not strictly compatible with the reference manual.
25701 GNAT, in order to be exactly compatible with the standard,
25702 does not provide this capability. In HP Ada 83, the
25703 point of this definition is to deal with a call like:
25705 @smallexample @c ada
25706 TO_ADDRESS (16#12777#);
25710 Normally, according to Ada 83 semantics, one would expect this to be
25711 ambiguous, since it matches both the @code{INTEGER} and
25712 @code{UNSIGNED_LONGWORD} forms of @code{TO_ADDRESS}.
25713 However, in HP Ada 83, there is no ambiguity, since the
25714 definition using @i{universal_integer} takes precedence.
25716 In GNAT, since the version with @i{universal_integer} cannot be supplied,
25718 not possible to be 100% compatible. Since there are many programs using
25719 numeric constants for the argument to @code{TO_ADDRESS}, the decision in
25721 to change the name of the function in the @code{UNSIGNED_LONGWORD} case,
25722 so the declarations provided in the GNAT version of @code{AUX_Dec} are:
25724 @smallexample @c ada
25725 function To_Address (X : Integer) return Address;
25726 pragma Pure_Function (To_Address);
25728 function To_Address_Long (X : Unsigned_Longword) return Address;
25729 pragma Pure_Function (To_Address_Long);
25733 This means that programs using @code{TO_ADDRESS} for
25734 @code{UNSIGNED_LONGWORD} must change the name to @code{TO_ADDRESS_LONG}.
25736 @node Tasking and Task-Related Features
25737 @section Tasking and Task-Related Features
25740 This section compares the treatment of tasking in GNAT
25741 and in HP Ada for OpenVMS Alpha.
25742 The GNAT description applies to both Alpha and I64 OpenVMS.
25743 For detailed information on tasking in
25744 HP Ada, see the @cite{HP Ada Language Reference Manual} and the
25745 relevant run-time reference manual.
25748 * Implementation of Tasks in HP Ada for OpenVMS Alpha Systems::
25749 * Assigning Task IDs::
25750 * Task IDs and Delays::
25751 * Task-Related Pragmas::
25752 * Scheduling and Task Priority::
25754 * External Interrupts::
25757 @node Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
25758 @subsection Implementation of Tasks in HP Ada for OpenVMS Alpha Systems
25761 On OpenVMS Alpha systems, each Ada task (except a passive
25762 task) is implemented as a single stream of execution
25763 that is created and managed by the kernel. On these
25764 systems, HP Ada tasking support is based on DECthreads,
25765 an implementation of the POSIX standard for threads.
25767 Also, on OpenVMS Alpha systems, HP Ada tasks and foreign
25768 code that calls DECthreads routines can be used together.
25769 The interaction between Ada tasks and DECthreads routines
25770 can have some benefits. For example when on OpenVMS Alpha,
25771 HP Ada can call C code that is already threaded.
25773 GNAT uses the facilities of DECthreads,
25774 and Ada tasks are mapped to threads.
25776 @node Assigning Task IDs
25777 @subsection Assigning Task IDs
25780 The HP Ada Run-Time Library always assigns @code{%TASK 1} to
25781 the environment task that executes the main program. On
25782 OpenVMS Alpha systems, @code{%TASK 0} is often used for tasks
25783 that have been created but are not yet activated.
25785 On OpenVMS Alpha systems, task IDs are assigned at
25786 activation. On GNAT systems, task IDs are also assigned at
25787 task creation but do not have the same form or values as
25788 task ID values in HP Ada. There is no null task, and the
25789 environment task does not have a specific task ID value.
25791 @node Task IDs and Delays
25792 @subsection Task IDs and Delays
25795 On OpenVMS Alpha systems, tasking delays are implemented
25796 using Timer System Services. The Task ID is used for the
25797 identification of the timer request (the @code{REQIDT} parameter).
25798 If Timers are used in the application take care not to use
25799 @code{0} for the identification, because cancelling such a timer
25800 will cancel all timers and may lead to unpredictable results.
25802 @node Task-Related Pragmas
25803 @subsection Task-Related Pragmas
25806 Ada supplies the pragma @code{TASK_STORAGE}, which allows
25807 specification of the size of the guard area for a task
25808 stack. (The guard area forms an area of memory that has no
25809 read or write access and thus helps in the detection of
25810 stack overflow.) On OpenVMS Alpha systems, if the pragma
25811 @code{TASK_STORAGE} specifies a value of zero, a minimal guard
25812 area is created. In the absence of a pragma @code{TASK_STORAGE},
25813 a default guard area is created.
25815 GNAT supplies the following task-related pragmas:
25818 @item @code{TASK_INFO}
25820 This pragma appears within a task definition and
25821 applies to the task in which it appears. The argument
25822 must be of type @code{SYSTEM.TASK_INFO.TASK_INFO_TYPE}.
25824 @item @code{TASK_STORAGE}
25826 GNAT implements pragma @code{TASK_STORAGE} in the same way as HP Ada.
25827 Both HP Ada and GNAT supply the pragmas @code{PASSIVE},
25828 @code{SUPPRESS}, and @code{VOLATILE}.
25830 @node Scheduling and Task Priority
25831 @subsection Scheduling and Task Priority
25834 HP Ada implements the Ada language requirement that
25835 when two tasks are eligible for execution and they have
25836 different priorities, the lower priority task does not
25837 execute while the higher priority task is waiting. The HP
25838 Ada Run-Time Library keeps a task running until either the
25839 task is suspended or a higher priority task becomes ready.
25841 On OpenVMS Alpha systems, the default strategy is round-
25842 robin with preemption. Tasks of equal priority take turns
25843 at the processor. A task is run for a certain period of
25844 time and then placed at the tail of the ready queue for
25845 its priority level.
25847 HP Ada provides the implementation-defined pragma @code{TIME_SLICE},
25848 which can be used to enable or disable round-robin
25849 scheduling of tasks with the same priority.
25850 See the relevant HP Ada run-time reference manual for
25851 information on using the pragmas to control HP Ada task
25854 GNAT follows the scheduling rules of Annex D (Real-Time
25855 Annex) of the @cite{Ada Reference Manual}. In general, this
25856 scheduling strategy is fully compatible with HP Ada
25857 although it provides some additional constraints (as
25858 fully documented in Annex D).
25859 GNAT implements time slicing control in a manner compatible with
25860 HP Ada 83, by means of the pragma @code{Time_Slice}, whose semantics
25861 are identical to the HP Ada 83 pragma of the same name.
25862 Note that it is not possible to mix GNAT tasking and
25863 HP Ada 83 tasking in the same program, since the two run-time
25864 libraries are not compatible.
25866 @node The Task Stack
25867 @subsection The Task Stack
25870 In HP Ada, a task stack is allocated each time a
25871 non-passive task is activated. As soon as the task is
25872 terminated, the storage for the task stack is deallocated.
25873 If you specify a size of zero (bytes) with @code{T'STORAGE_SIZE},
25874 a default stack size is used. Also, regardless of the size
25875 specified, some additional space is allocated for task
25876 management purposes. On OpenVMS Alpha systems, at least
25877 one page is allocated.
25879 GNAT handles task stacks in a similar manner. In accordance with
25880 the Ada rules, it provides the pragma @code{STORAGE_SIZE} as
25881 an alternative method for controlling the task stack size.
25882 The specification of the attribute @code{T'STORAGE_SIZE} is also
25883 supported in a manner compatible with HP Ada.
25885 @node External Interrupts
25886 @subsection External Interrupts
25889 On HP Ada, external interrupts can be associated with task entries.
25890 GNAT is compatible with HP Ada in its handling of external interrupts.
25892 @node Pragmas and Pragma-Related Features
25893 @section Pragmas and Pragma-Related Features
25896 Both HP Ada and GNAT supply all language-defined pragmas
25897 as specified by the Ada 83 standard. GNAT also supplies all
25898 language-defined pragmas introduced by Ada 95 and Ada 2005.
25899 In addition, GNAT implements the implementation-defined pragmas
25903 @item @code{AST_ENTRY}
25905 @item @code{COMMON_OBJECT}
25907 @item @code{COMPONENT_ALIGNMENT}
25909 @item @code{EXPORT_EXCEPTION}
25911 @item @code{EXPORT_FUNCTION}
25913 @item @code{EXPORT_OBJECT}
25915 @item @code{EXPORT_PROCEDURE}
25917 @item @code{EXPORT_VALUED_PROCEDURE}
25919 @item @code{FLOAT_REPRESENTATION}
25923 @item @code{IMPORT_EXCEPTION}
25925 @item @code{IMPORT_FUNCTION}
25927 @item @code{IMPORT_OBJECT}
25929 @item @code{IMPORT_PROCEDURE}
25931 @item @code{IMPORT_VALUED_PROCEDURE}
25933 @item @code{INLINE_GENERIC}
25935 @item @code{INTERFACE_NAME}
25937 @item @code{LONG_FLOAT}
25939 @item @code{MAIN_STORAGE}
25941 @item @code{PASSIVE}
25943 @item @code{PSECT_OBJECT}
25945 @item @code{SHARE_GENERIC}
25947 @item @code{SUPPRESS_ALL}
25949 @item @code{TASK_STORAGE}
25951 @item @code{TIME_SLICE}
25957 These pragmas are all fully implemented, with the exception of @code{TITLE},
25958 @code{PASSIVE}, and @code{SHARE_GENERIC}, which are
25959 recognized, but which have no
25960 effect in GNAT. The effect of @code{PASSIVE} may be obtained by the
25961 use of Ada protected objects. In GNAT, all generics are inlined.
25963 Unlike HP Ada, the GNAT ``@code{EXPORT_}@i{subprogram}'' pragmas require
25964 a separate subprogram specification which must appear before the
25967 GNAT also supplies a number of implementation-defined pragmas as follows:
25969 @item @code{ABORT_DEFER}
25971 @item @code{ADA_83}
25973 @item @code{ADA_95}
25975 @item @code{ADA_05}
25977 @item @code{ANNOTATE}
25979 @item @code{ASSERT}
25981 @item @code{C_PASS_BY_COPY}
25983 @item @code{CPP_CLASS}
25985 @item @code{CPP_CONSTRUCTOR}
25987 @item @code{CPP_DESTRUCTOR}
25991 @item @code{EXTEND_SYSTEM}
25993 @item @code{LINKER_ALIAS}
25995 @item @code{LINKER_SECTION}
25997 @item @code{MACHINE_ATTRIBUTE}
25999 @item @code{NO_RETURN}
26001 @item @code{PURE_FUNCTION}
26003 @item @code{SOURCE_FILE_NAME}
26005 @item @code{SOURCE_REFERENCE}
26007 @item @code{TASK_INFO}
26009 @item @code{UNCHECKED_UNION}
26011 @item @code{UNIMPLEMENTED_UNIT}
26013 @item @code{UNIVERSAL_DATA}
26015 @item @code{UNSUPPRESS}
26017 @item @code{WARNINGS}
26019 @item @code{WEAK_EXTERNAL}
26023 For full details on these GNAT implementation-defined pragmas,
26024 see @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT Reference
26028 * Restrictions on the Pragma INLINE::
26029 * Restrictions on the Pragma INTERFACE::
26030 * Restrictions on the Pragma SYSTEM_NAME::
26033 @node Restrictions on the Pragma INLINE
26034 @subsection Restrictions on Pragma @code{INLINE}
26037 HP Ada enforces the following restrictions on the pragma @code{INLINE}:
26039 @item Parameters cannot have a task type.
26041 @item Function results cannot be task types, unconstrained
26042 array types, or unconstrained types with discriminants.
26044 @item Bodies cannot declare the following:
26046 @item Subprogram body or stub (imported subprogram is allowed)
26050 @item Generic declarations
26052 @item Instantiations
26056 @item Access types (types derived from access types allowed)
26058 @item Array or record types
26060 @item Dependent tasks
26062 @item Direct recursive calls of subprogram or containing
26063 subprogram, directly or via a renaming
26069 In GNAT, the only restriction on pragma @code{INLINE} is that the
26070 body must occur before the call if both are in the same
26071 unit, and the size must be appropriately small. There are
26072 no other specific restrictions which cause subprograms to
26073 be incapable of being inlined.
26075 @node Restrictions on the Pragma INTERFACE
26076 @subsection Restrictions on Pragma @code{INTERFACE}
26079 The following restrictions on pragma @code{INTERFACE}
26080 are enforced by both HP Ada and GNAT:
26082 @item Languages accepted: Ada, Bliss, C, Fortran, Default.
26083 Default is the default on OpenVMS Alpha systems.
26085 @item Parameter passing: Language specifies default
26086 mechanisms but can be overridden with an @code{EXPORT} pragma.
26089 @item Ada: Use internal Ada rules.
26091 @item Bliss, C: Parameters must be mode @code{in}; cannot be
26092 record or task type. Result cannot be a string, an
26093 array, or a record.
26095 @item Fortran: Parameters cannot have a task type. Result cannot
26096 be a string, an array, or a record.
26101 GNAT is entirely upwards compatible with HP Ada, and in addition allows
26102 record parameters for all languages.
26104 @node Restrictions on the Pragma SYSTEM_NAME
26105 @subsection Restrictions on Pragma @code{SYSTEM_NAME}
26108 For HP Ada for OpenVMS Alpha, the enumeration literal
26109 for the type @code{NAME} is @code{OPENVMS_AXP}.
26110 In GNAT, the enumeration
26111 literal for the type @code{NAME} is @code{SYSTEM_NAME_GNAT}.
26113 @node Library of Predefined Units
26114 @section Library of Predefined Units
26117 A library of predefined units is provided as part of the
26118 HP Ada and GNAT implementations. HP Ada does not provide
26119 the package @code{MACHINE_CODE} but instead recommends importing
26122 The GNAT versions of the HP Ada Run-Time Library (@code{ADA$PREDEFINED:})
26123 units are taken from the OpenVMS Alpha version, not the OpenVMS VAX
26125 The HP Ada Predefined Library units are modified to remove post-Ada 83
26126 incompatibilities and to make them interoperable with GNAT
26127 (@pxref{Changes to DECLIB}, for details).
26128 The units are located in the @file{DECLIB} directory.
26130 The GNAT RTL is contained in
26131 the @file{ADALIB} directory, and
26132 the default search path is set up to find @code{DECLIB} units in preference
26133 to @code{ADALIB} units with the same name (@code{TEXT_IO},
26134 @code{SEQUENTIAL_IO}, and @code{DIRECT_IO}, for example).
26137 * Changes to DECLIB::
26140 @node Changes to DECLIB
26141 @subsection Changes to @code{DECLIB}
26144 The changes made to the HP Ada predefined library for GNAT and post-Ada 83
26145 compatibility are minor and include the following:
26148 @item Adjusting the location of pragmas and record representation
26149 clauses to obey Ada 95 (and thus Ada 2005) rules
26151 @item Adding the proper notation to generic formal parameters
26152 that take unconstrained types in instantiation
26154 @item Adding pragma @code{ELABORATE_BODY} to package specs
26155 that have package bodies not otherwise allowed
26157 @item Replacing occurrences of the identifier ``@code{PROTECTED}'' by
26158 ``@code{PROTECTD}''.
26159 Currently these are found only in the @code{STARLET} package spec.
26161 @item Changing @code{SYSTEM.ADDRESS} to @code{SYSTEM.SHORT_ADDRESS}
26162 where the address size is constrained to 32 bits.
26166 None of the above changes is visible to users.
26172 On OpenVMS Alpha, HP Ada provides the following strongly-typed bindings:
26175 @item Command Language Interpreter (CLI interface)
26177 @item DECtalk Run-Time Library (DTK interface)
26179 @item Librarian utility routines (LBR interface)
26181 @item General Purpose Run-Time Library (LIB interface)
26183 @item Math Run-Time Library (MTH interface)
26185 @item National Character Set Run-Time Library (NCS interface)
26187 @item Compiled Code Support Run-Time Library (OTS interface)
26189 @item Parallel Processing Run-Time Library (PPL interface)
26191 @item Screen Management Run-Time Library (SMG interface)
26193 @item Sort Run-Time Library (SOR interface)
26195 @item String Run-Time Library (STR interface)
26197 @item STARLET System Library
26200 @item X Window System Version 11R4 and 11R5 (X, XLIB interface)
26202 @item X Windows Toolkit (XT interface)
26204 @item X/Motif Version 1.1.3 and 1.2 (XM interface)
26208 GNAT provides implementations of these HP bindings in the @code{DECLIB}
26209 directory, on both the Alpha and I64 OpenVMS platforms.
26211 The X/Motif bindings used to build @code{DECLIB} are whatever versions are
26213 HP Ada @file{ADA$PREDEFINED} directory with extension @file{.ADC}.
26214 A pragma @code{Linker_Options} has been added to packages @code{Xm},
26215 @code{Xt}, and @code{X_Lib}
26216 causing the default X/Motif sharable image libraries to be linked in. This
26217 is done via options files named @file{xm.opt}, @file{xt.opt}, and
26218 @file{x_lib.opt} (also located in the @file{DECLIB} directory).
26220 It may be necessary to edit these options files to update or correct the
26221 library names if, for example, the newer X/Motif bindings from
26222 @file{ADA$EXAMPLES}
26223 had been (previous to installing GNAT) copied and renamed to supersede the
26224 default @file{ADA$PREDEFINED} versions.
26227 * Shared Libraries and Options Files::
26228 * Interfaces to C::
26231 @node Shared Libraries and Options Files
26232 @subsection Shared Libraries and Options Files
26235 When using the HP Ada
26236 predefined X and Motif bindings, the linking with their sharable images is
26237 done automatically by @command{GNAT LINK}.
26238 When using other X and Motif bindings, you need
26239 to add the corresponding sharable images to the command line for
26240 @code{GNAT LINK}. When linking with shared libraries, or with
26241 @file{.OPT} files, you must
26242 also add them to the command line for @command{GNAT LINK}.
26244 A shared library to be used with GNAT is built in the same way as other
26245 libraries under VMS. The VMS Link command can be used in standard fashion.
26247 @node Interfaces to C
26248 @subsection Interfaces to C
26252 provides the following Ada types and operations:
26255 @item C types package (@code{C_TYPES})
26257 @item C strings (@code{C_TYPES.NULL_TERMINATED})
26259 @item Other_types (@code{SHORT_INT})
26263 Interfacing to C with GNAT, you can use the above approach
26264 described for HP Ada or the facilities of Annex B of
26265 the @cite{Ada Reference Manual} (packages @code{INTERFACES.C},
26266 @code{INTERFACES.C.STRINGS} and @code{INTERFACES.C.POINTERS}). For more
26267 information, see @ref{Interfacing to C,,, gnat_rm, GNAT Reference Manual}.
26269 The @option{-gnatF} qualifier forces default and explicit
26270 @code{External_Name} parameters in pragmas @code{Import} and @code{Export}
26271 to be uppercased for compatibility with the default behavior
26272 of HP C. The qualifier has no effect on @code{Link_Name} parameters.
26274 @node Main Program Definition
26275 @section Main Program Definition
26278 The following section discusses differences in the
26279 definition of main programs on HP Ada and GNAT.
26280 On HP Ada, main programs are defined to meet the
26281 following conditions:
26283 @item Procedure with no formal parameters (returns @code{0} upon
26286 @item Procedure with no formal parameters (returns @code{42} when
26287 an unhandled exception is raised)
26289 @item Function with no formal parameters whose returned value
26290 is of a discrete type
26292 @item Procedure with one @code{out} formal of a discrete type for
26293 which a specification of pragma @code{EXPORT_VALUED_PROCEDURE} is given.
26298 When declared with the pragma @code{EXPORT_VALUED_PROCEDURE},
26299 a main function or main procedure returns a discrete
26300 value whose size is less than 64 bits (32 on VAX systems),
26301 the value is zero- or sign-extended as appropriate.
26302 On GNAT, main programs are defined as follows:
26304 @item Must be a non-generic, parameterless subprogram that
26305 is either a procedure or function returning an Ada
26306 @code{STANDARD.INTEGER} (the predefined type)
26308 @item Cannot be a generic subprogram or an instantiation of a
26312 @node Implementation-Defined Attributes
26313 @section Implementation-Defined Attributes
26316 GNAT provides all HP Ada implementation-defined
26319 @node Compiler and Run-Time Interfacing
26320 @section Compiler and Run-Time Interfacing
26323 HP Ada provides the following qualifiers to pass options to the linker
26326 @item @option{/WAIT} and @option{/SUBMIT}
26328 @item @option{/COMMAND}
26330 @item @option{/@r{[}NO@r{]}MAP}
26332 @item @option{/OUTPUT=@var{file-spec}}
26334 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
26338 To pass options to the linker, GNAT provides the following
26342 @item @option{/EXECUTABLE=@var{exec-name}}
26344 @item @option{/VERBOSE}
26346 @item @option{/@r{[}NO@r{]}DEBUG} and @option{/@r{[}NO@r{]}TRACEBACK}
26350 For more information on these switches, see
26351 @ref{Switches for gnatlink}.
26352 In HP Ada, the command-line switch @option{/OPTIMIZE} is available
26353 to control optimization. HP Ada also supplies the
26356 @item @code{OPTIMIZE}
26358 @item @code{INLINE}
26360 @item @code{INLINE_GENERIC}
26362 @item @code{SUPPRESS_ALL}
26364 @item @code{PASSIVE}
26368 In GNAT, optimization is controlled strictly by command
26369 line parameters, as described in the corresponding section of this guide.
26370 The HP pragmas for control of optimization are
26371 recognized but ignored.
26373 Note that in GNAT, the default is optimization off, whereas in HP Ada
26374 the default is that optimization is turned on.
26376 @node Program Compilation and Library Management
26377 @section Program Compilation and Library Management
26380 HP Ada and GNAT provide a comparable set of commands to
26381 build programs. HP Ada also provides a program library,
26382 which is a concept that does not exist on GNAT. Instead,
26383 GNAT provides directories of sources that are compiled as
26386 The following table summarizes
26387 the HP Ada commands and provides
26388 equivalent GNAT commands. In this table, some GNAT
26389 equivalents reflect the fact that GNAT does not use the
26390 concept of a program library. Instead, it uses a model
26391 in which collections of source and object files are used
26392 in a manner consistent with other languages like C and
26393 Fortran. Therefore, standard system file commands are used
26394 to manipulate these elements. Those GNAT commands are marked with
26396 Note that, unlike HP Ada, none of the GNAT commands accepts wild cards.
26399 @multitable @columnfractions .35 .65
26401 @item @emph{HP Ada Command}
26402 @tab @emph{GNAT Equivalent / Description}
26404 @item @command{ADA}
26405 @tab @command{GNAT COMPILE}@*
26406 Invokes the compiler to compile one or more Ada source files.
26408 @item @command{ACS ATTACH}@*
26409 @tab [No equivalent]@*
26410 Switches control of terminal from current process running the program
26413 @item @command{ACS CHECK}
26414 @tab @command{GNAT MAKE /DEPENDENCY_LIST}@*
26415 Forms the execution closure of one
26416 or more compiled units and checks completeness and currency.
26418 @item @command{ACS COMPILE}
26419 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
26420 Forms the execution closure of one or
26421 more specified units, checks completeness and currency,
26422 identifies units that have revised source files, compiles same,
26423 and recompiles units that are or will become obsolete.
26424 Also completes incomplete generic instantiations.
26426 @item @command{ACS COPY FOREIGN}
26428 Copies a foreign object file into the program library as a
26431 @item @command{ACS COPY UNIT}
26433 Copies a compiled unit from one program library to another.
26435 @item @command{ACS CREATE LIBRARY}
26436 @tab Create /directory (*)@*
26437 Creates a program library.
26439 @item @command{ACS CREATE SUBLIBRARY}
26440 @tab Create /directory (*)@*
26441 Creates a program sublibrary.
26443 @item @command{ACS DELETE LIBRARY}
26445 Deletes a program library and its contents.
26447 @item @command{ACS DELETE SUBLIBRARY}
26449 Deletes a program sublibrary and its contents.
26451 @item @command{ACS DELETE UNIT}
26452 @tab Delete file (*)@*
26453 On OpenVMS systems, deletes one or more compiled units from
26454 the current program library.
26456 @item @command{ACS DIRECTORY}
26457 @tab Directory (*)@*
26458 On OpenVMS systems, lists units contained in the current
26461 @item @command{ACS ENTER FOREIGN}
26463 Allows the import of a foreign body as an Ada library
26464 spec and enters a reference to a pointer.
26466 @item @command{ACS ENTER UNIT}
26468 Enters a reference (pointer) from the current program library to
26469 a unit compiled into another program library.
26471 @item @command{ACS EXIT}
26472 @tab [No equivalent]@*
26473 Exits from the program library manager.
26475 @item @command{ACS EXPORT}
26477 Creates an object file that contains system-specific object code
26478 for one or more units. With GNAT, object files can simply be copied
26479 into the desired directory.
26481 @item @command{ACS EXTRACT SOURCE}
26483 Allows access to the copied source file for each Ada compilation unit
26485 @item @command{ACS HELP}
26486 @tab @command{HELP GNAT}@*
26487 Provides online help.
26489 @item @command{ACS LINK}
26490 @tab @command{GNAT LINK}@*
26491 Links an object file containing Ada units into an executable file.
26493 @item @command{ACS LOAD}
26495 Loads (partially compiles) Ada units into the program library.
26496 Allows loading a program from a collection of files into a library
26497 without knowing the relationship among units.
26499 @item @command{ACS MERGE}
26501 Merges into the current program library, one or more units from
26502 another library where they were modified.
26504 @item @command{ACS RECOMPILE}
26505 @tab @command{GNAT MAKE /ACTIONS=COMPILE}@*
26506 Recompiles from external or copied source files any obsolete
26507 unit in the closure. Also, completes any incomplete generic
26510 @item @command{ACS REENTER}
26511 @tab @command{GNAT MAKE}@*
26512 Reenters current references to units compiled after last entered
26513 with the @command{ACS ENTER UNIT} command.
26515 @item @command{ACS SET LIBRARY}
26516 @tab Set default (*)@*
26517 Defines a program library to be the compilation context as well
26518 as the target library for compiler output and commands in general.
26520 @item @command{ACS SET PRAGMA}
26521 @tab Edit @file{gnat.adc} (*)@*
26522 Redefines specified values of the library characteristics
26523 @code{LONG_ FLOAT}, @code{MEMORY_SIZE}, @code{SYSTEM_NAME},
26524 and @code{Float_Representation}.
26526 @item @command{ACS SET SOURCE}
26527 @tab Define @code{ADA_INCLUDE_PATH} path (*)@*
26528 Defines the source file search list for the @command{ACS COMPILE} command.
26530 @item @command{ACS SHOW LIBRARY}
26531 @tab Directory (*)@*
26532 Lists information about one or more program libraries.
26534 @item @command{ACS SHOW PROGRAM}
26535 @tab [No equivalent]@*
26536 Lists information about the execution closure of one or
26537 more units in the program library.
26539 @item @command{ACS SHOW SOURCE}
26540 @tab Show logical @code{ADA_INCLUDE_PATH}@*
26541 Shows the source file search used when compiling units.
26543 @item @command{ACS SHOW VERSION}
26544 @tab Compile with @option{VERBOSE} option
26545 Displays the version number of the compiler and program library
26548 @item @command{ACS SPAWN}
26549 @tab [No equivalent]@*
26550 Creates a subprocess of the current process (same as @command{DCL SPAWN}
26553 @item @command{ACS VERIFY}
26554 @tab [No equivalent]@*
26555 Performs a series of consistency checks on a program library to
26556 determine whether the library structure and library files are in
26563 @section Input-Output
26566 On OpenVMS Alpha systems, HP Ada uses OpenVMS Record
26567 Management Services (RMS) to perform operations on
26571 HP Ada and GNAT predefine an identical set of input-
26572 output packages. To make the use of the
26573 generic @code{TEXT_IO} operations more convenient, HP Ada
26574 provides predefined library packages that instantiate the
26575 integer and floating-point operations for the predefined
26576 integer and floating-point types as shown in the following table.
26578 @multitable @columnfractions .45 .55
26579 @item @emph{Package Name} @tab Instantiation
26581 @item @code{INTEGER_TEXT_IO}
26582 @tab @code{INTEGER_IO(INTEGER)}
26584 @item @code{SHORT_INTEGER_TEXT_IO}
26585 @tab @code{INTEGER_IO(SHORT_INTEGER)}
26587 @item @code{SHORT_SHORT_INTEGER_TEXT_IO}
26588 @tab @code{INTEGER_IO(SHORT_SHORT_INTEGER)}
26590 @item @code{FLOAT_TEXT_IO}
26591 @tab @code{FLOAT_IO(FLOAT)}
26593 @item @code{LONG_FLOAT_TEXT_IO}
26594 @tab @code{FLOAT_IO(LONG_FLOAT)}
26598 The HP Ada predefined packages and their operations
26599 are implemented using OpenVMS Alpha files and input-output
26600 facilities. HP Ada supports asynchronous input-output on OpenVMS Alpha.
26601 Familiarity with the following is recommended:
26603 @item RMS file organizations and access methods
26605 @item OpenVMS file specifications and directories
26607 @item OpenVMS File Definition Language (FDL)
26611 GNAT provides I/O facilities that are completely
26612 compatible with HP Ada. The distribution includes the
26613 standard HP Ada versions of all I/O packages, operating
26614 in a manner compatible with HP Ada. In particular, the
26615 following packages are by default the HP Ada (Ada 83)
26616 versions of these packages rather than the renamings
26617 suggested in Annex J of the Ada Reference Manual:
26619 @item @code{TEXT_IO}
26621 @item @code{SEQUENTIAL_IO}
26623 @item @code{DIRECT_IO}
26627 The use of the standard child package syntax (for
26628 example, @code{ADA.TEXT_IO}) retrieves the post-Ada 83 versions of these
26630 GNAT provides HP-compatible predefined instantiations
26631 of the @code{TEXT_IO} packages, and also
26632 provides the standard predefined instantiations required
26633 by the @cite{Ada Reference Manual}.
26635 For further information on how GNAT interfaces to the file
26636 system or how I/O is implemented in programs written in
26637 mixed languages, see @ref{Implementation of the Standard I/O,,,
26638 gnat_rm, GNAT Reference Manual}.
26639 This chapter covers the following:
26641 @item Standard I/O packages
26643 @item @code{FORM} strings
26645 @item @code{ADA.DIRECT_IO}
26647 @item @code{ADA.SEQUENTIAL_IO}
26649 @item @code{ADA.TEXT_IO}
26651 @item Stream pointer positioning
26653 @item Reading and writing non-regular files
26655 @item @code{GET_IMMEDIATE}
26657 @item Treating @code{TEXT_IO} files as streams
26664 @node Implementation Limits
26665 @section Implementation Limits
26668 The following table lists implementation limits for HP Ada
26670 @multitable @columnfractions .60 .20 .20
26672 @item @emph{Compilation Parameter}
26677 @item In a subprogram or entry declaration, maximum number of
26678 formal parameters that are of an unconstrained record type
26683 @item Maximum identifier length (number of characters)
26688 @item Maximum number of characters in a source line
26693 @item Maximum collection size (number of bytes)
26698 @item Maximum number of discriminants for a record type
26703 @item Maximum number of formal parameters in an entry or
26704 subprogram declaration
26709 @item Maximum number of dimensions in an array type
26714 @item Maximum number of library units and subunits in a compilation.
26719 @item Maximum number of library units and subunits in an execution.
26724 @item Maximum number of objects declared with the pragma @code{COMMON_OBJECT}
26725 or @code{PSECT_OBJECT}
26730 @item Maximum number of enumeration literals in an enumeration type
26736 @item Maximum number of lines in a source file
26741 @item Maximum number of bits in any object
26746 @item Maximum size of the static portion of a stack frame (approximate)
26751 @node Tools and Utilities
26752 @section Tools and Utilities
26755 The following table lists some of the OpenVMS development tools
26756 available for HP Ada, and the corresponding tools for
26757 use with @value{EDITION} on Alpha and I64 platforms.
26758 Aside from the debugger, all the OpenVMS tools identified are part
26759 of the DECset package.
26762 @c Specify table in TeX since Texinfo does a poor job
26766 \settabs\+Language-Sensitive Editor\quad
26767 &Product with HP Ada\quad
26770 &\it Product with HP Ada
26771 & \it Product with GNAT Pro\cr
26773 \+Code Management System
26777 \+Language-Sensitive Editor
26779 & emacs or HP LSE (Alpha)\cr
26789 & OpenVMS Debug (I64)\cr
26791 \+Source Code Analyzer /
26808 \+Coverage Analyzer
26812 \+Module Management
26814 & Not applicable\cr
26824 @c This is the Texinfo version of the table. It renders poorly in pdf, hence
26825 @c the TeX version above for the printed version
26827 @c @multitable @columnfractions .3 .4 .4
26828 @multitable {Source Code Analyzer /}{Tool with HP Ada}{Tool with GNAT Pro}
26830 @tab @i{Tool with HP Ada}
26831 @tab @i{Tool with @value{EDITION}}
26832 @item Code Management@*System
26835 @item Language-Sensitive@*Editor
26837 @tab emacs or HP LSE (Alpha)
26846 @tab OpenVMS Debug (I64)
26847 @item Source Code Analyzer /@*Cross Referencer
26851 @tab HP Digital Test@*Manager (DTM)
26853 @item Performance and@*Coverage Analyzer
26856 @item Module Management@*System
26858 @tab Not applicable
26865 @c **************************************
26866 @node Platform-Specific Information for the Run-Time Libraries
26867 @appendix Platform-Specific Information for the Run-Time Libraries
26868 @cindex Tasking and threads libraries
26869 @cindex Threads libraries and tasking
26870 @cindex Run-time libraries (platform-specific information)
26873 The GNAT run-time implementation may vary with respect to both the
26874 underlying threads library and the exception handling scheme.
26875 For threads support, one or more of the following are supplied:
26877 @item @b{native threads library}, a binding to the thread package from
26878 the underlying operating system
26880 @item @b{pthreads library} (Sparc Solaris only), a binding to the Solaris
26881 POSIX thread package
26885 For exception handling, either or both of two models are supplied:
26887 @item @b{Zero-Cost Exceptions} (``ZCX''),@footnote{
26888 Most programs should experience a substantial speed improvement by
26889 being compiled with a ZCX run-time.
26890 This is especially true for
26891 tasking applications or applications with many exception handlers.}
26892 @cindex Zero-Cost Exceptions
26893 @cindex ZCX (Zero-Cost Exceptions)
26894 which uses binder-generated tables that
26895 are interrogated at run time to locate a handler
26897 @item @b{setjmp / longjmp} (``SJLJ''),
26898 @cindex setjmp/longjmp Exception Model
26899 @cindex SJLJ (setjmp/longjmp Exception Model)
26900 which uses dynamically-set data to establish
26901 the set of handlers
26905 This appendix summarizes which combinations of threads and exception support
26906 are supplied on various GNAT platforms.
26907 It then shows how to select a particular library either
26908 permanently or temporarily,
26909 explains the properties of (and tradeoffs among) the various threads
26910 libraries, and provides some additional
26911 information about several specific platforms.
26914 * Summary of Run-Time Configurations::
26915 * Specifying a Run-Time Library::
26916 * Choosing the Scheduling Policy::
26917 * Solaris-Specific Considerations::
26918 * Linux-Specific Considerations::
26919 * AIX-Specific Considerations::
26920 * Irix-Specific Considerations::
26921 * RTX-Specific Considerations::
26924 @node Summary of Run-Time Configurations
26925 @section Summary of Run-Time Configurations
26927 @multitable @columnfractions .30 .70
26928 @item @b{alpha-openvms}
26929 @item @code{@ @ }@i{rts-native (default)}
26930 @item @code{@ @ @ @ }Tasking @tab native VMS threads
26931 @item @code{@ @ @ @ }Exceptions @tab ZCX
26933 @item @b{alpha-tru64}
26934 @item @code{@ @ }@i{rts-native (default)}
26935 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
26936 @item @code{@ @ @ @ }Exceptions @tab ZCX
26938 @item @code{@ @ }@i{rts-sjlj}
26939 @item @code{@ @ @ @ }Tasking @tab native TRU64 threads
26940 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26942 @item @b{ia64-hp_linux}
26943 @item @code{@ @ }@i{rts-native (default)}
26944 @item @code{@ @ @ @ }Tasking @tab pthread library
26945 @item @code{@ @ @ @ }Exceptions @tab ZCX
26947 @item @b{ia64-hpux}
26948 @item @code{@ @ }@i{rts-native (default)}
26949 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26950 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26952 @item @b{ia64-openvms}
26953 @item @code{@ @ }@i{rts-native (default)}
26954 @item @code{@ @ @ @ }Tasking @tab native VMS threads
26955 @item @code{@ @ @ @ }Exceptions @tab ZCX
26957 @item @b{ia64-sgi_linux}
26958 @item @code{@ @ }@i{rts-native (default)}
26959 @item @code{@ @ @ @ }Tasking @tab pthread library
26960 @item @code{@ @ @ @ }Exceptions @tab ZCX
26962 @item @b{mips-irix}
26963 @item @code{@ @ }@i{rts-native (default)}
26964 @item @code{@ @ @ @ }Tasking @tab native IRIX threads
26965 @item @code{@ @ @ @ }Exceptions @tab ZCX
26968 @item @code{@ @ }@i{rts-native (default)}
26969 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26970 @item @code{@ @ @ @ }Exceptions @tab ZCX
26972 @item @code{@ @ }@i{rts-sjlj}
26973 @item @code{@ @ @ @ }Tasking @tab native HP-UX threads
26974 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26977 @item @code{@ @ }@i{rts-native (default)}
26978 @item @code{@ @ @ @ }Tasking @tab native AIX threads
26979 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26981 @item @b{ppc-darwin}
26982 @item @code{@ @ }@i{rts-native (default)}
26983 @item @code{@ @ @ @ }Tasking @tab native MacOS threads
26984 @item @code{@ @ @ @ }Exceptions @tab ZCX
26986 @item @b{sparc-solaris} @tab
26987 @item @code{@ @ }@i{rts-native (default)}
26988 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
26989 @item @code{@ @ @ @ }Exceptions @tab ZCX
26991 @item @code{@ @ }@i{rts-pthread}
26992 @item @code{@ @ @ @ }Tasking @tab pthread library
26993 @item @code{@ @ @ @ }Exceptions @tab ZCX
26995 @item @code{@ @ }@i{rts-sjlj}
26996 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
26997 @item @code{@ @ @ @ }Exceptions @tab SJLJ
26999 @item @b{sparc64-solaris} @tab
27000 @item @code{@ @ }@i{rts-native (default)}
27001 @item @code{@ @ @ @ }Tasking @tab native Solaris threads library
27002 @item @code{@ @ @ @ }Exceptions @tab ZCX
27004 @item @b{x86-linux}
27005 @item @code{@ @ }@i{rts-native (default)}
27006 @item @code{@ @ @ @ }Tasking @tab pthread library
27007 @item @code{@ @ @ @ }Exceptions @tab ZCX
27009 @item @code{@ @ }@i{rts-sjlj}
27010 @item @code{@ @ @ @ }Tasking @tab pthread library
27011 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27014 @item @code{@ @ }@i{rts-native (default)}
27015 @item @code{@ @ @ @ }Tasking @tab native LynxOS threads
27016 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27018 @item @b{x86-solaris}
27019 @item @code{@ @ }@i{rts-native (default)}
27020 @item @code{@ @ @ @ }Tasking @tab native Solaris threads
27021 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27023 @item @b{x86-windows}
27024 @item @code{@ @ }@i{rts-native (default)}
27025 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
27026 @item @code{@ @ @ @ }Exceptions @tab ZCX
27028 @item @code{@ @ }@i{rts-sjlj (default)}
27029 @item @code{@ @ @ @ }Tasking @tab native Win32 threads
27030 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27032 @item @b{x86-windows-rtx}
27033 @item @code{@ @ }@i{rts-rtx-rtss (default)}
27034 @item @code{@ @ @ @ }Tasking @tab RTX real-time subsystem RTSS threads (kernel mode)
27035 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27037 @item @code{@ @ }@i{rts-rtx-w32}
27038 @item @code{@ @ @ @ }Tasking @tab RTX Win32 threads (user mode)
27039 @item @code{@ @ @ @ }Exceptions @tab ZCX
27041 @item @b{x86_64-linux}
27042 @item @code{@ @ }@i{rts-native (default)}
27043 @item @code{@ @ @ @ }Tasking @tab pthread library
27044 @item @code{@ @ @ @ }Exceptions @tab ZCX
27046 @item @code{@ @ }@i{rts-sjlj}
27047 @item @code{@ @ @ @ }Tasking @tab pthread library
27048 @item @code{@ @ @ @ }Exceptions @tab SJLJ
27052 @node Specifying a Run-Time Library
27053 @section Specifying a Run-Time Library
27056 The @file{adainclude} subdirectory containing the sources of the GNAT
27057 run-time library, and the @file{adalib} subdirectory containing the
27058 @file{ALI} files and the static and/or shared GNAT library, are located
27059 in the gcc target-dependent area:
27062 target=$prefix/lib/gcc/gcc-@i{dumpmachine}/gcc-@i{dumpversion}/
27066 As indicated above, on some platforms several run-time libraries are supplied.
27067 These libraries are installed in the target dependent area and
27068 contain a complete source and binary subdirectory. The detailed description
27069 below explains the differences between the different libraries in terms of
27070 their thread support.
27072 The default run-time library (when GNAT is installed) is @emph{rts-native}.
27073 This default run time is selected by the means of soft links.
27074 For example on x86-linux:
27080 +--- adainclude----------+
27082 +--- adalib-----------+ |
27084 +--- rts-native | |
27086 | +--- adainclude <---+
27088 | +--- adalib <----+
27099 If the @i{rts-sjlj} library is to be selected on a permanent basis,
27100 these soft links can be modified with the following commands:
27104 $ rm -f adainclude adalib
27105 $ ln -s rts-sjlj/adainclude adainclude
27106 $ ln -s rts-sjlj/adalib adalib
27110 Alternatively, you can specify @file{rts-sjlj/adainclude} in the file
27111 @file{$target/ada_source_path} and @file{rts-sjlj/adalib} in
27112 @file{$target/ada_object_path}.
27114 Selecting another run-time library temporarily can be
27115 achieved by using the @option{--RTS} switch, e.g., @option{--RTS=sjlj}
27116 @cindex @option{--RTS} option
27118 @node Choosing the Scheduling Policy
27119 @section Choosing the Scheduling Policy
27122 When using a POSIX threads implementation, you have a choice of several
27123 scheduling policies: @code{SCHED_FIFO},
27124 @cindex @code{SCHED_FIFO} scheduling policy
27126 @cindex @code{SCHED_RR} scheduling policy
27127 and @code{SCHED_OTHER}.
27128 @cindex @code{SCHED_OTHER} scheduling policy
27129 Typically, the default is @code{SCHED_OTHER}, while using @code{SCHED_FIFO}
27130 or @code{SCHED_RR} requires special (e.g., root) privileges.
27132 By default, GNAT uses the @code{SCHED_OTHER} policy. To specify
27134 @cindex @code{SCHED_FIFO} scheduling policy
27135 you can use one of the following:
27139 @code{pragma Time_Slice (0.0)}
27140 @cindex pragma Time_Slice
27142 the corresponding binder option @option{-T0}
27143 @cindex @option{-T0} option
27145 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
27146 @cindex pragma Task_Dispatching_Policy
27150 To specify @code{SCHED_RR},
27151 @cindex @code{SCHED_RR} scheduling policy
27152 you should use @code{pragma Time_Slice} with a
27153 value greater than @code{0.0}, or else use the corresponding @option{-T}
27156 @node Solaris-Specific Considerations
27157 @section Solaris-Specific Considerations
27158 @cindex Solaris Sparc threads libraries
27161 This section addresses some topics related to the various threads libraries
27165 * Solaris Threads Issues::
27168 @node Solaris Threads Issues
27169 @subsection Solaris Threads Issues
27172 GNAT under Solaris/Sparc 32 bits comes with an alternate tasking run-time
27173 library based on POSIX threads --- @emph{rts-pthread}.
27174 @cindex rts-pthread threads library
27175 This run-time library has the advantage of being mostly shared across all
27176 POSIX-compliant thread implementations, and it also provides under
27177 @w{Solaris 8} the @code{PTHREAD_PRIO_INHERIT}
27178 @cindex @code{PTHREAD_PRIO_INHERIT} policy (under rts-pthread)
27179 and @code{PTHREAD_PRIO_PROTECT}
27180 @cindex @code{PTHREAD_PRIO_PROTECT} policy (under rts-pthread)
27181 semantics that can be selected using the predefined pragma
27182 @code{Locking_Policy}
27183 @cindex pragma Locking_Policy (under rts-pthread)
27185 @code{Inheritance_Locking} and @code{Ceiling_Locking} as the policy.
27186 @cindex @code{Inheritance_Locking} (under rts-pthread)
27187 @cindex @code{Ceiling_Locking} (under rts-pthread)
27189 As explained above, the native run-time library is based on the Solaris thread
27190 library (@code{libthread}) and is the default library.
27192 When the Solaris threads library is used (this is the default), programs
27193 compiled with GNAT can automatically take advantage of
27194 and can thus execute on multiple processors.
27195 The user can alternatively specify a processor on which the program should run
27196 to emulate a single-processor system. The multiprocessor / uniprocessor choice
27198 setting the environment variable @env{GNAT_PROCESSOR}
27199 @cindex @env{GNAT_PROCESSOR} environment variable (on Sparc Solaris)
27200 to one of the following:
27204 Use the default configuration (run the program on all
27205 available processors) - this is the same as having @code{GNAT_PROCESSOR}
27209 Let the run-time implementation choose one processor and run the program on
27212 @item 0 .. Last_Proc
27213 Run the program on the specified processor.
27214 @code{Last_Proc} is equal to @code{_SC_NPROCESSORS_CONF - 1}
27215 (where @code{_SC_NPROCESSORS_CONF} is a system variable).
27218 @node Linux-Specific Considerations
27219 @section Linux-Specific Considerations
27220 @cindex Linux threads libraries
27223 On GNU/Linux without NPTL support (usually system with GNU C Library
27224 older than 2.3), the signal model is not POSIX compliant, which means
27225 that to send a signal to the process, you need to send the signal to all
27226 threads, e.g.@: by using @code{killpg()}.
27228 @node AIX-Specific Considerations
27229 @section AIX-Specific Considerations
27230 @cindex AIX resolver library
27233 On AIX, the resolver library initializes some internal structure on
27234 the first call to @code{get*by*} functions, which are used to implement
27235 @code{GNAT.Sockets.Get_Host_By_Name} and
27236 @code{GNAT.Sockets.Get_Host_By_Address}.
27237 If such initialization occurs within an Ada task, and the stack size for
27238 the task is the default size, a stack overflow may occur.
27240 To avoid this overflow, the user should either ensure that the first call
27241 to @code{GNAT.Sockets.Get_Host_By_Name} or
27242 @code{GNAT.Sockets.Get_Host_By_Addrss}
27243 occurs in the environment task, or use @code{pragma Storage_Size} to
27244 specify a sufficiently large size for the stack of the task that contains
27247 @node Irix-Specific Considerations
27248 @section Irix-Specific Considerations
27249 @cindex Irix libraries
27252 The GCC support libraries coming with the Irix compiler have moved to
27253 their canonical place with respect to the general Irix ABI related
27254 conventions. Running applications built with the default shared GNAT
27255 run-time now requires the LD_LIBRARY_PATH environment variable to
27256 include this location. A possible way to achieve this is to issue the
27257 following command line on a bash prompt:
27261 $ LD_LIBRARY_PATH=$LD_LIBRARY_PATH:`dirname \`gcc --print-file-name=libgcc_s.so\``
27265 @node RTX-Specific Considerations
27266 @section RTX-Specific Considerations
27267 @cindex RTX libraries
27270 The Real-time Extension (RTX) to Windows is based on the Windows Win32
27271 API. Applications can be built to work in two different modes:
27275 Windows executables that run in Ring 3 to utilize memory protection
27276 (@emph{rts-rtx-w32}).
27279 Real-time subsystem (RTSS) executables that run in Ring 0, where
27280 performance can be optimized with RTSS applications taking precedent
27281 over all Windows applications (@emph{rts-rtx-rtss}).
27285 @c *******************************
27286 @node Example of Binder Output File
27287 @appendix Example of Binder Output File
27290 This Appendix displays the source code for @command{gnatbind}'s output
27291 file generated for a simple ``Hello World'' program.
27292 Comments have been added for clarification purposes.
27294 @smallexample @c adanocomment
27298 -- The package is called Ada_Main unless this name is actually used
27299 -- as a unit name in the partition, in which case some other unique
27303 package ada_main is
27305 Elab_Final_Code : Integer;
27306 pragma Import (C, Elab_Final_Code, "__gnat_inside_elab_final_code");
27308 -- The main program saves the parameters (argument count,
27309 -- argument values, environment pointer) in global variables
27310 -- for later access by other units including
27311 -- Ada.Command_Line.
27313 gnat_argc : Integer;
27314 gnat_argv : System.Address;
27315 gnat_envp : System.Address;
27317 -- The actual variables are stored in a library routine. This
27318 -- is useful for some shared library situations, where there
27319 -- are problems if variables are not in the library.
27321 pragma Import (C, gnat_argc);
27322 pragma Import (C, gnat_argv);
27323 pragma Import (C, gnat_envp);
27325 -- The exit status is similarly an external location
27327 gnat_exit_status : Integer;
27328 pragma Import (C, gnat_exit_status);
27330 GNAT_Version : constant String :=
27331 "GNAT Version: 6.0.0w (20061115)";
27332 pragma Export (C, GNAT_Version, "__gnat_version");
27334 -- This is the generated adafinal routine that performs
27335 -- finalization at the end of execution. In the case where
27336 -- Ada is the main program, this main program makes a call
27337 -- to adafinal at program termination.
27339 procedure adafinal;
27340 pragma Export (C, adafinal, "adafinal");
27342 -- This is the generated adainit routine that performs
27343 -- initialization at the start of execution. In the case
27344 -- where Ada is the main program, this main program makes
27345 -- a call to adainit at program startup.
27348 pragma Export (C, adainit, "adainit");
27350 -- This routine is called at the start of execution. It is
27351 -- a dummy routine that is used by the debugger to breakpoint
27352 -- at the start of execution.
27354 procedure Break_Start;
27355 pragma Import (C, Break_Start, "__gnat_break_start");
27357 -- This is the actual generated main program (it would be
27358 -- suppressed if the no main program switch were used). As
27359 -- required by standard system conventions, this program has
27360 -- the external name main.
27364 argv : System.Address;
27365 envp : System.Address)
27367 pragma Export (C, main, "main");
27369 -- The following set of constants give the version
27370 -- identification values for every unit in the bound
27371 -- partition. This identification is computed from all
27372 -- dependent semantic units, and corresponds to the
27373 -- string that would be returned by use of the
27374 -- Body_Version or Version attributes.
27376 type Version_32 is mod 2 ** 32;
27377 u00001 : constant Version_32 := 16#7880BEB3#;
27378 u00002 : constant Version_32 := 16#0D24CBD0#;
27379 u00003 : constant Version_32 := 16#3283DBEB#;
27380 u00004 : constant Version_32 := 16#2359F9ED#;
27381 u00005 : constant Version_32 := 16#664FB847#;
27382 u00006 : constant Version_32 := 16#68E803DF#;
27383 u00007 : constant Version_32 := 16#5572E604#;
27384 u00008 : constant Version_32 := 16#46B173D8#;
27385 u00009 : constant Version_32 := 16#156A40CF#;
27386 u00010 : constant Version_32 := 16#033DABE0#;
27387 u00011 : constant Version_32 := 16#6AB38FEA#;
27388 u00012 : constant Version_32 := 16#22B6217D#;
27389 u00013 : constant Version_32 := 16#68A22947#;
27390 u00014 : constant Version_32 := 16#18CC4A56#;
27391 u00015 : constant Version_32 := 16#08258E1B#;
27392 u00016 : constant Version_32 := 16#367D5222#;
27393 u00017 : constant Version_32 := 16#20C9ECA4#;
27394 u00018 : constant Version_32 := 16#50D32CB6#;
27395 u00019 : constant Version_32 := 16#39A8BB77#;
27396 u00020 : constant Version_32 := 16#5CF8FA2B#;
27397 u00021 : constant Version_32 := 16#2F1EB794#;
27398 u00022 : constant Version_32 := 16#31AB6444#;
27399 u00023 : constant Version_32 := 16#1574B6E9#;
27400 u00024 : constant Version_32 := 16#5109C189#;
27401 u00025 : constant Version_32 := 16#56D770CD#;
27402 u00026 : constant Version_32 := 16#02F9DE3D#;
27403 u00027 : constant Version_32 := 16#08AB6B2C#;
27404 u00028 : constant Version_32 := 16#3FA37670#;
27405 u00029 : constant Version_32 := 16#476457A0#;
27406 u00030 : constant Version_32 := 16#731E1B6E#;
27407 u00031 : constant Version_32 := 16#23C2E789#;
27408 u00032 : constant Version_32 := 16#0F1BD6A1#;
27409 u00033 : constant Version_32 := 16#7C25DE96#;
27410 u00034 : constant Version_32 := 16#39ADFFA2#;
27411 u00035 : constant Version_32 := 16#571DE3E7#;
27412 u00036 : constant Version_32 := 16#5EB646AB#;
27413 u00037 : constant Version_32 := 16#4249379B#;
27414 u00038 : constant Version_32 := 16#0357E00A#;
27415 u00039 : constant Version_32 := 16#3784FB72#;
27416 u00040 : constant Version_32 := 16#2E723019#;
27417 u00041 : constant Version_32 := 16#623358EA#;
27418 u00042 : constant Version_32 := 16#107F9465#;
27419 u00043 : constant Version_32 := 16#6843F68A#;
27420 u00044 : constant Version_32 := 16#63305874#;
27421 u00045 : constant Version_32 := 16#31E56CE1#;
27422 u00046 : constant Version_32 := 16#02917970#;
27423 u00047 : constant Version_32 := 16#6CCBA70E#;
27424 u00048 : constant Version_32 := 16#41CD4204#;
27425 u00049 : constant Version_32 := 16#572E3F58#;
27426 u00050 : constant Version_32 := 16#20729FF5#;
27427 u00051 : constant Version_32 := 16#1D4F93E8#;
27428 u00052 : constant Version_32 := 16#30B2EC3D#;
27429 u00053 : constant Version_32 := 16#34054F96#;
27430 u00054 : constant Version_32 := 16#5A199860#;
27431 u00055 : constant Version_32 := 16#0E7F912B#;
27432 u00056 : constant Version_32 := 16#5760634A#;
27433 u00057 : constant Version_32 := 16#5D851835#;
27435 -- The following Export pragmas export the version numbers
27436 -- with symbolic names ending in B (for body) or S
27437 -- (for spec) so that they can be located in a link. The
27438 -- information provided here is sufficient to track down
27439 -- the exact versions of units used in a given build.
27441 pragma Export (C, u00001, "helloB");
27442 pragma Export (C, u00002, "system__standard_libraryB");
27443 pragma Export (C, u00003, "system__standard_libraryS");
27444 pragma Export (C, u00004, "adaS");
27445 pragma Export (C, u00005, "ada__text_ioB");
27446 pragma Export (C, u00006, "ada__text_ioS");
27447 pragma Export (C, u00007, "ada__exceptionsB");
27448 pragma Export (C, u00008, "ada__exceptionsS");
27449 pragma Export (C, u00009, "gnatS");
27450 pragma Export (C, u00010, "gnat__heap_sort_aB");
27451 pragma Export (C, u00011, "gnat__heap_sort_aS");
27452 pragma Export (C, u00012, "systemS");
27453 pragma Export (C, u00013, "system__exception_tableB");
27454 pragma Export (C, u00014, "system__exception_tableS");
27455 pragma Export (C, u00015, "gnat__htableB");
27456 pragma Export (C, u00016, "gnat__htableS");
27457 pragma Export (C, u00017, "system__exceptionsS");
27458 pragma Export (C, u00018, "system__machine_state_operationsB");
27459 pragma Export (C, u00019, "system__machine_state_operationsS");
27460 pragma Export (C, u00020, "system__machine_codeS");
27461 pragma Export (C, u00021, "system__storage_elementsB");
27462 pragma Export (C, u00022, "system__storage_elementsS");
27463 pragma Export (C, u00023, "system__secondary_stackB");
27464 pragma Export (C, u00024, "system__secondary_stackS");
27465 pragma Export (C, u00025, "system__parametersB");
27466 pragma Export (C, u00026, "system__parametersS");
27467 pragma Export (C, u00027, "system__soft_linksB");
27468 pragma Export (C, u00028, "system__soft_linksS");
27469 pragma Export (C, u00029, "system__stack_checkingB");
27470 pragma Export (C, u00030, "system__stack_checkingS");
27471 pragma Export (C, u00031, "system__tracebackB");
27472 pragma Export (C, u00032, "system__tracebackS");
27473 pragma Export (C, u00033, "ada__streamsS");
27474 pragma Export (C, u00034, "ada__tagsB");
27475 pragma Export (C, u00035, "ada__tagsS");
27476 pragma Export (C, u00036, "system__string_opsB");
27477 pragma Export (C, u00037, "system__string_opsS");
27478 pragma Export (C, u00038, "interfacesS");
27479 pragma Export (C, u00039, "interfaces__c_streamsB");
27480 pragma Export (C, u00040, "interfaces__c_streamsS");
27481 pragma Export (C, u00041, "system__file_ioB");
27482 pragma Export (C, u00042, "system__file_ioS");
27483 pragma Export (C, u00043, "ada__finalizationB");
27484 pragma Export (C, u00044, "ada__finalizationS");
27485 pragma Export (C, u00045, "system__finalization_rootB");
27486 pragma Export (C, u00046, "system__finalization_rootS");
27487 pragma Export (C, u00047, "system__finalization_implementationB");
27488 pragma Export (C, u00048, "system__finalization_implementationS");
27489 pragma Export (C, u00049, "system__string_ops_concat_3B");
27490 pragma Export (C, u00050, "system__string_ops_concat_3S");
27491 pragma Export (C, u00051, "system__stream_attributesB");
27492 pragma Export (C, u00052, "system__stream_attributesS");
27493 pragma Export (C, u00053, "ada__io_exceptionsS");
27494 pragma Export (C, u00054, "system__unsigned_typesS");
27495 pragma Export (C, u00055, "system__file_control_blockS");
27496 pragma Export (C, u00056, "ada__finalization__list_controllerB");
27497 pragma Export (C, u00057, "ada__finalization__list_controllerS");
27499 -- BEGIN ELABORATION ORDER
27502 -- gnat.heap_sort_a (spec)
27503 -- gnat.heap_sort_a (body)
27504 -- gnat.htable (spec)
27505 -- gnat.htable (body)
27506 -- interfaces (spec)
27508 -- system.machine_code (spec)
27509 -- system.parameters (spec)
27510 -- system.parameters (body)
27511 -- interfaces.c_streams (spec)
27512 -- interfaces.c_streams (body)
27513 -- system.standard_library (spec)
27514 -- ada.exceptions (spec)
27515 -- system.exception_table (spec)
27516 -- system.exception_table (body)
27517 -- ada.io_exceptions (spec)
27518 -- system.exceptions (spec)
27519 -- system.storage_elements (spec)
27520 -- system.storage_elements (body)
27521 -- system.machine_state_operations (spec)
27522 -- system.machine_state_operations (body)
27523 -- system.secondary_stack (spec)
27524 -- system.stack_checking (spec)
27525 -- system.soft_links (spec)
27526 -- system.soft_links (body)
27527 -- system.stack_checking (body)
27528 -- system.secondary_stack (body)
27529 -- system.standard_library (body)
27530 -- system.string_ops (spec)
27531 -- system.string_ops (body)
27534 -- ada.streams (spec)
27535 -- system.finalization_root (spec)
27536 -- system.finalization_root (body)
27537 -- system.string_ops_concat_3 (spec)
27538 -- system.string_ops_concat_3 (body)
27539 -- system.traceback (spec)
27540 -- system.traceback (body)
27541 -- ada.exceptions (body)
27542 -- system.unsigned_types (spec)
27543 -- system.stream_attributes (spec)
27544 -- system.stream_attributes (body)
27545 -- system.finalization_implementation (spec)
27546 -- system.finalization_implementation (body)
27547 -- ada.finalization (spec)
27548 -- ada.finalization (body)
27549 -- ada.finalization.list_controller (spec)
27550 -- ada.finalization.list_controller (body)
27551 -- system.file_control_block (spec)
27552 -- system.file_io (spec)
27553 -- system.file_io (body)
27554 -- ada.text_io (spec)
27555 -- ada.text_io (body)
27557 -- END ELABORATION ORDER
27561 -- The following source file name pragmas allow the generated file
27562 -- names to be unique for different main programs. They are needed
27563 -- since the package name will always be Ada_Main.
27565 pragma Source_File_Name (ada_main, Spec_File_Name => "b~hello.ads");
27566 pragma Source_File_Name (ada_main, Body_File_Name => "b~hello.adb");
27568 -- Generated package body for Ada_Main starts here
27570 package body ada_main is
27572 -- The actual finalization is performed by calling the
27573 -- library routine in System.Standard_Library.Adafinal
27575 procedure Do_Finalize;
27576 pragma Import (C, Do_Finalize, "system__standard_library__adafinal");
27583 procedure adainit is
27585 -- These booleans are set to True once the associated unit has
27586 -- been elaborated. It is also used to avoid elaborating the
27587 -- same unit twice.
27590 pragma Import (Ada, E040, "interfaces__c_streams_E");
27593 pragma Import (Ada, E008, "ada__exceptions_E");
27596 pragma Import (Ada, E014, "system__exception_table_E");
27599 pragma Import (Ada, E053, "ada__io_exceptions_E");
27602 pragma Import (Ada, E017, "system__exceptions_E");
27605 pragma Import (Ada, E024, "system__secondary_stack_E");
27608 pragma Import (Ada, E030, "system__stack_checking_E");
27611 pragma Import (Ada, E028, "system__soft_links_E");
27614 pragma Import (Ada, E035, "ada__tags_E");
27617 pragma Import (Ada, E033, "ada__streams_E");
27620 pragma Import (Ada, E046, "system__finalization_root_E");
27623 pragma Import (Ada, E048, "system__finalization_implementation_E");
27626 pragma Import (Ada, E044, "ada__finalization_E");
27629 pragma Import (Ada, E057, "ada__finalization__list_controller_E");
27632 pragma Import (Ada, E055, "system__file_control_block_E");
27635 pragma Import (Ada, E042, "system__file_io_E");
27638 pragma Import (Ada, E006, "ada__text_io_E");
27640 -- Set_Globals is a library routine that stores away the
27641 -- value of the indicated set of global values in global
27642 -- variables within the library.
27644 procedure Set_Globals
27645 (Main_Priority : Integer;
27646 Time_Slice_Value : Integer;
27647 WC_Encoding : Character;
27648 Locking_Policy : Character;
27649 Queuing_Policy : Character;
27650 Task_Dispatching_Policy : Character;
27651 Adafinal : System.Address;
27652 Unreserve_All_Interrupts : Integer;
27653 Exception_Tracebacks : Integer);
27654 @findex __gnat_set_globals
27655 pragma Import (C, Set_Globals, "__gnat_set_globals");
27657 -- SDP_Table_Build is a library routine used to build the
27658 -- exception tables. See unit Ada.Exceptions in files
27659 -- a-except.ads/adb for full details of how zero cost
27660 -- exception handling works. This procedure, the call to
27661 -- it, and the two following tables are all omitted if the
27662 -- build is in longjmp/setjmp exception mode.
27664 @findex SDP_Table_Build
27665 @findex Zero Cost Exceptions
27666 procedure SDP_Table_Build
27667 (SDP_Addresses : System.Address;
27668 SDP_Count : Natural;
27669 Elab_Addresses : System.Address;
27670 Elab_Addr_Count : Natural);
27671 pragma Import (C, SDP_Table_Build, "__gnat_SDP_Table_Build");
27673 -- Table of Unit_Exception_Table addresses. Used for zero
27674 -- cost exception handling to build the top level table.
27676 ST : aliased constant array (1 .. 23) of System.Address := (
27678 Ada.Text_Io'UET_Address,
27679 Ada.Exceptions'UET_Address,
27680 Gnat.Heap_Sort_A'UET_Address,
27681 System.Exception_Table'UET_Address,
27682 System.Machine_State_Operations'UET_Address,
27683 System.Secondary_Stack'UET_Address,
27684 System.Parameters'UET_Address,
27685 System.Soft_Links'UET_Address,
27686 System.Stack_Checking'UET_Address,
27687 System.Traceback'UET_Address,
27688 Ada.Streams'UET_Address,
27689 Ada.Tags'UET_Address,
27690 System.String_Ops'UET_Address,
27691 Interfaces.C_Streams'UET_Address,
27692 System.File_Io'UET_Address,
27693 Ada.Finalization'UET_Address,
27694 System.Finalization_Root'UET_Address,
27695 System.Finalization_Implementation'UET_Address,
27696 System.String_Ops_Concat_3'UET_Address,
27697 System.Stream_Attributes'UET_Address,
27698 System.File_Control_Block'UET_Address,
27699 Ada.Finalization.List_Controller'UET_Address);
27701 -- Table of addresses of elaboration routines. Used for
27702 -- zero cost exception handling to make sure these
27703 -- addresses are included in the top level procedure
27706 EA : aliased constant array (1 .. 23) of System.Address := (
27707 adainit'Code_Address,
27708 Do_Finalize'Code_Address,
27709 Ada.Exceptions'Elab_Spec'Address,
27710 System.Exceptions'Elab_Spec'Address,
27711 Interfaces.C_Streams'Elab_Spec'Address,
27712 System.Exception_Table'Elab_Body'Address,
27713 Ada.Io_Exceptions'Elab_Spec'Address,
27714 System.Stack_Checking'Elab_Spec'Address,
27715 System.Soft_Links'Elab_Body'Address,
27716 System.Secondary_Stack'Elab_Body'Address,
27717 Ada.Tags'Elab_Spec'Address,
27718 Ada.Tags'Elab_Body'Address,
27719 Ada.Streams'Elab_Spec'Address,
27720 System.Finalization_Root'Elab_Spec'Address,
27721 Ada.Exceptions'Elab_Body'Address,
27722 System.Finalization_Implementation'Elab_Spec'Address,
27723 System.Finalization_Implementation'Elab_Body'Address,
27724 Ada.Finalization'Elab_Spec'Address,
27725 Ada.Finalization.List_Controller'Elab_Spec'Address,
27726 System.File_Control_Block'Elab_Spec'Address,
27727 System.File_Io'Elab_Body'Address,
27728 Ada.Text_Io'Elab_Spec'Address,
27729 Ada.Text_Io'Elab_Body'Address);
27731 -- Start of processing for adainit
27735 -- Call SDP_Table_Build to build the top level procedure
27736 -- table for zero cost exception handling (omitted in
27737 -- longjmp/setjmp mode).
27739 SDP_Table_Build (ST'Address, 23, EA'Address, 23);
27741 -- Call Set_Globals to record various information for
27742 -- this partition. The values are derived by the binder
27743 -- from information stored in the ali files by the compiler.
27745 @findex __gnat_set_globals
27747 (Main_Priority => -1,
27748 -- Priority of main program, -1 if no pragma Priority used
27750 Time_Slice_Value => -1,
27751 -- Time slice from Time_Slice pragma, -1 if none used
27753 WC_Encoding => 'b',
27754 -- Wide_Character encoding used, default is brackets
27756 Locking_Policy => ' ',
27757 -- Locking_Policy used, default of space means not
27758 -- specified, otherwise it is the first character of
27759 -- the policy name.
27761 Queuing_Policy => ' ',
27762 -- Queuing_Policy used, default of space means not
27763 -- specified, otherwise it is the first character of
27764 -- the policy name.
27766 Task_Dispatching_Policy => ' ',
27767 -- Task_Dispatching_Policy used, default of space means
27768 -- not specified, otherwise first character of the
27771 Adafinal => System.Null_Address,
27772 -- Address of Adafinal routine, not used anymore
27774 Unreserve_All_Interrupts => 0,
27775 -- Set true if pragma Unreserve_All_Interrupts was used
27777 Exception_Tracebacks => 0);
27778 -- Indicates if exception tracebacks are enabled
27780 Elab_Final_Code := 1;
27782 -- Now we have the elaboration calls for all units in the partition.
27783 -- The Elab_Spec and Elab_Body attributes generate references to the
27784 -- implicit elaboration procedures generated by the compiler for
27785 -- each unit that requires elaboration.
27788 Interfaces.C_Streams'Elab_Spec;
27792 Ada.Exceptions'Elab_Spec;
27795 System.Exception_Table'Elab_Body;
27799 Ada.Io_Exceptions'Elab_Spec;
27803 System.Exceptions'Elab_Spec;
27807 System.Stack_Checking'Elab_Spec;
27810 System.Soft_Links'Elab_Body;
27815 System.Secondary_Stack'Elab_Body;
27819 Ada.Tags'Elab_Spec;
27822 Ada.Tags'Elab_Body;
27826 Ada.Streams'Elab_Spec;
27830 System.Finalization_Root'Elab_Spec;
27834 Ada.Exceptions'Elab_Body;
27838 System.Finalization_Implementation'Elab_Spec;
27841 System.Finalization_Implementation'Elab_Body;
27845 Ada.Finalization'Elab_Spec;
27849 Ada.Finalization.List_Controller'Elab_Spec;
27853 System.File_Control_Block'Elab_Spec;
27857 System.File_Io'Elab_Body;
27861 Ada.Text_Io'Elab_Spec;
27864 Ada.Text_Io'Elab_Body;
27868 Elab_Final_Code := 0;
27876 procedure adafinal is
27885 -- main is actually a function, as in the ANSI C standard,
27886 -- defined to return the exit status. The three parameters
27887 -- are the argument count, argument values and environment
27890 @findex Main Program
27893 argv : System.Address;
27894 envp : System.Address)
27897 -- The initialize routine performs low level system
27898 -- initialization using a standard library routine which
27899 -- sets up signal handling and performs any other
27900 -- required setup. The routine can be found in file
27903 @findex __gnat_initialize
27904 procedure initialize;
27905 pragma Import (C, initialize, "__gnat_initialize");
27907 -- The finalize routine performs low level system
27908 -- finalization using a standard library routine. The
27909 -- routine is found in file a-final.c and in the standard
27910 -- distribution is a dummy routine that does nothing, so
27911 -- really this is a hook for special user finalization.
27913 @findex __gnat_finalize
27914 procedure finalize;
27915 pragma Import (C, finalize, "__gnat_finalize");
27917 -- We get to the main program of the partition by using
27918 -- pragma Import because if we try to with the unit and
27919 -- call it Ada style, then not only do we waste time
27920 -- recompiling it, but also, we don't really know the right
27921 -- switches (e.g.@: identifier character set) to be used
27924 procedure Ada_Main_Program;
27925 pragma Import (Ada, Ada_Main_Program, "_ada_hello");
27927 -- Start of processing for main
27930 -- Save global variables
27936 -- Call low level system initialization
27940 -- Call our generated Ada initialization routine
27944 -- This is the point at which we want the debugger to get
27949 -- Now we call the main program of the partition
27953 -- Perform Ada finalization
27957 -- Perform low level system finalization
27961 -- Return the proper exit status
27962 return (gnat_exit_status);
27965 -- This section is entirely comments, so it has no effect on the
27966 -- compilation of the Ada_Main package. It provides the list of
27967 -- object files and linker options, as well as some standard
27968 -- libraries needed for the link. The gnatlink utility parses
27969 -- this b~hello.adb file to read these comment lines to generate
27970 -- the appropriate command line arguments for the call to the
27971 -- system linker. The BEGIN/END lines are used for sentinels for
27972 -- this parsing operation.
27974 -- The exact file names will of course depend on the environment,
27975 -- host/target and location of files on the host system.
27977 @findex Object file list
27978 -- BEGIN Object file/option list
27981 -- -L/usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/
27982 -- /usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/libgnat.a
27983 -- END Object file/option list
27989 The Ada code in the above example is exactly what is generated by the
27990 binder. We have added comments to more clearly indicate the function
27991 of each part of the generated @code{Ada_Main} package.
27993 The code is standard Ada in all respects, and can be processed by any
27994 tools that handle Ada. In particular, it is possible to use the debugger
27995 in Ada mode to debug the generated @code{Ada_Main} package. For example,
27996 suppose that for reasons that you do not understand, your program is crashing
27997 during elaboration of the body of @code{Ada.Text_IO}. To locate this bug,
27998 you can place a breakpoint on the call:
28000 @smallexample @c ada
28001 Ada.Text_Io'Elab_Body;
28005 and trace the elaboration routine for this package to find out where
28006 the problem might be (more usually of course you would be debugging
28007 elaboration code in your own application).
28009 @node Elaboration Order Handling in GNAT
28010 @appendix Elaboration Order Handling in GNAT
28011 @cindex Order of elaboration
28012 @cindex Elaboration control
28015 * Elaboration Code::
28016 * Checking the Elaboration Order::
28017 * Controlling the Elaboration Order::
28018 * Controlling Elaboration in GNAT - Internal Calls::
28019 * Controlling Elaboration in GNAT - External Calls::
28020 * Default Behavior in GNAT - Ensuring Safety::
28021 * Treatment of Pragma Elaborate::
28022 * Elaboration Issues for Library Tasks::
28023 * Mixing Elaboration Models::
28024 * What to Do If the Default Elaboration Behavior Fails::
28025 * Elaboration for Access-to-Subprogram Values::
28026 * Summary of Procedures for Elaboration Control::
28027 * Other Elaboration Order Considerations::
28031 This chapter describes the handling of elaboration code in Ada and
28032 in GNAT, and discusses how the order of elaboration of program units can
28033 be controlled in GNAT, either automatically or with explicit programming
28036 @node Elaboration Code
28037 @section Elaboration Code
28040 Ada provides rather general mechanisms for executing code at elaboration
28041 time, that is to say before the main program starts executing. Such code arises
28045 @item Initializers for variables.
28046 Variables declared at the library level, in package specs or bodies, can
28047 require initialization that is performed at elaboration time, as in:
28048 @smallexample @c ada
28050 Sqrt_Half : Float := Sqrt (0.5);
28054 @item Package initialization code
28055 Code in a @code{BEGIN-END} section at the outer level of a package body is
28056 executed as part of the package body elaboration code.
28058 @item Library level task allocators
28059 Tasks that are declared using task allocators at the library level
28060 start executing immediately and hence can execute at elaboration time.
28064 Subprogram calls are possible in any of these contexts, which means that
28065 any arbitrary part of the program may be executed as part of the elaboration
28066 code. It is even possible to write a program which does all its work at
28067 elaboration time, with a null main program, although stylistically this
28068 would usually be considered an inappropriate way to structure
28071 An important concern arises in the context of elaboration code:
28072 we have to be sure that it is executed in an appropriate order. What we
28073 have is a series of elaboration code sections, potentially one section
28074 for each unit in the program. It is important that these execute
28075 in the correct order. Correctness here means that, taking the above
28076 example of the declaration of @code{Sqrt_Half},
28077 if some other piece of
28078 elaboration code references @code{Sqrt_Half},
28079 then it must run after the
28080 section of elaboration code that contains the declaration of
28083 There would never be any order of elaboration problem if we made a rule
28084 that whenever you @code{with} a unit, you must elaborate both the spec and body
28085 of that unit before elaborating the unit doing the @code{with}'ing:
28087 @smallexample @c ada
28091 package Unit_2 is @dots{}
28097 would require that both the body and spec of @code{Unit_1} be elaborated
28098 before the spec of @code{Unit_2}. However, a rule like that would be far too
28099 restrictive. In particular, it would make it impossible to have routines
28100 in separate packages that were mutually recursive.
28102 You might think that a clever enough compiler could look at the actual
28103 elaboration code and determine an appropriate correct order of elaboration,
28104 but in the general case, this is not possible. Consider the following
28107 In the body of @code{Unit_1}, we have a procedure @code{Func_1}
28109 the variable @code{Sqrt_1}, which is declared in the elaboration code
28110 of the body of @code{Unit_1}:
28112 @smallexample @c ada
28114 Sqrt_1 : Float := Sqrt (0.1);
28119 The elaboration code of the body of @code{Unit_1} also contains:
28121 @smallexample @c ada
28124 if expression_1 = 1 then
28125 Q := Unit_2.Func_2;
28132 @code{Unit_2} is exactly parallel,
28133 it has a procedure @code{Func_2} that references
28134 the variable @code{Sqrt_2}, which is declared in the elaboration code of
28135 the body @code{Unit_2}:
28137 @smallexample @c ada
28139 Sqrt_2 : Float := Sqrt (0.1);
28144 The elaboration code of the body of @code{Unit_2} also contains:
28146 @smallexample @c ada
28149 if expression_2 = 2 then
28150 Q := Unit_1.Func_1;
28157 Now the question is, which of the following orders of elaboration is
28182 If you carefully analyze the flow here, you will see that you cannot tell
28183 at compile time the answer to this question.
28184 If @code{expression_1} is not equal to 1,
28185 and @code{expression_2} is not equal to 2,
28186 then either order is acceptable, because neither of the function calls is
28187 executed. If both tests evaluate to true, then neither order is acceptable
28188 and in fact there is no correct order.
28190 If one of the two expressions is true, and the other is false, then one
28191 of the above orders is correct, and the other is incorrect. For example,
28192 if @code{expression_1} /= 1 and @code{expression_2} = 2,
28193 then the call to @code{Func_1}
28194 will occur, but not the call to @code{Func_2.}
28195 This means that it is essential
28196 to elaborate the body of @code{Unit_1} before
28197 the body of @code{Unit_2}, so the first
28198 order of elaboration is correct and the second is wrong.
28200 By making @code{expression_1} and @code{expression_2}
28201 depend on input data, or perhaps
28202 the time of day, we can make it impossible for the compiler or binder
28203 to figure out which of these expressions will be true, and hence it
28204 is impossible to guarantee a safe order of elaboration at run time.
28206 @node Checking the Elaboration Order
28207 @section Checking the Elaboration Order
28210 In some languages that involve the same kind of elaboration problems,
28211 e.g.@: Java and C++, the programmer is expected to worry about these
28212 ordering problems himself, and it is common to
28213 write a program in which an incorrect elaboration order gives
28214 surprising results, because it references variables before they
28216 Ada is designed to be a safe language, and a programmer-beware approach is
28217 clearly not sufficient. Consequently, the language provides three lines
28221 @item Standard rules
28222 Some standard rules restrict the possible choice of elaboration
28223 order. In particular, if you @code{with} a unit, then its spec is always
28224 elaborated before the unit doing the @code{with}. Similarly, a parent
28225 spec is always elaborated before the child spec, and finally
28226 a spec is always elaborated before its corresponding body.
28228 @item Dynamic elaboration checks
28229 @cindex Elaboration checks
28230 @cindex Checks, elaboration
28231 Dynamic checks are made at run time, so that if some entity is accessed
28232 before it is elaborated (typically by means of a subprogram call)
28233 then the exception (@code{Program_Error}) is raised.
28235 @item Elaboration control
28236 Facilities are provided for the programmer to specify the desired order
28240 Let's look at these facilities in more detail. First, the rules for
28241 dynamic checking. One possible rule would be simply to say that the
28242 exception is raised if you access a variable which has not yet been
28243 elaborated. The trouble with this approach is that it could require
28244 expensive checks on every variable reference. Instead Ada has two
28245 rules which are a little more restrictive, but easier to check, and
28249 @item Restrictions on calls
28250 A subprogram can only be called at elaboration time if its body
28251 has been elaborated. The rules for elaboration given above guarantee
28252 that the spec of the subprogram has been elaborated before the
28253 call, but not the body. If this rule is violated, then the
28254 exception @code{Program_Error} is raised.
28256 @item Restrictions on instantiations
28257 A generic unit can only be instantiated if the body of the generic
28258 unit has been elaborated. Again, the rules for elaboration given above
28259 guarantee that the spec of the generic unit has been elaborated
28260 before the instantiation, but not the body. If this rule is
28261 violated, then the exception @code{Program_Error} is raised.
28265 The idea is that if the body has been elaborated, then any variables
28266 it references must have been elaborated; by checking for the body being
28267 elaborated we guarantee that none of its references causes any
28268 trouble. As we noted above, this is a little too restrictive, because a
28269 subprogram that has no non-local references in its body may in fact be safe
28270 to call. However, it really would be unsafe to rely on this, because
28271 it would mean that the caller was aware of details of the implementation
28272 in the body. This goes against the basic tenets of Ada.
28274 A plausible implementation can be described as follows.
28275 A Boolean variable is associated with each subprogram
28276 and each generic unit. This variable is initialized to False, and is set to
28277 True at the point body is elaborated. Every call or instantiation checks the
28278 variable, and raises @code{Program_Error} if the variable is False.
28280 Note that one might think that it would be good enough to have one Boolean
28281 variable for each package, but that would not deal with cases of trying
28282 to call a body in the same package as the call
28283 that has not been elaborated yet.
28284 Of course a compiler may be able to do enough analysis to optimize away
28285 some of the Boolean variables as unnecessary, and @code{GNAT} indeed
28286 does such optimizations, but still the easiest conceptual model is to
28287 think of there being one variable per subprogram.
28289 @node Controlling the Elaboration Order
28290 @section Controlling the Elaboration Order
28293 In the previous section we discussed the rules in Ada which ensure
28294 that @code{Program_Error} is raised if an incorrect elaboration order is
28295 chosen. This prevents erroneous executions, but we need mechanisms to
28296 specify a correct execution and avoid the exception altogether.
28297 To achieve this, Ada provides a number of features for controlling
28298 the order of elaboration. We discuss these features in this section.
28300 First, there are several ways of indicating to the compiler that a given
28301 unit has no elaboration problems:
28304 @item packages that do not require a body
28305 A library package that does not require a body does not permit
28306 a body (this rule was introduced in Ada 95).
28307 Thus if we have a such a package, as in:
28309 @smallexample @c ada
28312 package Definitions is
28314 type m is new integer;
28316 type a is array (1 .. 10) of m;
28317 type b is array (1 .. 20) of m;
28325 A package that @code{with}'s @code{Definitions} may safely instantiate
28326 @code{Definitions.Subp} because the compiler can determine that there
28327 definitely is no package body to worry about in this case
28330 @cindex pragma Pure
28332 Places sufficient restrictions on a unit to guarantee that
28333 no call to any subprogram in the unit can result in an
28334 elaboration problem. This means that the compiler does not need
28335 to worry about the point of elaboration of such units, and in
28336 particular, does not need to check any calls to any subprograms
28339 @item pragma Preelaborate
28340 @findex Preelaborate
28341 @cindex pragma Preelaborate
28342 This pragma places slightly less stringent restrictions on a unit than
28344 but these restrictions are still sufficient to ensure that there
28345 are no elaboration problems with any calls to the unit.
28347 @item pragma Elaborate_Body
28348 @findex Elaborate_Body
28349 @cindex pragma Elaborate_Body
28350 This pragma requires that the body of a unit be elaborated immediately
28351 after its spec. Suppose a unit @code{A} has such a pragma,
28352 and unit @code{B} does
28353 a @code{with} of unit @code{A}. Recall that the standard rules require
28354 the spec of unit @code{A}
28355 to be elaborated before the @code{with}'ing unit; given the pragma in
28356 @code{A}, we also know that the body of @code{A}
28357 will be elaborated before @code{B}, so
28358 that calls to @code{A} are safe and do not need a check.
28363 unlike pragma @code{Pure} and pragma @code{Preelaborate},
28365 @code{Elaborate_Body} does not guarantee that the program is
28366 free of elaboration problems, because it may not be possible
28367 to satisfy the requested elaboration order.
28368 Let's go back to the example with @code{Unit_1} and @code{Unit_2}.
28370 marks @code{Unit_1} as @code{Elaborate_Body},
28371 and not @code{Unit_2,} then the order of
28372 elaboration will be:
28384 Now that means that the call to @code{Func_1} in @code{Unit_2}
28385 need not be checked,
28386 it must be safe. But the call to @code{Func_2} in
28387 @code{Unit_1} may still fail if
28388 @code{Expression_1} is equal to 1,
28389 and the programmer must still take
28390 responsibility for this not being the case.
28392 If all units carry a pragma @code{Elaborate_Body}, then all problems are
28393 eliminated, except for calls entirely within a body, which are
28394 in any case fully under programmer control. However, using the pragma
28395 everywhere is not always possible.
28396 In particular, for our @code{Unit_1}/@code{Unit_2} example, if
28397 we marked both of them as having pragma @code{Elaborate_Body}, then
28398 clearly there would be no possible elaboration order.
28400 The above pragmas allow a server to guarantee safe use by clients, and
28401 clearly this is the preferable approach. Consequently a good rule
28402 is to mark units as @code{Pure} or @code{Preelaborate} if possible,
28403 and if this is not possible,
28404 mark them as @code{Elaborate_Body} if possible.
28405 As we have seen, there are situations where neither of these
28406 three pragmas can be used.
28407 So we also provide methods for clients to control the
28408 order of elaboration of the servers on which they depend:
28411 @item pragma Elaborate (unit)
28413 @cindex pragma Elaborate
28414 This pragma is placed in the context clause, after a @code{with} clause,
28415 and it requires that the body of the named unit be elaborated before
28416 the unit in which the pragma occurs. The idea is to use this pragma
28417 if the current unit calls at elaboration time, directly or indirectly,
28418 some subprogram in the named unit.
28420 @item pragma Elaborate_All (unit)
28421 @findex Elaborate_All
28422 @cindex pragma Elaborate_All
28423 This is a stronger version of the Elaborate pragma. Consider the
28427 Unit A @code{with}'s unit B and calls B.Func in elab code
28428 Unit B @code{with}'s unit C, and B.Func calls C.Func
28432 Now if we put a pragma @code{Elaborate (B)}
28433 in unit @code{A}, this ensures that the
28434 body of @code{B} is elaborated before the call, but not the
28435 body of @code{C}, so
28436 the call to @code{C.Func} could still cause @code{Program_Error} to
28439 The effect of a pragma @code{Elaborate_All} is stronger, it requires
28440 not only that the body of the named unit be elaborated before the
28441 unit doing the @code{with}, but also the bodies of all units that the
28442 named unit uses, following @code{with} links transitively. For example,
28443 if we put a pragma @code{Elaborate_All (B)} in unit @code{A},
28445 not only that the body of @code{B} be elaborated before @code{A},
28447 body of @code{C}, because @code{B} @code{with}'s @code{C}.
28451 We are now in a position to give a usage rule in Ada for avoiding
28452 elaboration problems, at least if dynamic dispatching and access to
28453 subprogram values are not used. We will handle these cases separately
28456 The rule is simple. If a unit has elaboration code that can directly or
28457 indirectly make a call to a subprogram in a @code{with}'ed unit, or instantiate
28458 a generic package in a @code{with}'ed unit,
28459 then if the @code{with}'ed unit does not have
28460 pragma @code{Pure} or @code{Preelaborate}, then the client should have
28461 a pragma @code{Elaborate_All}
28462 for the @code{with}'ed unit. By following this rule a client is
28463 assured that calls can be made without risk of an exception.
28465 For generic subprogram instantiations, the rule can be relaxed to
28466 require only a pragma @code{Elaborate} since elaborating the body
28467 of a subprogram cannot cause any transitive elaboration (we are
28468 not calling the subprogram in this case, just elaborating its
28471 If this rule is not followed, then a program may be in one of four
28475 @item No order exists
28476 No order of elaboration exists which follows the rules, taking into
28477 account any @code{Elaborate}, @code{Elaborate_All},
28478 or @code{Elaborate_Body} pragmas. In
28479 this case, an Ada compiler must diagnose the situation at bind
28480 time, and refuse to build an executable program.
28482 @item One or more orders exist, all incorrect
28483 One or more acceptable elaboration orders exist, and all of them
28484 generate an elaboration order problem. In this case, the binder
28485 can build an executable program, but @code{Program_Error} will be raised
28486 when the program is run.
28488 @item Several orders exist, some right, some incorrect
28489 One or more acceptable elaboration orders exists, and some of them
28490 work, and some do not. The programmer has not controlled
28491 the order of elaboration, so the binder may or may not pick one of
28492 the correct orders, and the program may or may not raise an
28493 exception when it is run. This is the worst case, because it means
28494 that the program may fail when moved to another compiler, or even
28495 another version of the same compiler.
28497 @item One or more orders exists, all correct
28498 One ore more acceptable elaboration orders exist, and all of them
28499 work. In this case the program runs successfully. This state of
28500 affairs can be guaranteed by following the rule we gave above, but
28501 may be true even if the rule is not followed.
28505 Note that one additional advantage of following our rules on the use
28506 of @code{Elaborate} and @code{Elaborate_All}
28507 is that the program continues to stay in the ideal (all orders OK) state
28508 even if maintenance
28509 changes some bodies of some units. Conversely, if a program that does
28510 not follow this rule happens to be safe at some point, this state of affairs
28511 may deteriorate silently as a result of maintenance changes.
28513 You may have noticed that the above discussion did not mention
28514 the use of @code{Elaborate_Body}. This was a deliberate omission. If you
28515 @code{with} an @code{Elaborate_Body} unit, it still may be the case that
28516 code in the body makes calls to some other unit, so it is still necessary
28517 to use @code{Elaborate_All} on such units.
28519 @node Controlling Elaboration in GNAT - Internal Calls
28520 @section Controlling Elaboration in GNAT - Internal Calls
28523 In the case of internal calls, i.e., calls within a single package, the
28524 programmer has full control over the order of elaboration, and it is up
28525 to the programmer to elaborate declarations in an appropriate order. For
28528 @smallexample @c ada
28531 function One return Float;
28535 function One return Float is
28544 will obviously raise @code{Program_Error} at run time, because function
28545 One will be called before its body is elaborated. In this case GNAT will
28546 generate a warning that the call will raise @code{Program_Error}:
28552 2. function One return Float;
28554 4. Q : Float := One;
28556 >>> warning: cannot call "One" before body is elaborated
28557 >>> warning: Program_Error will be raised at run time
28560 6. function One return Float is
28573 Note that in this particular case, it is likely that the call is safe, because
28574 the function @code{One} does not access any global variables.
28575 Nevertheless in Ada, we do not want the validity of the check to depend on
28576 the contents of the body (think about the separate compilation case), so this
28577 is still wrong, as we discussed in the previous sections.
28579 The error is easily corrected by rearranging the declarations so that the
28580 body of @code{One} appears before the declaration containing the call
28581 (note that in Ada 95 and Ada 2005,
28582 declarations can appear in any order, so there is no restriction that
28583 would prevent this reordering, and if we write:
28585 @smallexample @c ada
28588 function One return Float;
28590 function One return Float is
28601 then all is well, no warning is generated, and no
28602 @code{Program_Error} exception
28604 Things are more complicated when a chain of subprograms is executed:
28606 @smallexample @c ada
28609 function A return Integer;
28610 function B return Integer;
28611 function C return Integer;
28613 function B return Integer is begin return A; end;
28614 function C return Integer is begin return B; end;
28618 function A return Integer is begin return 1; end;
28624 Now the call to @code{C}
28625 at elaboration time in the declaration of @code{X} is correct, because
28626 the body of @code{C} is already elaborated,
28627 and the call to @code{B} within the body of
28628 @code{C} is correct, but the call
28629 to @code{A} within the body of @code{B} is incorrect, because the body
28630 of @code{A} has not been elaborated, so @code{Program_Error}
28631 will be raised on the call to @code{A}.
28632 In this case GNAT will generate a
28633 warning that @code{Program_Error} may be
28634 raised at the point of the call. Let's look at the warning:
28640 2. function A return Integer;
28641 3. function B return Integer;
28642 4. function C return Integer;
28644 6. function B return Integer is begin return A; end;
28646 >>> warning: call to "A" before body is elaborated may
28647 raise Program_Error
28648 >>> warning: "B" called at line 7
28649 >>> warning: "C" called at line 9
28651 7. function C return Integer is begin return B; end;
28653 9. X : Integer := C;
28655 11. function A return Integer is begin return 1; end;
28665 Note that the message here says ``may raise'', instead of the direct case,
28666 where the message says ``will be raised''. That's because whether
28668 actually called depends in general on run-time flow of control.
28669 For example, if the body of @code{B} said
28671 @smallexample @c ada
28674 function B return Integer is
28676 if some-condition-depending-on-input-data then
28687 then we could not know until run time whether the incorrect call to A would
28688 actually occur, so @code{Program_Error} might
28689 or might not be raised. It is possible for a compiler to
28690 do a better job of analyzing bodies, to
28691 determine whether or not @code{Program_Error}
28692 might be raised, but it certainly
28693 couldn't do a perfect job (that would require solving the halting problem
28694 and is provably impossible), and because this is a warning anyway, it does
28695 not seem worth the effort to do the analysis. Cases in which it
28696 would be relevant are rare.
28698 In practice, warnings of either of the forms given
28699 above will usually correspond to
28700 real errors, and should be examined carefully and eliminated.
28701 In the rare case where a warning is bogus, it can be suppressed by any of
28702 the following methods:
28706 Compile with the @option{-gnatws} switch set
28709 Suppress @code{Elaboration_Check} for the called subprogram
28712 Use pragma @code{Warnings_Off} to turn warnings off for the call
28716 For the internal elaboration check case,
28717 GNAT by default generates the
28718 necessary run-time checks to ensure
28719 that @code{Program_Error} is raised if any
28720 call fails an elaboration check. Of course this can only happen if a
28721 warning has been issued as described above. The use of pragma
28722 @code{Suppress (Elaboration_Check)} may (but is not guaranteed to) suppress
28723 some of these checks, meaning that it may be possible (but is not
28724 guaranteed) for a program to be able to call a subprogram whose body
28725 is not yet elaborated, without raising a @code{Program_Error} exception.
28727 @node Controlling Elaboration in GNAT - External Calls
28728 @section Controlling Elaboration in GNAT - External Calls
28731 The previous section discussed the case in which the execution of a
28732 particular thread of elaboration code occurred entirely within a
28733 single unit. This is the easy case to handle, because a programmer
28734 has direct and total control over the order of elaboration, and
28735 furthermore, checks need only be generated in cases which are rare
28736 and which the compiler can easily detect.
28737 The situation is more complex when separate compilation is taken into account.
28738 Consider the following:
28740 @smallexample @c ada
28744 function Sqrt (Arg : Float) return Float;
28747 package body Math is
28748 function Sqrt (Arg : Float) return Float is
28757 X : Float := Math.Sqrt (0.5);
28770 where @code{Main} is the main program. When this program is executed, the
28771 elaboration code must first be executed, and one of the jobs of the
28772 binder is to determine the order in which the units of a program are
28773 to be elaborated. In this case we have four units: the spec and body
28775 the spec of @code{Stuff} and the body of @code{Main}).
28776 In what order should the four separate sections of elaboration code
28779 There are some restrictions in the order of elaboration that the binder
28780 can choose. In particular, if unit U has a @code{with}
28781 for a package @code{X}, then you
28782 are assured that the spec of @code{X}
28783 is elaborated before U , but you are
28784 not assured that the body of @code{X}
28785 is elaborated before U.
28786 This means that in the above case, the binder is allowed to choose the
28797 but that's not good, because now the call to @code{Math.Sqrt}
28798 that happens during
28799 the elaboration of the @code{Stuff}
28800 spec happens before the body of @code{Math.Sqrt} is
28801 elaborated, and hence causes @code{Program_Error} exception to be raised.
28802 At first glance, one might say that the binder is misbehaving, because
28803 obviously you want to elaborate the body of something you @code{with}
28805 that is not a general rule that can be followed in all cases. Consider
28807 @smallexample @c ada
28810 package X is @dots{}
28812 package Y is @dots{}
28815 package body Y is @dots{}
28818 package body X is @dots{}
28824 This is a common arrangement, and, apart from the order of elaboration
28825 problems that might arise in connection with elaboration code, this works fine.
28826 A rule that says that you must first elaborate the body of anything you
28827 @code{with} cannot work in this case:
28828 the body of @code{X} @code{with}'s @code{Y},
28829 which means you would have to
28830 elaborate the body of @code{Y} first, but that @code{with}'s @code{X},
28832 you have to elaborate the body of @code{X} first, but @dots{} and we have a
28833 loop that cannot be broken.
28835 It is true that the binder can in many cases guess an order of elaboration
28836 that is unlikely to cause a @code{Program_Error}
28837 exception to be raised, and it tries to do so (in the
28838 above example of @code{Math/Stuff/Spec}, the GNAT binder will
28840 elaborate the body of @code{Math} right after its spec, so all will be well).
28842 However, a program that blindly relies on the binder to be helpful can
28843 get into trouble, as we discussed in the previous sections, so
28845 provides a number of facilities for assisting the programmer in
28846 developing programs that are robust with respect to elaboration order.
28848 @node Default Behavior in GNAT - Ensuring Safety
28849 @section Default Behavior in GNAT - Ensuring Safety
28852 The default behavior in GNAT ensures elaboration safety. In its
28853 default mode GNAT implements the
28854 rule we previously described as the right approach. Let's restate it:
28858 @emph{If a unit has elaboration code that can directly or indirectly make a
28859 call to a subprogram in a @code{with}'ed unit, or instantiate a generic
28860 package in a @code{with}'ed unit, then if the @code{with}'ed unit
28861 does not have pragma @code{Pure} or
28862 @code{Preelaborate}, then the client should have an
28863 @code{Elaborate_All} pragma for the @code{with}'ed unit.}
28865 @emph{In the case of instantiating a generic subprogram, it is always
28866 sufficient to have only an @code{Elaborate} pragma for the
28867 @code{with}'ed unit.}
28871 By following this rule a client is assured that calls and instantiations
28872 can be made without risk of an exception.
28874 In this mode GNAT traces all calls that are potentially made from
28875 elaboration code, and puts in any missing implicit @code{Elaborate}
28876 and @code{Elaborate_All} pragmas.
28877 The advantage of this approach is that no elaboration problems
28878 are possible if the binder can find an elaboration order that is
28879 consistent with these implicit @code{Elaborate} and
28880 @code{Elaborate_All} pragmas. The
28881 disadvantage of this approach is that no such order may exist.
28883 If the binder does not generate any diagnostics, then it means that it has
28884 found an elaboration order that is guaranteed to be safe. However, the binder
28885 may still be relying on implicitly generated @code{Elaborate} and
28886 @code{Elaborate_All} pragmas so portability to other compilers than GNAT is not
28889 If it is important to guarantee portability, then the compilations should
28892 (warn on elaboration problems) switch. This will cause warning messages
28893 to be generated indicating the missing @code{Elaborate} and
28894 @code{Elaborate_All} pragmas.
28895 Consider the following source program:
28897 @smallexample @c ada
28902 m : integer := k.r;
28909 where it is clear that there
28910 should be a pragma @code{Elaborate_All}
28911 for unit @code{k}. An implicit pragma will be generated, and it is
28912 likely that the binder will be able to honor it. However, if you want
28913 to port this program to some other Ada compiler than GNAT.
28914 it is safer to include the pragma explicitly in the source. If this
28915 unit is compiled with the
28917 switch, then the compiler outputs a warning:
28924 3. m : integer := k.r;
28926 >>> warning: call to "r" may raise Program_Error
28927 >>> warning: missing pragma Elaborate_All for "k"
28935 and these warnings can be used as a guide for supplying manually
28936 the missing pragmas. It is usually a bad idea to use this warning
28937 option during development. That's because it will warn you when
28938 you need to put in a pragma, but cannot warn you when it is time
28939 to take it out. So the use of pragma @code{Elaborate_All} may lead to
28940 unnecessary dependencies and even false circularities.
28942 This default mode is more restrictive than the Ada Reference
28943 Manual, and it is possible to construct programs which will compile
28944 using the dynamic model described there, but will run into a
28945 circularity using the safer static model we have described.
28947 Of course any Ada compiler must be able to operate in a mode
28948 consistent with the requirements of the Ada Reference Manual,
28949 and in particular must have the capability of implementing the
28950 standard dynamic model of elaboration with run-time checks.
28952 In GNAT, this standard mode can be achieved either by the use of
28953 the @option{-gnatE} switch on the compiler (@command{gcc} or
28954 @command{gnatmake}) command, or by the use of the configuration pragma:
28956 @smallexample @c ada
28957 pragma Elaboration_Checks (DYNAMIC);
28961 Either approach will cause the unit affected to be compiled using the
28962 standard dynamic run-time elaboration checks described in the Ada
28963 Reference Manual. The static model is generally preferable, since it
28964 is clearly safer to rely on compile and link time checks rather than
28965 run-time checks. However, in the case of legacy code, it may be
28966 difficult to meet the requirements of the static model. This
28967 issue is further discussed in
28968 @ref{What to Do If the Default Elaboration Behavior Fails}.
28970 Note that the static model provides a strict subset of the allowed
28971 behavior and programs of the Ada Reference Manual, so if you do
28972 adhere to the static model and no circularities exist,
28973 then you are assured that your program will
28974 work using the dynamic model, providing that you remove any
28975 pragma Elaborate statements from the source.
28977 @node Treatment of Pragma Elaborate
28978 @section Treatment of Pragma Elaborate
28979 @cindex Pragma Elaborate
28982 The use of @code{pragma Elaborate}
28983 should generally be avoided in Ada 95 and Ada 2005 programs,
28984 since there is no guarantee that transitive calls
28985 will be properly handled. Indeed at one point, this pragma was placed
28986 in Annex J (Obsolescent Features), on the grounds that it is never useful.
28988 Now that's a bit restrictive. In practice, the case in which
28989 @code{pragma Elaborate} is useful is when the caller knows that there
28990 are no transitive calls, or that the called unit contains all necessary
28991 transitive @code{pragma Elaborate} statements, and legacy code often
28992 contains such uses.
28994 Strictly speaking the static mode in GNAT should ignore such pragmas,
28995 since there is no assurance at compile time that the necessary safety
28996 conditions are met. In practice, this would cause GNAT to be incompatible
28997 with correctly written Ada 83 code that had all necessary
28998 @code{pragma Elaborate} statements in place. Consequently, we made the
28999 decision that GNAT in its default mode will believe that if it encounters
29000 a @code{pragma Elaborate} then the programmer knows what they are doing,
29001 and it will trust that no elaboration errors can occur.
29003 The result of this decision is two-fold. First to be safe using the
29004 static mode, you should remove all @code{pragma Elaborate} statements.
29005 Second, when fixing circularities in existing code, you can selectively
29006 use @code{pragma Elaborate} statements to convince the static mode of
29007 GNAT that it need not generate an implicit @code{pragma Elaborate_All}
29010 When using the static mode with @option{-gnatwl}, any use of
29011 @code{pragma Elaborate} will generate a warning about possible
29014 @node Elaboration Issues for Library Tasks
29015 @section Elaboration Issues for Library Tasks
29016 @cindex Library tasks, elaboration issues
29017 @cindex Elaboration of library tasks
29020 In this section we examine special elaboration issues that arise for
29021 programs that declare library level tasks.
29023 Generally the model of execution of an Ada program is that all units are
29024 elaborated, and then execution of the program starts. However, the
29025 declaration of library tasks definitely does not fit this model. The
29026 reason for this is that library tasks start as soon as they are declared
29027 (more precisely, as soon as the statement part of the enclosing package
29028 body is reached), that is to say before elaboration
29029 of the program is complete. This means that if such a task calls a
29030 subprogram, or an entry in another task, the callee may or may not be
29031 elaborated yet, and in the standard
29032 Reference Manual model of dynamic elaboration checks, you can even
29033 get timing dependent Program_Error exceptions, since there can be
29034 a race between the elaboration code and the task code.
29036 The static model of elaboration in GNAT seeks to avoid all such
29037 dynamic behavior, by being conservative, and the conservative
29038 approach in this particular case is to assume that all the code
29039 in a task body is potentially executed at elaboration time if
29040 a task is declared at the library level.
29042 This can definitely result in unexpected circularities. Consider
29043 the following example
29045 @smallexample @c ada
29051 type My_Int is new Integer;
29053 function Ident (M : My_Int) return My_Int;
29057 package body Decls is
29058 task body Lib_Task is
29064 function Ident (M : My_Int) return My_Int is
29072 procedure Put_Val (Arg : Decls.My_Int);
29076 package body Utils is
29077 procedure Put_Val (Arg : Decls.My_Int) is
29079 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
29086 Decls.Lib_Task.Start;
29091 If the above example is compiled in the default static elaboration
29092 mode, then a circularity occurs. The circularity comes from the call
29093 @code{Utils.Put_Val} in the task body of @code{Decls.Lib_Task}. Since
29094 this call occurs in elaboration code, we need an implicit pragma
29095 @code{Elaborate_All} for @code{Utils}. This means that not only must
29096 the spec and body of @code{Utils} be elaborated before the body
29097 of @code{Decls}, but also the spec and body of any unit that is
29098 @code{with'ed} by the body of @code{Utils} must also be elaborated before
29099 the body of @code{Decls}. This is the transitive implication of
29100 pragma @code{Elaborate_All} and it makes sense, because in general
29101 the body of @code{Put_Val} might have a call to something in a
29102 @code{with'ed} unit.
29104 In this case, the body of Utils (actually its spec) @code{with's}
29105 @code{Decls}. Unfortunately this means that the body of @code{Decls}
29106 must be elaborated before itself, in case there is a call from the
29107 body of @code{Utils}.
29109 Here is the exact chain of events we are worrying about:
29113 In the body of @code{Decls} a call is made from within the body of a library
29114 task to a subprogram in the package @code{Utils}. Since this call may
29115 occur at elaboration time (given that the task is activated at elaboration
29116 time), we have to assume the worst, i.e., that the
29117 call does happen at elaboration time.
29120 This means that the body and spec of @code{Util} must be elaborated before
29121 the body of @code{Decls} so that this call does not cause an access before
29125 Within the body of @code{Util}, specifically within the body of
29126 @code{Util.Put_Val} there may be calls to any unit @code{with}'ed
29130 One such @code{with}'ed package is package @code{Decls}, so there
29131 might be a call to a subprogram in @code{Decls} in @code{Put_Val}.
29132 In fact there is such a call in this example, but we would have to
29133 assume that there was such a call even if it were not there, since
29134 we are not supposed to write the body of @code{Decls} knowing what
29135 is in the body of @code{Utils}; certainly in the case of the
29136 static elaboration model, the compiler does not know what is in
29137 other bodies and must assume the worst.
29140 This means that the spec and body of @code{Decls} must also be
29141 elaborated before we elaborate the unit containing the call, but
29142 that unit is @code{Decls}! This means that the body of @code{Decls}
29143 must be elaborated before itself, and that's a circularity.
29147 Indeed, if you add an explicit pragma @code{Elaborate_All} for @code{Utils} in
29148 the body of @code{Decls} you will get a true Ada Reference Manual
29149 circularity that makes the program illegal.
29151 In practice, we have found that problems with the static model of
29152 elaboration in existing code often arise from library tasks, so
29153 we must address this particular situation.
29155 Note that if we compile and run the program above, using the dynamic model of
29156 elaboration (that is to say use the @option{-gnatE} switch),
29157 then it compiles, binds,
29158 links, and runs, printing the expected result of 2. Therefore in some sense
29159 the circularity here is only apparent, and we need to capture
29160 the properties of this program that distinguish it from other library-level
29161 tasks that have real elaboration problems.
29163 We have four possible answers to this question:
29168 Use the dynamic model of elaboration.
29170 If we use the @option{-gnatE} switch, then as noted above, the program works.
29171 Why is this? If we examine the task body, it is apparent that the task cannot
29173 @code{accept} statement until after elaboration has been completed, because
29174 the corresponding entry call comes from the main program, not earlier.
29175 This is why the dynamic model works here. But that's really giving
29176 up on a precise analysis, and we prefer to take this approach only if we cannot
29178 problem in any other manner. So let us examine two ways to reorganize
29179 the program to avoid the potential elaboration problem.
29182 Split library tasks into separate packages.
29184 Write separate packages, so that library tasks are isolated from
29185 other declarations as much as possible. Let us look at a variation on
29188 @smallexample @c ada
29196 package body Decls1 is
29197 task body Lib_Task is
29205 type My_Int is new Integer;
29206 function Ident (M : My_Int) return My_Int;
29210 package body Decls2 is
29211 function Ident (M : My_Int) return My_Int is
29219 procedure Put_Val (Arg : Decls2.My_Int);
29223 package body Utils is
29224 procedure Put_Val (Arg : Decls2.My_Int) is
29226 Text_IO.Put_Line (Decls2.My_Int'Image (Decls2.Ident (Arg)));
29233 Decls1.Lib_Task.Start;
29238 All we have done is to split @code{Decls} into two packages, one
29239 containing the library task, and one containing everything else. Now
29240 there is no cycle, and the program compiles, binds, links and executes
29241 using the default static model of elaboration.
29244 Declare separate task types.
29246 A significant part of the problem arises because of the use of the
29247 single task declaration form. This means that the elaboration of
29248 the task type, and the elaboration of the task itself (i.e.@: the
29249 creation of the task) happen at the same time. A good rule
29250 of style in Ada is to always create explicit task types. By
29251 following the additional step of placing task objects in separate
29252 packages from the task type declaration, many elaboration problems
29253 are avoided. Here is another modified example of the example program:
29255 @smallexample @c ada
29257 task type Lib_Task_Type is
29261 type My_Int is new Integer;
29263 function Ident (M : My_Int) return My_Int;
29267 package body Decls is
29268 task body Lib_Task_Type is
29274 function Ident (M : My_Int) return My_Int is
29282 procedure Put_Val (Arg : Decls.My_Int);
29286 package body Utils is
29287 procedure Put_Val (Arg : Decls.My_Int) is
29289 Text_IO.Put_Line (Decls.My_Int'Image (Decls.Ident (Arg)));
29295 Lib_Task : Decls.Lib_Task_Type;
29301 Declst.Lib_Task.Start;
29306 What we have done here is to replace the @code{task} declaration in
29307 package @code{Decls} with a @code{task type} declaration. Then we
29308 introduce a separate package @code{Declst} to contain the actual
29309 task object. This separates the elaboration issues for
29310 the @code{task type}
29311 declaration, which causes no trouble, from the elaboration issues
29312 of the task object, which is also unproblematic, since it is now independent
29313 of the elaboration of @code{Utils}.
29314 This separation of concerns also corresponds to
29315 a generally sound engineering principle of separating declarations
29316 from instances. This version of the program also compiles, binds, links,
29317 and executes, generating the expected output.
29320 Use No_Entry_Calls_In_Elaboration_Code restriction.
29321 @cindex No_Entry_Calls_In_Elaboration_Code
29323 The previous two approaches described how a program can be restructured
29324 to avoid the special problems caused by library task bodies. in practice,
29325 however, such restructuring may be difficult to apply to existing legacy code,
29326 so we must consider solutions that do not require massive rewriting.
29328 Let us consider more carefully why our original sample program works
29329 under the dynamic model of elaboration. The reason is that the code
29330 in the task body blocks immediately on the @code{accept}
29331 statement. Now of course there is nothing to prohibit elaboration
29332 code from making entry calls (for example from another library level task),
29333 so we cannot tell in isolation that
29334 the task will not execute the accept statement during elaboration.
29336 However, in practice it is very unusual to see elaboration code
29337 make any entry calls, and the pattern of tasks starting
29338 at elaboration time and then immediately blocking on @code{accept} or
29339 @code{select} statements is very common. What this means is that
29340 the compiler is being too pessimistic when it analyzes the
29341 whole package body as though it might be executed at elaboration
29344 If we know that the elaboration code contains no entry calls, (a very safe
29345 assumption most of the time, that could almost be made the default
29346 behavior), then we can compile all units of the program under control
29347 of the following configuration pragma:
29350 pragma Restrictions (No_Entry_Calls_In_Elaboration_Code);
29354 This pragma can be placed in the @file{gnat.adc} file in the usual
29355 manner. If we take our original unmodified program and compile it
29356 in the presence of a @file{gnat.adc} containing the above pragma,
29357 then once again, we can compile, bind, link, and execute, obtaining
29358 the expected result. In the presence of this pragma, the compiler does
29359 not trace calls in a task body, that appear after the first @code{accept}
29360 or @code{select} statement, and therefore does not report a potential
29361 circularity in the original program.
29363 The compiler will check to the extent it can that the above
29364 restriction is not violated, but it is not always possible to do a
29365 complete check at compile time, so it is important to use this
29366 pragma only if the stated restriction is in fact met, that is to say
29367 no task receives an entry call before elaboration of all units is completed.
29371 @node Mixing Elaboration Models
29372 @section Mixing Elaboration Models
29374 So far, we have assumed that the entire program is either compiled
29375 using the dynamic model or static model, ensuring consistency. It
29376 is possible to mix the two models, but rules have to be followed
29377 if this mixing is done to ensure that elaboration checks are not
29380 The basic rule is that @emph{a unit compiled with the static model cannot
29381 be @code{with'ed} by a unit compiled with the dynamic model}. The
29382 reason for this is that in the static model, a unit assumes that
29383 its clients guarantee to use (the equivalent of) pragma
29384 @code{Elaborate_All} so that no elaboration checks are required
29385 in inner subprograms, and this assumption is violated if the
29386 client is compiled with dynamic checks.
29388 The precise rule is as follows. A unit that is compiled with dynamic
29389 checks can only @code{with} a unit that meets at least one of the
29390 following criteria:
29395 The @code{with'ed} unit is itself compiled with dynamic elaboration
29396 checks (that is with the @option{-gnatE} switch.
29399 The @code{with'ed} unit is an internal GNAT implementation unit from
29400 the System, Interfaces, Ada, or GNAT hierarchies.
29403 The @code{with'ed} unit has pragma Preelaborate or pragma Pure.
29406 The @code{with'ing} unit (that is the client) has an explicit pragma
29407 @code{Elaborate_All} for the @code{with'ed} unit.
29412 If this rule is violated, that is if a unit with dynamic elaboration
29413 checks @code{with's} a unit that does not meet one of the above four
29414 criteria, then the binder (@code{gnatbind}) will issue a warning
29415 similar to that in the following example:
29418 warning: "x.ads" has dynamic elaboration checks and with's
29419 warning: "y.ads" which has static elaboration checks
29423 These warnings indicate that the rule has been violated, and that as a result
29424 elaboration checks may be missed in the resulting executable file.
29425 This warning may be suppressed using the @option{-ws} binder switch
29426 in the usual manner.
29428 One useful application of this mixing rule is in the case of a subsystem
29429 which does not itself @code{with} units from the remainder of the
29430 application. In this case, the entire subsystem can be compiled with
29431 dynamic checks to resolve a circularity in the subsystem, while
29432 allowing the main application that uses this subsystem to be compiled
29433 using the more reliable default static model.
29435 @node What to Do If the Default Elaboration Behavior Fails
29436 @section What to Do If the Default Elaboration Behavior Fails
29439 If the binder cannot find an acceptable order, it outputs detailed
29440 diagnostics. For example:
29446 error: elaboration circularity detected
29447 info: "proc (body)" must be elaborated before "pack (body)"
29448 info: reason: Elaborate_All probably needed in unit "pack (body)"
29449 info: recompile "pack (body)" with -gnatwl
29450 info: for full details
29451 info: "proc (body)"
29452 info: is needed by its spec:
29453 info: "proc (spec)"
29454 info: which is withed by:
29455 info: "pack (body)"
29456 info: "pack (body)" must be elaborated before "proc (body)"
29457 info: reason: pragma Elaborate in unit "proc (body)"
29463 In this case we have a cycle that the binder cannot break. On the one
29464 hand, there is an explicit pragma Elaborate in @code{proc} for
29465 @code{pack}. This means that the body of @code{pack} must be elaborated
29466 before the body of @code{proc}. On the other hand, there is elaboration
29467 code in @code{pack} that calls a subprogram in @code{proc}. This means
29468 that for maximum safety, there should really be a pragma
29469 Elaborate_All in @code{pack} for @code{proc} which would require that
29470 the body of @code{proc} be elaborated before the body of
29471 @code{pack}. Clearly both requirements cannot be satisfied.
29472 Faced with a circularity of this kind, you have three different options.
29475 @item Fix the program
29476 The most desirable option from the point of view of long-term maintenance
29477 is to rearrange the program so that the elaboration problems are avoided.
29478 One useful technique is to place the elaboration code into separate
29479 child packages. Another is to move some of the initialization code to
29480 explicitly called subprograms, where the program controls the order
29481 of initialization explicitly. Although this is the most desirable option,
29482 it may be impractical and involve too much modification, especially in
29483 the case of complex legacy code.
29485 @item Perform dynamic checks
29486 If the compilations are done using the
29488 (dynamic elaboration check) switch, then GNAT behaves in a quite different
29489 manner. Dynamic checks are generated for all calls that could possibly result
29490 in raising an exception. With this switch, the compiler does not generate
29491 implicit @code{Elaborate} or @code{Elaborate_All} pragmas. The behavior then is
29492 exactly as specified in the @cite{Ada Reference Manual}.
29493 The binder will generate
29494 an executable program that may or may not raise @code{Program_Error}, and then
29495 it is the programmer's job to ensure that it does not raise an exception. Note
29496 that it is important to compile all units with the switch, it cannot be used
29499 @item Suppress checks
29500 The drawback of dynamic checks is that they generate a
29501 significant overhead at run time, both in space and time. If you
29502 are absolutely sure that your program cannot raise any elaboration
29503 exceptions, and you still want to use the dynamic elaboration model,
29504 then you can use the configuration pragma
29505 @code{Suppress (Elaboration_Check)} to suppress all such checks. For
29506 example this pragma could be placed in the @file{gnat.adc} file.
29508 @item Suppress checks selectively
29509 When you know that certain calls or instantiations in elaboration code cannot
29510 possibly lead to an elaboration error, and the binder nevertheless complains
29511 about implicit @code{Elaborate} and @code{Elaborate_All} pragmas that lead to
29512 elaboration circularities, it is possible to remove those warnings locally and
29513 obtain a program that will bind. Clearly this can be unsafe, and it is the
29514 responsibility of the programmer to make sure that the resulting program has no
29515 elaboration anomalies. The pragma @code{Suppress (Elaboration_Check)} can be
29516 used with different granularity to suppress warnings and break elaboration
29521 Place the pragma that names the called subprogram in the declarative part
29522 that contains the call.
29525 Place the pragma in the declarative part, without naming an entity. This
29526 disables warnings on all calls in the corresponding declarative region.
29529 Place the pragma in the package spec that declares the called subprogram,
29530 and name the subprogram. This disables warnings on all elaboration calls to
29534 Place the pragma in the package spec that declares the called subprogram,
29535 without naming any entity. This disables warnings on all elaboration calls to
29536 all subprograms declared in this spec.
29538 @item Use Pragma Elaborate
29539 As previously described in section @xref{Treatment of Pragma Elaborate},
29540 GNAT in static mode assumes that a @code{pragma} Elaborate indicates correctly
29541 that no elaboration checks are required on calls to the designated unit.
29542 There may be cases in which the caller knows that no transitive calls
29543 can occur, so that a @code{pragma Elaborate} will be sufficient in a
29544 case where @code{pragma Elaborate_All} would cause a circularity.
29548 These five cases are listed in order of decreasing safety, and therefore
29549 require increasing programmer care in their application. Consider the
29552 @smallexample @c adanocomment
29554 function F1 return Integer;
29559 function F2 return Integer;
29560 function Pure (x : integer) return integer;
29561 -- pragma Suppress (Elaboration_Check, On => Pure); -- (3)
29562 -- pragma Suppress (Elaboration_Check); -- (4)
29566 package body Pack1 is
29567 function F1 return Integer is
29571 Val : integer := Pack2.Pure (11); -- Elab. call (1)
29574 -- pragma Suppress(Elaboration_Check, Pack2.F2); -- (1)
29575 -- pragma Suppress(Elaboration_Check); -- (2)
29577 X1 := Pack2.F2 + 1; -- Elab. call (2)
29582 package body Pack2 is
29583 function F2 return Integer is
29587 function Pure (x : integer) return integer is
29589 return x ** 3 - 3 * x;
29593 with Pack1, Ada.Text_IO;
29596 Ada.Text_IO.Put_Line(Pack1.X1'Img); -- 101
29599 In the absence of any pragmas, an attempt to bind this program produces
29600 the following diagnostics:
29606 error: elaboration circularity detected
29607 info: "pack1 (body)" must be elaborated before "pack1 (body)"
29608 info: reason: Elaborate_All probably needed in unit "pack1 (body)"
29609 info: recompile "pack1 (body)" with -gnatwl for full details
29610 info: "pack1 (body)"
29611 info: must be elaborated along with its spec:
29612 info: "pack1 (spec)"
29613 info: which is withed by:
29614 info: "pack2 (body)"
29615 info: which must be elaborated along with its spec:
29616 info: "pack2 (spec)"
29617 info: which is withed by:
29618 info: "pack1 (body)"
29621 The sources of the circularity are the two calls to @code{Pack2.Pure} and
29622 @code{Pack2.F2} in the body of @code{Pack1}. We can see that the call to
29623 F2 is safe, even though F2 calls F1, because the call appears after the
29624 elaboration of the body of F1. Therefore the pragma (1) is safe, and will
29625 remove the warning on the call. It is also possible to use pragma (2)
29626 because there are no other potentially unsafe calls in the block.
29629 The call to @code{Pure} is safe because this function does not depend on the
29630 state of @code{Pack2}. Therefore any call to this function is safe, and it
29631 is correct to place pragma (3) in the corresponding package spec.
29634 Finally, we could place pragma (4) in the spec of @code{Pack2} to disable
29635 warnings on all calls to functions declared therein. Note that this is not
29636 necessarily safe, and requires more detailed examination of the subprogram
29637 bodies involved. In particular, a call to @code{F2} requires that @code{F1}
29638 be already elaborated.
29642 It is hard to generalize on which of these four approaches should be
29643 taken. Obviously if it is possible to fix the program so that the default
29644 treatment works, this is preferable, but this may not always be practical.
29645 It is certainly simple enough to use
29647 but the danger in this case is that, even if the GNAT binder
29648 finds a correct elaboration order, it may not always do so,
29649 and certainly a binder from another Ada compiler might not. A
29650 combination of testing and analysis (for which the warnings generated
29653 switch can be useful) must be used to ensure that the program is free
29654 of errors. One switch that is useful in this testing is the
29655 @option{^-p (pessimistic elaboration order)^/PESSIMISTIC_ELABORATION_ORDER^}
29658 Normally the binder tries to find an order that has the best chance
29659 of avoiding elaboration problems. However, if this switch is used, the binder
29660 plays a devil's advocate role, and tries to choose the order that
29661 has the best chance of failing. If your program works even with this
29662 switch, then it has a better chance of being error free, but this is still
29665 For an example of this approach in action, consider the C-tests (executable
29666 tests) from the ACVC suite. If these are compiled and run with the default
29667 treatment, then all but one of them succeed without generating any error
29668 diagnostics from the binder. However, there is one test that fails, and
29669 this is not surprising, because the whole point of this test is to ensure
29670 that the compiler can handle cases where it is impossible to determine
29671 a correct order statically, and it checks that an exception is indeed
29672 raised at run time.
29674 This one test must be compiled and run using the
29676 switch, and then it passes. Alternatively, the entire suite can
29677 be run using this switch. It is never wrong to run with the dynamic
29678 elaboration switch if your code is correct, and we assume that the
29679 C-tests are indeed correct (it is less efficient, but efficiency is
29680 not a factor in running the ACVC tests.)
29682 @node Elaboration for Access-to-Subprogram Values
29683 @section Elaboration for Access-to-Subprogram Values
29684 @cindex Access-to-subprogram
29687 Access-to-subprogram types (introduced in Ada 95) complicate
29688 the handling of elaboration. The trouble is that it becomes
29689 impossible to tell at compile time which procedure
29690 is being called. This means that it is not possible for the binder
29691 to analyze the elaboration requirements in this case.
29693 If at the point at which the access value is created
29694 (i.e., the evaluation of @code{P'Access} for a subprogram @code{P}),
29695 the body of the subprogram is
29696 known to have been elaborated, then the access value is safe, and its use
29697 does not require a check. This may be achieved by appropriate arrangement
29698 of the order of declarations if the subprogram is in the current unit,
29699 or, if the subprogram is in another unit, by using pragma
29700 @code{Pure}, @code{Preelaborate}, or @code{Elaborate_Body}
29701 on the referenced unit.
29703 If the referenced body is not known to have been elaborated at the point
29704 the access value is created, then any use of the access value must do a
29705 dynamic check, and this dynamic check will fail and raise a
29706 @code{Program_Error} exception if the body has not been elaborated yet.
29707 GNAT will generate the necessary checks, and in addition, if the
29709 switch is set, will generate warnings that such checks are required.
29711 The use of dynamic dispatching for tagged types similarly generates
29712 a requirement for dynamic checks, and premature calls to any primitive
29713 operation of a tagged type before the body of the operation has been
29714 elaborated, will result in the raising of @code{Program_Error}.
29716 @node Summary of Procedures for Elaboration Control
29717 @section Summary of Procedures for Elaboration Control
29718 @cindex Elaboration control
29721 First, compile your program with the default options, using none of
29722 the special elaboration control switches. If the binder successfully
29723 binds your program, then you can be confident that, apart from issues
29724 raised by the use of access-to-subprogram types and dynamic dispatching,
29725 the program is free of elaboration errors. If it is important that the
29726 program be portable, then use the
29728 switch to generate warnings about missing @code{Elaborate} or
29729 @code{Elaborate_All} pragmas, and supply the missing pragmas.
29731 If the program fails to bind using the default static elaboration
29732 handling, then you can fix the program to eliminate the binder
29733 message, or recompile the entire program with the
29734 @option{-gnatE} switch to generate dynamic elaboration checks,
29735 and, if you are sure there really are no elaboration problems,
29736 use a global pragma @code{Suppress (Elaboration_Check)}.
29738 @node Other Elaboration Order Considerations
29739 @section Other Elaboration Order Considerations
29741 This section has been entirely concerned with the issue of finding a valid
29742 elaboration order, as defined by the Ada Reference Manual. In a case
29743 where several elaboration orders are valid, the task is to find one
29744 of the possible valid elaboration orders (and the static model in GNAT
29745 will ensure that this is achieved).
29747 The purpose of the elaboration rules in the Ada Reference Manual is to
29748 make sure that no entity is accessed before it has been elaborated. For
29749 a subprogram, this means that the spec and body must have been elaborated
29750 before the subprogram is called. For an object, this means that the object
29751 must have been elaborated before its value is read or written. A violation
29752 of either of these two requirements is an access before elaboration order,
29753 and this section has been all about avoiding such errors.
29755 In the case where more than one order of elaboration is possible, in the
29756 sense that access before elaboration errors are avoided, then any one of
29757 the orders is ``correct'' in the sense that it meets the requirements of
29758 the Ada Reference Manual, and no such error occurs.
29760 However, it may be the case for a given program, that there are
29761 constraints on the order of elaboration that come not from consideration
29762 of avoiding elaboration errors, but rather from extra-lingual logic
29763 requirements. Consider this example:
29765 @smallexample @c ada
29766 with Init_Constants;
29767 package Constants is
29772 package Init_Constants is
29773 procedure P; -- require a body
29774 end Init_Constants;
29777 package body Init_Constants is
29778 procedure P is begin null; end;
29782 end Init_Constants;
29786 Z : Integer := Constants.X + Constants.Y;
29790 with Text_IO; use Text_IO;
29793 Put_Line (Calc.Z'Img);
29798 In this example, there is more than one valid order of elaboration. For
29799 example both the following are correct orders:
29802 Init_Constants spec
29805 Init_Constants body
29810 Init_Constants spec
29811 Init_Constants body
29818 There is no language rule to prefer one or the other, both are correct
29819 from an order of elaboration point of view. But the programmatic effects
29820 of the two orders are very different. In the first, the elaboration routine
29821 of @code{Calc} initializes @code{Z} to zero, and then the main program
29822 runs with this value of zero. But in the second order, the elaboration
29823 routine of @code{Calc} runs after the body of Init_Constants has set
29824 @code{X} and @code{Y} and thus @code{Z} is set to 7 before @code{Main}
29827 One could perhaps by applying pretty clever non-artificial intelligence
29828 to the situation guess that it is more likely that the second order of
29829 elaboration is the one desired, but there is no formal linguistic reason
29830 to prefer one over the other. In fact in this particular case, GNAT will
29831 prefer the second order, because of the rule that bodies are elaborated
29832 as soon as possible, but it's just luck that this is what was wanted
29833 (if indeed the second order was preferred).
29835 If the program cares about the order of elaboration routines in a case like
29836 this, it is important to specify the order required. In this particular
29837 case, that could have been achieved by adding to the spec of Calc:
29839 @smallexample @c ada
29840 pragma Elaborate_All (Constants);
29844 which requires that the body (if any) and spec of @code{Constants},
29845 as well as the body and spec of any unit @code{with}'ed by
29846 @code{Constants} be elaborated before @code{Calc} is elaborated.
29848 Clearly no automatic method can always guess which alternative you require,
29849 and if you are working with legacy code that had constraints of this kind
29850 which were not properly specified by adding @code{Elaborate} or
29851 @code{Elaborate_All} pragmas, then indeed it is possible that two different
29852 compilers can choose different orders.
29854 However, GNAT does attempt to diagnose the common situation where there
29855 are uninitialized variables in the visible part of a package spec, and the
29856 corresponding package body has an elaboration block that directly or
29857 indirectly initialized one or more of these variables. This is the situation
29858 in which a pragma Elaborate_Body is usually desirable, and GNAT will generate
29859 a warning that suggests this addition if it detects this situation.
29861 The @code{gnatbind}
29862 @option{^-p^/PESSIMISTIC_ELABORATION^} switch may be useful in smoking
29863 out problems. This switch causes bodies to be elaborated as late as possible
29864 instead of as early as possible. In the example above, it would have forced
29865 the choice of the first elaboration order. If you get different results
29866 when using this switch, and particularly if one set of results is right,
29867 and one is wrong as far as you are concerned, it shows that you have some
29868 missing @code{Elaborate} pragmas. For the example above, we have the
29872 gnatmake -f -q main
29875 gnatmake -f -q main -bargs -p
29881 It is of course quite unlikely that both these results are correct, so
29882 it is up to you in a case like this to investigate the source of the
29883 difference, by looking at the two elaboration orders that are chosen,
29884 and figuring out which is correct, and then adding the necessary
29885 @code{Elaborate} or @code{Elaborate_All} pragmas to ensure the desired order.
29889 @c *******************************
29890 @node Conditional Compilation
29891 @appendix Conditional Compilation
29892 @c *******************************
29893 @cindex Conditional compilation
29896 It is often necessary to arrange for a single source program
29897 to serve multiple purposes, where it is compiled in different
29898 ways to achieve these different goals. Some examples of the
29899 need for this feature are
29902 @item Adapting a program to a different hardware environment
29903 @item Adapting a program to a different target architecture
29904 @item Turning debugging features on and off
29905 @item Arranging for a program to compile with different compilers
29909 In C, or C++, the typical approach would be to use the preprocessor
29910 that is defined as part of the language. The Ada language does not
29911 contain such a feature. This is not an oversight, but rather a very
29912 deliberate design decision, based on the experience that overuse of
29913 the preprocessing features in C and C++ can result in programs that
29914 are extremely difficult to maintain. For example, if we have ten
29915 switches that can be on or off, this means that there are a thousand
29916 separate programs, any one of which might not even be syntactically
29917 correct, and even if syntactically correct, the resulting program
29918 might not work correctly. Testing all combinations can quickly become
29921 Nevertheless, the need to tailor programs certainly exists, and in
29922 this Appendix we will discuss how this can
29923 be achieved using Ada in general, and GNAT in particular.
29926 * Use of Boolean Constants::
29927 * Debugging - A Special Case::
29928 * Conditionalizing Declarations::
29929 * Use of Alternative Implementations::
29933 @node Use of Boolean Constants
29934 @section Use of Boolean Constants
29937 In the case where the difference is simply which code
29938 sequence is executed, the cleanest solution is to use Boolean
29939 constants to control which code is executed.
29941 @smallexample @c ada
29943 FP_Initialize_Required : constant Boolean := True;
29945 if FP_Initialize_Required then
29952 Not only will the code inside the @code{if} statement not be executed if
29953 the constant Boolean is @code{False}, but it will also be completely
29954 deleted from the program.
29955 However, the code is only deleted after the @code{if} statement
29956 has been checked for syntactic and semantic correctness.
29957 (In contrast, with preprocessors the code is deleted before the
29958 compiler ever gets to see it, so it is not checked until the switch
29960 @cindex Preprocessors (contrasted with conditional compilation)
29962 Typically the Boolean constants will be in a separate package,
29965 @smallexample @c ada
29968 FP_Initialize_Required : constant Boolean := True;
29969 Reset_Available : constant Boolean := False;
29976 The @code{Config} package exists in multiple forms for the various targets,
29977 with an appropriate script selecting the version of @code{Config} needed.
29978 Then any other unit requiring conditional compilation can do a @code{with}
29979 of @code{Config} to make the constants visible.
29982 @node Debugging - A Special Case
29983 @section Debugging - A Special Case
29986 A common use of conditional code is to execute statements (for example
29987 dynamic checks, or output of intermediate results) under control of a
29988 debug switch, so that the debugging behavior can be turned on and off.
29989 This can be done using a Boolean constant to control whether the code
29992 @smallexample @c ada
29995 Put_Line ("got to the first stage!");
30003 @smallexample @c ada
30005 if Debugging and then Temperature > 999.0 then
30006 raise Temperature_Crazy;
30012 Since this is a common case, there are special features to deal with
30013 this in a convenient manner. For the case of tests, Ada 2005 has added
30014 a pragma @code{Assert} that can be used for such tests. This pragma is modeled
30015 @cindex pragma @code{Assert}
30016 on the @code{Assert} pragma that has always been available in GNAT, so this
30017 feature may be used with GNAT even if you are not using Ada 2005 features.
30018 The use of pragma @code{Assert} is described in
30019 @ref{Pragma Assert,,, gnat_rm, GNAT Reference Manual}, but as an
30020 example, the last test could be written:
30022 @smallexample @c ada
30023 pragma Assert (Temperature <= 999.0, "Temperature Crazy");
30029 @smallexample @c ada
30030 pragma Assert (Temperature <= 999.0);
30034 In both cases, if assertions are active and the temperature is excessive,
30035 the exception @code{Assert_Failure} will be raised, with the given string in
30036 the first case or a string indicating the location of the pragma in the second
30037 case used as the exception message.
30039 You can turn assertions on and off by using the @code{Assertion_Policy}
30041 @cindex pragma @code{Assertion_Policy}
30042 This is an Ada 2005 pragma which is implemented in all modes by
30043 GNAT, but only in the latest versions of GNAT which include Ada 2005
30044 capability. Alternatively, you can use the @option{-gnata} switch
30045 @cindex @option{-gnata} switch
30046 to enable assertions from the command line (this is recognized by all versions
30049 For the example above with the @code{Put_Line}, the GNAT-specific pragma
30050 @code{Debug} can be used:
30051 @cindex pragma @code{Debug}
30053 @smallexample @c ada
30054 pragma Debug (Put_Line ("got to the first stage!"));
30058 If debug pragmas are enabled, the argument, which must be of the form of
30059 a procedure call, is executed (in this case, @code{Put_Line} will be called).
30060 Only one call can be present, but of course a special debugging procedure
30061 containing any code you like can be included in the program and then
30062 called in a pragma @code{Debug} argument as needed.
30064 One advantage of pragma @code{Debug} over the @code{if Debugging then}
30065 construct is that pragma @code{Debug} can appear in declarative contexts,
30066 such as at the very beginning of a procedure, before local declarations have
30069 Debug pragmas are enabled using either the @option{-gnata} switch that also
30070 controls assertions, or with a separate Debug_Policy pragma.
30071 @cindex pragma @code{Debug_Policy}
30072 The latter pragma is new in the Ada 2005 versions of GNAT (but it can be used
30073 in Ada 95 and Ada 83 programs as well), and is analogous to
30074 pragma @code{Assertion_Policy} to control assertions.
30076 @code{Assertion_Policy} and @code{Debug_Policy} are configuration pragmas,
30077 and thus they can appear in @file{gnat.adc} if you are not using a
30078 project file, or in the file designated to contain configuration pragmas
30080 They then apply to all subsequent compilations. In practice the use of
30081 the @option{-gnata} switch is often the most convenient method of controlling
30082 the status of these pragmas.
30084 Note that a pragma is not a statement, so in contexts where a statement
30085 sequence is required, you can't just write a pragma on its own. You have
30086 to add a @code{null} statement.
30088 @smallexample @c ada
30091 @dots{} -- some statements
30093 pragma Assert (Num_Cases < 10);
30100 @node Conditionalizing Declarations
30101 @section Conditionalizing Declarations
30104 In some cases, it may be necessary to conditionalize declarations to meet
30105 different requirements. For example we might want a bit string whose length
30106 is set to meet some hardware message requirement.
30108 In some cases, it may be possible to do this using declare blocks controlled
30109 by conditional constants:
30111 @smallexample @c ada
30113 if Small_Machine then
30115 X : Bit_String (1 .. 10);
30121 X : Large_Bit_String (1 .. 1000);
30130 Note that in this approach, both declarations are analyzed by the
30131 compiler so this can only be used where both declarations are legal,
30132 even though one of them will not be used.
30134 Another approach is to define integer constants, e.g.@: @code{Bits_Per_Word}, or
30135 Boolean constants, e.g.@: @code{Little_Endian}, and then write declarations
30136 that are parameterized by these constants. For example
30138 @smallexample @c ada
30141 Field1 at 0 range Boolean'Pos (Little_Endian) * 10 .. Bits_Per_Word;
30147 If @code{Bits_Per_Word} is set to 32, this generates either
30149 @smallexample @c ada
30152 Field1 at 0 range 0 .. 32;
30158 for the big endian case, or
30160 @smallexample @c ada
30163 Field1 at 0 range 10 .. 32;
30169 for the little endian case. Since a powerful subset of Ada expression
30170 notation is usable for creating static constants, clever use of this
30171 feature can often solve quite difficult problems in conditionalizing
30172 compilation (note incidentally that in Ada 95, the little endian
30173 constant was introduced as @code{System.Default_Bit_Order}, so you do not
30174 need to define this one yourself).
30177 @node Use of Alternative Implementations
30178 @section Use of Alternative Implementations
30181 In some cases, none of the approaches described above are adequate. This
30182 can occur for example if the set of declarations required is radically
30183 different for two different configurations.
30185 In this situation, the official Ada way of dealing with conditionalizing
30186 such code is to write separate units for the different cases. As long as
30187 this does not result in excessive duplication of code, this can be done
30188 without creating maintenance problems. The approach is to share common
30189 code as far as possible, and then isolate the code and declarations
30190 that are different. Subunits are often a convenient method for breaking
30191 out a piece of a unit that is to be conditionalized, with separate files
30192 for different versions of the subunit for different targets, where the
30193 build script selects the right one to give to the compiler.
30194 @cindex Subunits (and conditional compilation)
30196 As an example, consider a situation where a new feature in Ada 2005
30197 allows something to be done in a really nice way. But your code must be able
30198 to compile with an Ada 95 compiler. Conceptually you want to say:
30200 @smallexample @c ada
30203 @dots{} neat Ada 2005 code
30205 @dots{} not quite as neat Ada 95 code
30211 where @code{Ada_2005} is a Boolean constant.
30213 But this won't work when @code{Ada_2005} is set to @code{False},
30214 since the @code{then} clause will be illegal for an Ada 95 compiler.
30215 (Recall that although such unreachable code would eventually be deleted
30216 by the compiler, it still needs to be legal. If it uses features
30217 introduced in Ada 2005, it will be illegal in Ada 95.)
30219 So instead we write
30221 @smallexample @c ada
30222 procedure Insert is separate;
30226 Then we have two files for the subunit @code{Insert}, with the two sets of
30228 If the package containing this is called @code{File_Queries}, then we might
30232 @item @file{file_queries-insert-2005.adb}
30233 @item @file{file_queries-insert-95.adb}
30237 and the build script renames the appropriate file to
30240 file_queries-insert.adb
30244 and then carries out the compilation.
30246 This can also be done with project files' naming schemes. For example:
30248 @smallexample @c project
30249 For Body ("File_Queries.Insert") use "file_queries-insert-2005.ada";
30253 Note also that with project files it is desirable to use a different extension
30254 than @file{ads} / @file{adb} for alternative versions. Otherwise a naming
30255 conflict may arise through another commonly used feature: to declare as part
30256 of the project a set of directories containing all the sources obeying the
30257 default naming scheme.
30259 The use of alternative units is certainly feasible in all situations,
30260 and for example the Ada part of the GNAT run-time is conditionalized
30261 based on the target architecture using this approach. As a specific example,
30262 consider the implementation of the AST feature in VMS. There is one
30270 which is the same for all architectures, and three bodies:
30274 used for all non-VMS operating systems
30275 @item s-asthan-vms-alpha.adb
30276 used for VMS on the Alpha
30277 @item s-asthan-vms-ia64.adb
30278 used for VMS on the ia64
30282 The dummy version @file{s-asthan.adb} simply raises exceptions noting that
30283 this operating system feature is not available, and the two remaining
30284 versions interface with the corresponding versions of VMS to provide
30285 VMS-compatible AST handling. The GNAT build script knows the architecture
30286 and operating system, and automatically selects the right version,
30287 renaming it if necessary to @file{s-asthan.adb} before the run-time build.
30289 Another style for arranging alternative implementations is through Ada's
30290 access-to-subprogram facility.
30291 In case some functionality is to be conditionally included,
30292 you can declare an access-to-procedure variable @code{Ref} that is initialized
30293 to designate a ``do nothing'' procedure, and then invoke @code{Ref.all}
30295 In some library package, set @code{Ref} to @code{Proc'Access} for some
30296 procedure @code{Proc} that performs the relevant processing.
30297 The initialization only occurs if the library package is included in the
30299 The same idea can also be implemented using tagged types and dispatching
30303 @node Preprocessing
30304 @section Preprocessing
30305 @cindex Preprocessing
30308 Although it is quite possible to conditionalize code without the use of
30309 C-style preprocessing, as described earlier in this section, it is
30310 nevertheless convenient in some cases to use the C approach. Moreover,
30311 older Ada compilers have often provided some preprocessing capability,
30312 so legacy code may depend on this approach, even though it is not
30315 To accommodate such use, GNAT provides a preprocessor (modeled to a large
30316 extent on the various preprocessors that have been used
30317 with legacy code on other compilers, to enable easier transition).
30319 The preprocessor may be used in two separate modes. It can be used quite
30320 separately from the compiler, to generate a separate output source file
30321 that is then fed to the compiler as a separate step. This is the
30322 @code{gnatprep} utility, whose use is fully described in
30323 @ref{Preprocessing Using gnatprep}.
30324 @cindex @code{gnatprep}
30326 The preprocessing language allows such constructs as
30330 #if DEBUG or PRIORITY > 4 then
30331 bunch of declarations
30333 completely different bunch of declarations
30339 The values of the symbols @code{DEBUG} and @code{PRIORITY} can be
30340 defined either on the command line or in a separate file.
30342 The other way of running the preprocessor is even closer to the C style and
30343 often more convenient. In this approach the preprocessing is integrated into
30344 the compilation process. The compiler is fed the preprocessor input which
30345 includes @code{#if} lines etc, and then the compiler carries out the
30346 preprocessing internally and processes the resulting output.
30347 For more details on this approach, see @ref{Integrated Preprocessing}.
30350 @c *******************************
30351 @node Inline Assembler
30352 @appendix Inline Assembler
30353 @c *******************************
30356 If you need to write low-level software that interacts directly
30357 with the hardware, Ada provides two ways to incorporate assembly
30358 language code into your program. First, you can import and invoke
30359 external routines written in assembly language, an Ada feature fully
30360 supported by GNAT@. However, for small sections of code it may be simpler
30361 or more efficient to include assembly language statements directly
30362 in your Ada source program, using the facilities of the implementation-defined
30363 package @code{System.Machine_Code}, which incorporates the gcc
30364 Inline Assembler. The Inline Assembler approach offers a number of advantages,
30365 including the following:
30368 @item No need to use non-Ada tools
30369 @item Consistent interface over different targets
30370 @item Automatic usage of the proper calling conventions
30371 @item Access to Ada constants and variables
30372 @item Definition of intrinsic routines
30373 @item Possibility of inlining a subprogram comprising assembler code
30374 @item Code optimizer can take Inline Assembler code into account
30377 This chapter presents a series of examples to show you how to use
30378 the Inline Assembler. Although it focuses on the Intel x86,
30379 the general approach applies also to other processors.
30380 It is assumed that you are familiar with Ada
30381 and with assembly language programming.
30384 * Basic Assembler Syntax::
30385 * A Simple Example of Inline Assembler::
30386 * Output Variables in Inline Assembler::
30387 * Input Variables in Inline Assembler::
30388 * Inlining Inline Assembler Code::
30389 * Other Asm Functionality::
30392 @c ---------------------------------------------------------------------------
30393 @node Basic Assembler Syntax
30394 @section Basic Assembler Syntax
30397 The assembler used by GNAT and gcc is based not on the Intel assembly
30398 language, but rather on a language that descends from the AT&T Unix
30399 assembler @emph{as} (and which is often referred to as ``AT&T syntax'').
30400 The following table summarizes the main features of @emph{as} syntax
30401 and points out the differences from the Intel conventions.
30402 See the gcc @emph{as} and @emph{gas} (an @emph{as} macro
30403 pre-processor) documentation for further information.
30406 @item Register names
30407 gcc / @emph{as}: Prefix with ``%''; for example @code{%eax}
30409 Intel: No extra punctuation; for example @code{eax}
30411 @item Immediate operand
30412 gcc / @emph{as}: Prefix with ``$''; for example @code{$4}
30414 Intel: No extra punctuation; for example @code{4}
30417 gcc / @emph{as}: Prefix with ``$''; for example @code{$loc}
30419 Intel: No extra punctuation; for example @code{loc}
30421 @item Memory contents
30422 gcc / @emph{as}: No extra punctuation; for example @code{loc}
30424 Intel: Square brackets; for example @code{[loc]}
30426 @item Register contents
30427 gcc / @emph{as}: Parentheses; for example @code{(%eax)}
30429 Intel: Square brackets; for example @code{[eax]}
30431 @item Hexadecimal numbers
30432 gcc / @emph{as}: Leading ``0x'' (C language syntax); for example @code{0xA0}
30434 Intel: Trailing ``h''; for example @code{A0h}
30437 gcc / @emph{as}: Explicit in op code; for example @code{movw} to move
30440 Intel: Implicit, deduced by assembler; for example @code{mov}
30442 @item Instruction repetition
30443 gcc / @emph{as}: Split into two lines; for example
30449 Intel: Keep on one line; for example @code{rep stosl}
30451 @item Order of operands
30452 gcc / @emph{as}: Source first; for example @code{movw $4, %eax}
30454 Intel: Destination first; for example @code{mov eax, 4}
30457 @c ---------------------------------------------------------------------------
30458 @node A Simple Example of Inline Assembler
30459 @section A Simple Example of Inline Assembler
30462 The following example will generate a single assembly language statement,
30463 @code{nop}, which does nothing. Despite its lack of run-time effect,
30464 the example will be useful in illustrating the basics of
30465 the Inline Assembler facility.
30467 @smallexample @c ada
30469 with System.Machine_Code; use System.Machine_Code;
30470 procedure Nothing is
30477 @code{Asm} is a procedure declared in package @code{System.Machine_Code};
30478 here it takes one parameter, a @emph{template string} that must be a static
30479 expression and that will form the generated instruction.
30480 @code{Asm} may be regarded as a compile-time procedure that parses
30481 the template string and additional parameters (none here),
30482 from which it generates a sequence of assembly language instructions.
30484 The examples in this chapter will illustrate several of the forms
30485 for invoking @code{Asm}; a complete specification of the syntax
30486 is found in @ref{Machine Code Insertions,,, gnat_rm, GNAT Reference
30489 Under the standard GNAT conventions, the @code{Nothing} procedure
30490 should be in a file named @file{nothing.adb}.
30491 You can build the executable in the usual way:
30495 However, the interesting aspect of this example is not its run-time behavior
30496 but rather the generated assembly code.
30497 To see this output, invoke the compiler as follows:
30499 gcc -c -S -fomit-frame-pointer -gnatp @file{nothing.adb}
30501 where the options are:
30505 compile only (no bind or link)
30507 generate assembler listing
30508 @item -fomit-frame-pointer
30509 do not set up separate stack frames
30511 do not add runtime checks
30514 This gives a human-readable assembler version of the code. The resulting
30515 file will have the same name as the Ada source file, but with a @code{.s}
30516 extension. In our example, the file @file{nothing.s} has the following
30521 .file "nothing.adb"
30523 ___gnu_compiled_ada:
30526 .globl __ada_nothing
30538 The assembly code you included is clearly indicated by
30539 the compiler, between the @code{#APP} and @code{#NO_APP}
30540 delimiters. The character before the 'APP' and 'NOAPP'
30541 can differ on different targets. For example, GNU/Linux uses '#APP' while
30542 on NT you will see '/APP'.
30544 If you make a mistake in your assembler code (such as using the
30545 wrong size modifier, or using a wrong operand for the instruction) GNAT
30546 will report this error in a temporary file, which will be deleted when
30547 the compilation is finished. Generating an assembler file will help
30548 in such cases, since you can assemble this file separately using the
30549 @emph{as} assembler that comes with gcc.
30551 Assembling the file using the command
30554 as @file{nothing.s}
30557 will give you error messages whose lines correspond to the assembler
30558 input file, so you can easily find and correct any mistakes you made.
30559 If there are no errors, @emph{as} will generate an object file
30560 @file{nothing.out}.
30562 @c ---------------------------------------------------------------------------
30563 @node Output Variables in Inline Assembler
30564 @section Output Variables in Inline Assembler
30567 The examples in this section, showing how to access the processor flags,
30568 illustrate how to specify the destination operands for assembly language
30571 @smallexample @c ada
30573 with Interfaces; use Interfaces;
30574 with Ada.Text_IO; use Ada.Text_IO;
30575 with System.Machine_Code; use System.Machine_Code;
30576 procedure Get_Flags is
30577 Flags : Unsigned_32;
30580 Asm ("pushfl" & LF & HT & -- push flags on stack
30581 "popl %%eax" & LF & HT & -- load eax with flags
30582 "movl %%eax, %0", -- store flags in variable
30583 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
30584 Put_Line ("Flags register:" & Flags'Img);
30589 In order to have a nicely aligned assembly listing, we have separated
30590 multiple assembler statements in the Asm template string with linefeed
30591 (ASCII.LF) and horizontal tab (ASCII.HT) characters.
30592 The resulting section of the assembly output file is:
30599 movl %eax, -40(%ebp)
30604 It would have been legal to write the Asm invocation as:
30607 Asm ("pushfl popl %%eax movl %%eax, %0")
30610 but in the generated assembler file, this would come out as:
30614 pushfl popl %eax movl %eax, -40(%ebp)
30618 which is not so convenient for the human reader.
30620 We use Ada comments
30621 at the end of each line to explain what the assembler instructions
30622 actually do. This is a useful convention.
30624 When writing Inline Assembler instructions, you need to precede each register
30625 and variable name with a percent sign. Since the assembler already requires
30626 a percent sign at the beginning of a register name, you need two consecutive
30627 percent signs for such names in the Asm template string, thus @code{%%eax}.
30628 In the generated assembly code, one of the percent signs will be stripped off.
30630 Names such as @code{%0}, @code{%1}, @code{%2}, etc., denote input or output
30631 variables: operands you later define using @code{Input} or @code{Output}
30632 parameters to @code{Asm}.
30633 An output variable is illustrated in
30634 the third statement in the Asm template string:
30638 The intent is to store the contents of the eax register in a variable that can
30639 be accessed in Ada. Simply writing @code{movl %%eax, Flags} would not
30640 necessarily work, since the compiler might optimize by using a register
30641 to hold Flags, and the expansion of the @code{movl} instruction would not be
30642 aware of this optimization. The solution is not to store the result directly
30643 but rather to advise the compiler to choose the correct operand form;
30644 that is the purpose of the @code{%0} output variable.
30646 Information about the output variable is supplied in the @code{Outputs}
30647 parameter to @code{Asm}:
30649 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
30652 The output is defined by the @code{Asm_Output} attribute of the target type;
30653 the general format is
30655 Type'Asm_Output (constraint_string, variable_name)
30658 The constraint string directs the compiler how
30659 to store/access the associated variable. In the example
30661 Unsigned_32'Asm_Output ("=m", Flags);
30663 the @code{"m"} (memory) constraint tells the compiler that the variable
30664 @code{Flags} should be stored in a memory variable, thus preventing
30665 the optimizer from keeping it in a register. In contrast,
30667 Unsigned_32'Asm_Output ("=r", Flags);
30669 uses the @code{"r"} (register) constraint, telling the compiler to
30670 store the variable in a register.
30672 If the constraint is preceded by the equal character (@strong{=}), it tells
30673 the compiler that the variable will be used to store data into it.
30675 In the @code{Get_Flags} example, we used the @code{"g"} (global) constraint,
30676 allowing the optimizer to choose whatever it deems best.
30678 There are a fairly large number of constraints, but the ones that are
30679 most useful (for the Intel x86 processor) are the following:
30685 global (i.e.@: can be stored anywhere)
30703 use one of eax, ebx, ecx or edx
30705 use one of eax, ebx, ecx, edx, esi or edi
30708 The full set of constraints is described in the gcc and @emph{as}
30709 documentation; note that it is possible to combine certain constraints
30710 in one constraint string.
30712 You specify the association of an output variable with an assembler operand
30713 through the @code{%}@emph{n} notation, where @emph{n} is a non-negative
30715 @smallexample @c ada
30717 Asm ("pushfl" & LF & HT & -- push flags on stack
30718 "popl %%eax" & LF & HT & -- load eax with flags
30719 "movl %%eax, %0", -- store flags in variable
30720 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
30724 @code{%0} will be replaced in the expanded code by the appropriate operand,
30726 the compiler decided for the @code{Flags} variable.
30728 In general, you may have any number of output variables:
30731 Count the operands starting at 0; thus @code{%0}, @code{%1}, etc.
30733 Specify the @code{Outputs} parameter as a parenthesized comma-separated list
30734 of @code{Asm_Output} attributes
30738 @smallexample @c ada
30740 Asm ("movl %%eax, %0" & LF & HT &
30741 "movl %%ebx, %1" & LF & HT &
30743 Outputs => (Unsigned_32'Asm_Output ("=g", Var_A), -- %0 = Var_A
30744 Unsigned_32'Asm_Output ("=g", Var_B), -- %1 = Var_B
30745 Unsigned_32'Asm_Output ("=g", Var_C))); -- %2 = Var_C
30749 where @code{Var_A}, @code{Var_B}, and @code{Var_C} are variables
30750 in the Ada program.
30752 As a variation on the @code{Get_Flags} example, we can use the constraints
30753 string to direct the compiler to store the eax register into the @code{Flags}
30754 variable, instead of including the store instruction explicitly in the
30755 @code{Asm} template string:
30757 @smallexample @c ada
30759 with Interfaces; use Interfaces;
30760 with Ada.Text_IO; use Ada.Text_IO;
30761 with System.Machine_Code; use System.Machine_Code;
30762 procedure Get_Flags_2 is
30763 Flags : Unsigned_32;
30766 Asm ("pushfl" & LF & HT & -- push flags on stack
30767 "popl %%eax", -- save flags in eax
30768 Outputs => Unsigned_32'Asm_Output ("=a", Flags));
30769 Put_Line ("Flags register:" & Flags'Img);
30775 The @code{"a"} constraint tells the compiler that the @code{Flags}
30776 variable will come from the eax register. Here is the resulting code:
30784 movl %eax,-40(%ebp)
30789 The compiler generated the store of eax into Flags after
30790 expanding the assembler code.
30792 Actually, there was no need to pop the flags into the eax register;
30793 more simply, we could just pop the flags directly into the program variable:
30795 @smallexample @c ada
30797 with Interfaces; use Interfaces;
30798 with Ada.Text_IO; use Ada.Text_IO;
30799 with System.Machine_Code; use System.Machine_Code;
30800 procedure Get_Flags_3 is
30801 Flags : Unsigned_32;
30804 Asm ("pushfl" & LF & HT & -- push flags on stack
30805 "pop %0", -- save flags in Flags
30806 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
30807 Put_Line ("Flags register:" & Flags'Img);
30812 @c ---------------------------------------------------------------------------
30813 @node Input Variables in Inline Assembler
30814 @section Input Variables in Inline Assembler
30817 The example in this section illustrates how to specify the source operands
30818 for assembly language statements.
30819 The program simply increments its input value by 1:
30821 @smallexample @c ada
30823 with Interfaces; use Interfaces;
30824 with Ada.Text_IO; use Ada.Text_IO;
30825 with System.Machine_Code; use System.Machine_Code;
30826 procedure Increment is
30828 function Incr (Value : Unsigned_32) return Unsigned_32 is
30829 Result : Unsigned_32;
30832 Inputs => Unsigned_32'Asm_Input ("a", Value),
30833 Outputs => Unsigned_32'Asm_Output ("=a", Result));
30837 Value : Unsigned_32;
30841 Put_Line ("Value before is" & Value'Img);
30842 Value := Incr (Value);
30843 Put_Line ("Value after is" & Value'Img);
30848 The @code{Outputs} parameter to @code{Asm} specifies
30849 that the result will be in the eax register and that it is to be stored
30850 in the @code{Result} variable.
30852 The @code{Inputs} parameter looks much like the @code{Outputs} parameter,
30853 but with an @code{Asm_Input} attribute.
30854 The @code{"="} constraint, indicating an output value, is not present.
30856 You can have multiple input variables, in the same way that you can have more
30857 than one output variable.
30859 The parameter count (%0, %1) etc, now starts at the first input
30860 statement, and continues with the output statements.
30861 When both parameters use the same variable, the
30862 compiler will treat them as the same %n operand, which is the case here.
30864 Just as the @code{Outputs} parameter causes the register to be stored into the
30865 target variable after execution of the assembler statements, so does the
30866 @code{Inputs} parameter cause its variable to be loaded into the register
30867 before execution of the assembler statements.
30869 Thus the effect of the @code{Asm} invocation is:
30871 @item load the 32-bit value of @code{Value} into eax
30872 @item execute the @code{incl %eax} instruction
30873 @item store the contents of eax into the @code{Result} variable
30876 The resulting assembler file (with @option{-O2} optimization) contains:
30879 _increment__incr.1:
30892 @c ---------------------------------------------------------------------------
30893 @node Inlining Inline Assembler Code
30894 @section Inlining Inline Assembler Code
30897 For a short subprogram such as the @code{Incr} function in the previous
30898 section, the overhead of the call and return (creating / deleting the stack
30899 frame) can be significant, compared to the amount of code in the subprogram
30900 body. A solution is to apply Ada's @code{Inline} pragma to the subprogram,
30901 which directs the compiler to expand invocations of the subprogram at the
30902 point(s) of call, instead of setting up a stack frame for out-of-line calls.
30903 Here is the resulting program:
30905 @smallexample @c ada
30907 with Interfaces; use Interfaces;
30908 with Ada.Text_IO; use Ada.Text_IO;
30909 with System.Machine_Code; use System.Machine_Code;
30910 procedure Increment_2 is
30912 function Incr (Value : Unsigned_32) return Unsigned_32 is
30913 Result : Unsigned_32;
30916 Inputs => Unsigned_32'Asm_Input ("a", Value),
30917 Outputs => Unsigned_32'Asm_Output ("=a", Result));
30920 pragma Inline (Increment);
30922 Value : Unsigned_32;
30926 Put_Line ("Value before is" & Value'Img);
30927 Value := Increment (Value);
30928 Put_Line ("Value after is" & Value'Img);
30933 Compile the program with both optimization (@option{-O2}) and inlining
30934 (@option{-gnatn}) enabled.
30936 The @code{Incr} function is still compiled as usual, but at the
30937 point in @code{Increment} where our function used to be called:
30942 call _increment__incr.1
30947 the code for the function body directly appears:
30960 thus saving the overhead of stack frame setup and an out-of-line call.
30962 @c ---------------------------------------------------------------------------
30963 @node Other Asm Functionality
30964 @section Other @code{Asm} Functionality
30967 This section describes two important parameters to the @code{Asm}
30968 procedure: @code{Clobber}, which identifies register usage;
30969 and @code{Volatile}, which inhibits unwanted optimizations.
30972 * The Clobber Parameter::
30973 * The Volatile Parameter::
30976 @c ---------------------------------------------------------------------------
30977 @node The Clobber Parameter
30978 @subsection The @code{Clobber} Parameter
30981 One of the dangers of intermixing assembly language and a compiled language
30982 such as Ada is that the compiler needs to be aware of which registers are
30983 being used by the assembly code. In some cases, such as the earlier examples,
30984 the constraint string is sufficient to indicate register usage (e.g.,
30986 the eax register). But more generally, the compiler needs an explicit
30987 identification of the registers that are used by the Inline Assembly
30990 Using a register that the compiler doesn't know about
30991 could be a side effect of an instruction (like @code{mull}
30992 storing its result in both eax and edx).
30993 It can also arise from explicit register usage in your
30994 assembly code; for example:
30997 Asm ("movl %0, %%ebx" & LF & HT &
30999 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
31000 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out));
31004 where the compiler (since it does not analyze the @code{Asm} template string)
31005 does not know you are using the ebx register.
31007 In such cases you need to supply the @code{Clobber} parameter to @code{Asm},
31008 to identify the registers that will be used by your assembly code:
31012 Asm ("movl %0, %%ebx" & LF & HT &
31014 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
31015 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
31020 The Clobber parameter is a static string expression specifying the
31021 register(s) you are using. Note that register names are @emph{not} prefixed
31022 by a percent sign. Also, if more than one register is used then their names
31023 are separated by commas; e.g., @code{"eax, ebx"}
31025 The @code{Clobber} parameter has several additional uses:
31027 @item Use ``register'' name @code{cc} to indicate that flags might have changed
31028 @item Use ``register'' name @code{memory} if you changed a memory location
31031 @c ---------------------------------------------------------------------------
31032 @node The Volatile Parameter
31033 @subsection The @code{Volatile} Parameter
31034 @cindex Volatile parameter
31037 Compiler optimizations in the presence of Inline Assembler may sometimes have
31038 unwanted effects. For example, when an @code{Asm} invocation with an input
31039 variable is inside a loop, the compiler might move the loading of the input
31040 variable outside the loop, regarding it as a one-time initialization.
31042 If this effect is not desired, you can disable such optimizations by setting
31043 the @code{Volatile} parameter to @code{True}; for example:
31045 @smallexample @c ada
31047 Asm ("movl %0, %%ebx" & LF & HT &
31049 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
31050 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
31056 By default, @code{Volatile} is set to @code{False} unless there is no
31057 @code{Outputs} parameter.
31059 Although setting @code{Volatile} to @code{True} prevents unwanted
31060 optimizations, it will also disable other optimizations that might be
31061 important for efficiency. In general, you should set @code{Volatile}
31062 to @code{True} only if the compiler's optimizations have created
31064 @c END OF INLINE ASSEMBLER CHAPTER
31065 @c ===============================
31067 @c ***********************************
31068 @c * Compatibility and Porting Guide *
31069 @c ***********************************
31070 @node Compatibility and Porting Guide
31071 @appendix Compatibility and Porting Guide
31074 This chapter describes the compatibility issues that may arise between
31075 GNAT and other Ada compilation systems (including those for Ada 83),
31076 and shows how GNAT can expedite porting
31077 applications developed in other Ada environments.
31080 * Compatibility with Ada 83::
31081 * Compatibility between Ada 95 and Ada 2005::
31082 * Implementation-dependent characteristics::
31083 * Compatibility with Other Ada Systems::
31084 * Representation Clauses::
31086 @c Brief section is only in non-VMS version
31087 @c Full chapter is in VMS version
31088 * Compatibility with HP Ada 83::
31091 * Transitioning to 64-Bit GNAT for OpenVMS::
31095 @node Compatibility with Ada 83
31096 @section Compatibility with Ada 83
31097 @cindex Compatibility (between Ada 83 and Ada 95 / Ada 2005)
31100 Ada 95 and Ada 2005 are highly upwards compatible with Ada 83. In
31101 particular, the design intention was that the difficulties associated
31102 with moving from Ada 83 to Ada 95 or Ada 2005 should be no greater than those
31103 that occur when moving from one Ada 83 system to another.
31105 However, there are a number of points at which there are minor
31106 incompatibilities. The @cite{Ada 95 Annotated Reference Manual} contains
31107 full details of these issues,
31108 and should be consulted for a complete treatment.
31110 following subsections treat the most likely issues to be encountered.
31113 * Legal Ada 83 programs that are illegal in Ada 95::
31114 * More deterministic semantics::
31115 * Changed semantics::
31116 * Other language compatibility issues::
31119 @node Legal Ada 83 programs that are illegal in Ada 95
31120 @subsection Legal Ada 83 programs that are illegal in Ada 95
31122 Some legal Ada 83 programs are illegal (i.e., they will fail to compile) in
31123 Ada 95 and thus also in Ada 2005:
31126 @item Character literals
31127 Some uses of character literals are ambiguous. Since Ada 95 has introduced
31128 @code{Wide_Character} as a new predefined character type, some uses of
31129 character literals that were legal in Ada 83 are illegal in Ada 95.
31131 @smallexample @c ada
31132 for Char in 'A' .. 'Z' loop @dots{} end loop;
31136 The problem is that @code{'A'} and @code{'Z'} could be from either
31137 @code{Character} or @code{Wide_Character}. The simplest correction
31138 is to make the type explicit; e.g.:
31139 @smallexample @c ada
31140 for Char in Character range 'A' .. 'Z' loop @dots{} end loop;
31143 @item New reserved words
31144 The identifiers @code{abstract}, @code{aliased}, @code{protected},
31145 @code{requeue}, @code{tagged}, and @code{until} are reserved in Ada 95.
31146 Existing Ada 83 code using any of these identifiers must be edited to
31147 use some alternative name.
31149 @item Freezing rules
31150 The rules in Ada 95 are slightly different with regard to the point at
31151 which entities are frozen, and representation pragmas and clauses are
31152 not permitted past the freeze point. This shows up most typically in
31153 the form of an error message complaining that a representation item
31154 appears too late, and the appropriate corrective action is to move
31155 the item nearer to the declaration of the entity to which it refers.
31157 A particular case is that representation pragmas
31160 extended HP Ada 83 compatibility pragmas such as @code{Export_Procedure})
31162 cannot be applied to a subprogram body. If necessary, a separate subprogram
31163 declaration must be introduced to which the pragma can be applied.
31165 @item Optional bodies for library packages
31166 In Ada 83, a package that did not require a package body was nevertheless
31167 allowed to have one. This lead to certain surprises in compiling large
31168 systems (situations in which the body could be unexpectedly ignored by the
31169 binder). In Ada 95, if a package does not require a body then it is not
31170 permitted to have a body. To fix this problem, simply remove a redundant
31171 body if it is empty, or, if it is non-empty, introduce a dummy declaration
31172 into the spec that makes the body required. One approach is to add a private
31173 part to the package declaration (if necessary), and define a parameterless
31174 procedure called @code{Requires_Body}, which must then be given a dummy
31175 procedure body in the package body, which then becomes required.
31176 Another approach (assuming that this does not introduce elaboration
31177 circularities) is to add an @code{Elaborate_Body} pragma to the package spec,
31178 since one effect of this pragma is to require the presence of a package body.
31180 @item @code{Numeric_Error} is now the same as @code{Constraint_Error}
31181 In Ada 95, the exception @code{Numeric_Error} is a renaming of
31182 @code{Constraint_Error}.
31183 This means that it is illegal to have separate exception handlers for
31184 the two exceptions. The fix is simply to remove the handler for the
31185 @code{Numeric_Error} case (since even in Ada 83, a compiler was free to raise
31186 @code{Constraint_Error} in place of @code{Numeric_Error} in all cases).
31188 @item Indefinite subtypes in generics
31189 In Ada 83, it was permissible to pass an indefinite type (e.g.@: @code{String})
31190 as the actual for a generic formal private type, but then the instantiation
31191 would be illegal if there were any instances of declarations of variables
31192 of this type in the generic body. In Ada 95, to avoid this clear violation
31193 of the methodological principle known as the ``contract model'',
31194 the generic declaration explicitly indicates whether
31195 or not such instantiations are permitted. If a generic formal parameter
31196 has explicit unknown discriminants, indicated by using @code{(<>)} after the
31197 type name, then it can be instantiated with indefinite types, but no
31198 stand-alone variables can be declared of this type. Any attempt to declare
31199 such a variable will result in an illegality at the time the generic is
31200 declared. If the @code{(<>)} notation is not used, then it is illegal
31201 to instantiate the generic with an indefinite type.
31202 This is the potential incompatibility issue when porting Ada 83 code to Ada 95.
31203 It will show up as a compile time error, and
31204 the fix is usually simply to add the @code{(<>)} to the generic declaration.
31207 @node More deterministic semantics
31208 @subsection More deterministic semantics
31212 Conversions from real types to integer types round away from 0. In Ada 83
31213 the conversion Integer(2.5) could deliver either 2 or 3 as its value. This
31214 implementation freedom was intended to support unbiased rounding in
31215 statistical applications, but in practice it interfered with portability.
31216 In Ada 95 the conversion semantics are unambiguous, and rounding away from 0
31217 is required. Numeric code may be affected by this change in semantics.
31218 Note, though, that this issue is no worse than already existed in Ada 83
31219 when porting code from one vendor to another.
31222 The Real-Time Annex introduces a set of policies that define the behavior of
31223 features that were implementation dependent in Ada 83, such as the order in
31224 which open select branches are executed.
31227 @node Changed semantics
31228 @subsection Changed semantics
31231 The worst kind of incompatibility is one where a program that is legal in
31232 Ada 83 is also legal in Ada 95 but can have an effect in Ada 95 that was not
31233 possible in Ada 83. Fortunately this is extremely rare, but the one
31234 situation that you should be alert to is the change in the predefined type
31235 @code{Character} from 7-bit ASCII to 8-bit Latin-1.
31238 @item Range of type @code{Character}
31239 The range of @code{Standard.Character} is now the full 256 characters
31240 of Latin-1, whereas in most Ada 83 implementations it was restricted
31241 to 128 characters. Although some of the effects of
31242 this change will be manifest in compile-time rejection of legal
31243 Ada 83 programs it is possible for a working Ada 83 program to have
31244 a different effect in Ada 95, one that was not permitted in Ada 83.
31245 As an example, the expression
31246 @code{Character'Pos(Character'Last)} returned @code{127} in Ada 83 and now
31247 delivers @code{255} as its value.
31248 In general, you should look at the logic of any
31249 character-processing Ada 83 program and see whether it needs to be adapted
31250 to work correctly with Latin-1. Note that the predefined Ada 95 API has a
31251 character handling package that may be relevant if code needs to be adapted
31252 to account for the additional Latin-1 elements.
31253 The desirable fix is to
31254 modify the program to accommodate the full character set, but in some cases
31255 it may be convenient to define a subtype or derived type of Character that
31256 covers only the restricted range.
31260 @node Other language compatibility issues
31261 @subsection Other language compatibility issues
31264 @item @option{-gnat83} switch
31265 All implementations of GNAT provide a switch that causes GNAT to operate
31266 in Ada 83 mode. In this mode, some but not all compatibility problems
31267 of the type described above are handled automatically. For example, the
31268 new reserved words introduced in Ada 95 and Ada 2005 are treated simply
31269 as identifiers as in Ada 83.
31271 in practice, it is usually advisable to make the necessary modifications
31272 to the program to remove the need for using this switch.
31273 See @ref{Compiling Different Versions of Ada}.
31275 @item Support for removed Ada 83 pragmas and attributes
31276 A number of pragmas and attributes from Ada 83 were removed from Ada 95,
31277 generally because they were replaced by other mechanisms. Ada 95 and Ada 2005
31278 compilers are allowed, but not required, to implement these missing
31279 elements. In contrast with some other compilers, GNAT implements all
31280 such pragmas and attributes, eliminating this compatibility concern. These
31281 include @code{pragma Interface} and the floating point type attributes
31282 (@code{Emax}, @code{Mantissa}, etc.), among other items.
31286 @node Compatibility between Ada 95 and Ada 2005
31287 @section Compatibility between Ada 95 and Ada 2005
31288 @cindex Compatibility between Ada 95 and Ada 2005
31291 Although Ada 2005 was designed to be upwards compatible with Ada 95, there are
31292 a number of incompatibilities. Several are enumerated below;
31293 for a complete description please see the
31294 Annotated Ada 2005 Reference Manual, or section 9.1.1 in
31295 @cite{Rationale for Ada 2005}.
31298 @item New reserved words.
31299 The words @code{interface}, @code{overriding} and @code{synchronized} are
31300 reserved in Ada 2005.
31301 A pre-Ada 2005 program that uses any of these as an identifier will be
31304 @item New declarations in predefined packages.
31305 A number of packages in the predefined environment contain new declarations:
31306 @code{Ada.Exceptions}, @code{Ada.Real_Time}, @code{Ada.Strings},
31307 @code{Ada.Strings.Fixed}, @code{Ada.Strings.Bounded},
31308 @code{Ada.Strings.Unbounded}, @code{Ada.Strings.Wide_Fixed},
31309 @code{Ada.Strings.Wide_Bounded}, @code{Ada.Strings.Wide_Unbounded},
31310 @code{Ada.Tags}, @code{Ada.Text_IO}, and @code{Interfaces.C}.
31311 If an Ada 95 program does a @code{with} and @code{use} of any of these
31312 packages, the new declarations may cause name clashes.
31314 @item Access parameters.
31315 A nondispatching subprogram with an access parameter cannot be renamed
31316 as a dispatching operation. This was permitted in Ada 95.
31318 @item Access types, discriminants, and constraints.
31319 Rule changes in this area have led to some incompatibilities; for example,
31320 constrained subtypes of some access types are not permitted in Ada 2005.
31322 @item Aggregates for limited types.
31323 The allowance of aggregates for limited types in Ada 2005 raises the
31324 possibility of ambiguities in legal Ada 95 programs, since additional types
31325 now need to be considered in expression resolution.
31327 @item Fixed-point multiplication and division.
31328 Certain expressions involving ``*'' or ``/'' for a fixed-point type, which
31329 were legal in Ada 95 and invoked the predefined versions of these operations,
31331 The ambiguity may be resolved either by applying a type conversion to the
31332 expression, or by explicitly invoking the operation from package
31335 @item Return-by-reference types.
31336 The Ada 95 return-by-reference mechanism has been removed. Instead, the user
31337 can declare a function returning a value from an anonymous access type.
31341 @node Implementation-dependent characteristics
31342 @section Implementation-dependent characteristics
31344 Although the Ada language defines the semantics of each construct as
31345 precisely as practical, in some situations (for example for reasons of
31346 efficiency, or where the effect is heavily dependent on the host or target
31347 platform) the implementation is allowed some freedom. In porting Ada 83
31348 code to GNAT, you need to be aware of whether / how the existing code
31349 exercised such implementation dependencies. Such characteristics fall into
31350 several categories, and GNAT offers specific support in assisting the
31351 transition from certain Ada 83 compilers.
31354 * Implementation-defined pragmas::
31355 * Implementation-defined attributes::
31357 * Elaboration order::
31358 * Target-specific aspects::
31361 @node Implementation-defined pragmas
31362 @subsection Implementation-defined pragmas
31365 Ada compilers are allowed to supplement the language-defined pragmas, and
31366 these are a potential source of non-portability. All GNAT-defined pragmas
31367 are described in @ref{Implementation Defined Pragmas,,, gnat_rm, GNAT
31368 Reference Manual}, and these include several that are specifically
31369 intended to correspond to other vendors' Ada 83 pragmas.
31370 For migrating from VADS, the pragma @code{Use_VADS_Size} may be useful.
31371 For compatibility with HP Ada 83, GNAT supplies the pragmas
31372 @code{Extend_System}, @code{Ident}, @code{Inline_Generic},
31373 @code{Interface_Name}, @code{Passive}, @code{Suppress_All},
31374 and @code{Volatile}.
31375 Other relevant pragmas include @code{External} and @code{Link_With}.
31376 Some vendor-specific
31377 Ada 83 pragmas (@code{Share_Generic}, @code{Subtitle}, and @code{Title}) are
31379 avoiding compiler rejection of units that contain such pragmas; they are not
31380 relevant in a GNAT context and hence are not otherwise implemented.
31382 @node Implementation-defined attributes
31383 @subsection Implementation-defined attributes
31385 Analogous to pragmas, the set of attributes may be extended by an
31386 implementation. All GNAT-defined attributes are described in
31387 @ref{Implementation Defined Attributes,,, gnat_rm, GNAT Reference
31388 Manual}, and these include several that are specifically intended
31389 to correspond to other vendors' Ada 83 attributes. For migrating from VADS,
31390 the attribute @code{VADS_Size} may be useful. For compatibility with HP
31391 Ada 83, GNAT supplies the attributes @code{Bit}, @code{Machine_Size} and
31395 @subsection Libraries
31397 Vendors may supply libraries to supplement the standard Ada API. If Ada 83
31398 code uses vendor-specific libraries then there are several ways to manage
31399 this in Ada 95 or Ada 2005:
31402 If the source code for the libraries (specs and bodies) are
31403 available, then the libraries can be migrated in the same way as the
31406 If the source code for the specs but not the bodies are
31407 available, then you can reimplement the bodies.
31409 Some features introduced by Ada 95 obviate the need for library support. For
31410 example most Ada 83 vendors supplied a package for unsigned integers. The
31411 Ada 95 modular type feature is the preferred way to handle this need, so
31412 instead of migrating or reimplementing the unsigned integer package it may
31413 be preferable to retrofit the application using modular types.
31416 @node Elaboration order
31417 @subsection Elaboration order
31419 The implementation can choose any elaboration order consistent with the unit
31420 dependency relationship. This freedom means that some orders can result in
31421 Program_Error being raised due to an ``Access Before Elaboration'': an attempt
31422 to invoke a subprogram its body has been elaborated, or to instantiate a
31423 generic before the generic body has been elaborated. By default GNAT
31424 attempts to choose a safe order (one that will not encounter access before
31425 elaboration problems) by implicitly inserting @code{Elaborate} or
31426 @code{Elaborate_All} pragmas where
31427 needed. However, this can lead to the creation of elaboration circularities
31428 and a resulting rejection of the program by gnatbind. This issue is
31429 thoroughly described in @ref{Elaboration Order Handling in GNAT}.
31430 In brief, there are several
31431 ways to deal with this situation:
31435 Modify the program to eliminate the circularities, e.g.@: by moving
31436 elaboration-time code into explicitly-invoked procedures
31438 Constrain the elaboration order by including explicit @code{Elaborate_Body} or
31439 @code{Elaborate} pragmas, and then inhibit the generation of implicit
31440 @code{Elaborate_All}
31441 pragmas either globally (as an effect of the @option{-gnatE} switch) or locally
31442 (by selectively suppressing elaboration checks via pragma
31443 @code{Suppress(Elaboration_Check)} when it is safe to do so).
31446 @node Target-specific aspects
31447 @subsection Target-specific aspects
31449 Low-level applications need to deal with machine addresses, data
31450 representations, interfacing with assembler code, and similar issues. If
31451 such an Ada 83 application is being ported to different target hardware (for
31452 example where the byte endianness has changed) then you will need to
31453 carefully examine the program logic; the porting effort will heavily depend
31454 on the robustness of the original design. Moreover, Ada 95 (and thus
31455 Ada 2005) are sometimes
31456 incompatible with typical Ada 83 compiler practices regarding implicit
31457 packing, the meaning of the Size attribute, and the size of access values.
31458 GNAT's approach to these issues is described in @ref{Representation Clauses}.
31460 @node Compatibility with Other Ada Systems
31461 @section Compatibility with Other Ada Systems
31464 If programs avoid the use of implementation dependent and
31465 implementation defined features, as documented in the @cite{Ada
31466 Reference Manual}, there should be a high degree of portability between
31467 GNAT and other Ada systems. The following are specific items which
31468 have proved troublesome in moving Ada 95 programs from GNAT to other Ada 95
31469 compilers, but do not affect porting code to GNAT@.
31470 (As of @value{NOW}, GNAT is the only compiler available for Ada 2005;
31471 the following issues may or may not arise for Ada 2005 programs
31472 when other compilers appear.)
31475 @item Ada 83 Pragmas and Attributes
31476 Ada 95 compilers are allowed, but not required, to implement the missing
31477 Ada 83 pragmas and attributes that are no longer defined in Ada 95.
31478 GNAT implements all such pragmas and attributes, eliminating this as
31479 a compatibility concern, but some other Ada 95 compilers reject these
31480 pragmas and attributes.
31482 @item Specialized Needs Annexes
31483 GNAT implements the full set of special needs annexes. At the
31484 current time, it is the only Ada 95 compiler to do so. This means that
31485 programs making use of these features may not be portable to other Ada
31486 95 compilation systems.
31488 @item Representation Clauses
31489 Some other Ada 95 compilers implement only the minimal set of
31490 representation clauses required by the Ada 95 reference manual. GNAT goes
31491 far beyond this minimal set, as described in the next section.
31494 @node Representation Clauses
31495 @section Representation Clauses
31498 The Ada 83 reference manual was quite vague in describing both the minimal
31499 required implementation of representation clauses, and also their precise
31500 effects. Ada 95 (and thus also Ada 2005) are much more explicit, but the
31501 minimal set of capabilities required is still quite limited.
31503 GNAT implements the full required set of capabilities in
31504 Ada 95 and Ada 2005, but also goes much further, and in particular
31505 an effort has been made to be compatible with existing Ada 83 usage to the
31506 greatest extent possible.
31508 A few cases exist in which Ada 83 compiler behavior is incompatible with
31509 the requirements in Ada 95 (and thus also Ada 2005). These are instances of
31510 intentional or accidental dependence on specific implementation dependent
31511 characteristics of these Ada 83 compilers. The following is a list of
31512 the cases most likely to arise in existing Ada 83 code.
31515 @item Implicit Packing
31516 Some Ada 83 compilers allowed a Size specification to cause implicit
31517 packing of an array or record. This could cause expensive implicit
31518 conversions for change of representation in the presence of derived
31519 types, and the Ada design intends to avoid this possibility.
31520 Subsequent AI's were issued to make it clear that such implicit
31521 change of representation in response to a Size clause is inadvisable,
31522 and this recommendation is represented explicitly in the Ada 95 (and Ada 2005)
31523 Reference Manuals as implementation advice that is followed by GNAT@.
31524 The problem will show up as an error
31525 message rejecting the size clause. The fix is simply to provide
31526 the explicit pragma @code{Pack}, or for more fine tuned control, provide
31527 a Component_Size clause.
31529 @item Meaning of Size Attribute
31530 The Size attribute in Ada 95 (and Ada 2005) for discrete types is defined as
31531 the minimal number of bits required to hold values of the type. For example,
31532 on a 32-bit machine, the size of @code{Natural} will typically be 31 and not
31533 32 (since no sign bit is required). Some Ada 83 compilers gave 31, and
31534 some 32 in this situation. This problem will usually show up as a compile
31535 time error, but not always. It is a good idea to check all uses of the
31536 'Size attribute when porting Ada 83 code. The GNAT specific attribute
31537 Object_Size can provide a useful way of duplicating the behavior of
31538 some Ada 83 compiler systems.
31540 @item Size of Access Types
31541 A common assumption in Ada 83 code is that an access type is in fact a pointer,
31542 and that therefore it will be the same size as a System.Address value. This
31543 assumption is true for GNAT in most cases with one exception. For the case of
31544 a pointer to an unconstrained array type (where the bounds may vary from one
31545 value of the access type to another), the default is to use a ``fat pointer'',
31546 which is represented as two separate pointers, one to the bounds, and one to
31547 the array. This representation has a number of advantages, including improved
31548 efficiency. However, it may cause some difficulties in porting existing Ada 83
31549 code which makes the assumption that, for example, pointers fit in 32 bits on
31550 a machine with 32-bit addressing.
31552 To get around this problem, GNAT also permits the use of ``thin pointers'' for
31553 access types in this case (where the designated type is an unconstrained array
31554 type). These thin pointers are indeed the same size as a System.Address value.
31555 To specify a thin pointer, use a size clause for the type, for example:
31557 @smallexample @c ada
31558 type X is access all String;
31559 for X'Size use Standard'Address_Size;
31563 which will cause the type X to be represented using a single pointer.
31564 When using this representation, the bounds are right behind the array.
31565 This representation is slightly less efficient, and does not allow quite
31566 such flexibility in the use of foreign pointers or in using the
31567 Unrestricted_Access attribute to create pointers to non-aliased objects.
31568 But for any standard portable use of the access type it will work in
31569 a functionally correct manner and allow porting of existing code.
31570 Note that another way of forcing a thin pointer representation
31571 is to use a component size clause for the element size in an array,
31572 or a record representation clause for an access field in a record.
31576 @c This brief section is only in the non-VMS version
31577 @c The complete chapter on HP Ada is in the VMS version
31578 @node Compatibility with HP Ada 83
31579 @section Compatibility with HP Ada 83
31582 The VMS version of GNAT fully implements all the pragmas and attributes
31583 provided by HP Ada 83, as well as providing the standard HP Ada 83
31584 libraries, including Starlet. In addition, data layouts and parameter
31585 passing conventions are highly compatible. This means that porting
31586 existing HP Ada 83 code to GNAT in VMS systems should be easier than
31587 most other porting efforts. The following are some of the most
31588 significant differences between GNAT and HP Ada 83.
31591 @item Default floating-point representation
31592 In GNAT, the default floating-point format is IEEE, whereas in HP Ada 83,
31593 it is VMS format. GNAT does implement the necessary pragmas
31594 (Long_Float, Float_Representation) for changing this default.
31597 The package System in GNAT exactly corresponds to the definition in the
31598 Ada 95 reference manual, which means that it excludes many of the
31599 HP Ada 83 extensions. However, a separate package Aux_DEC is provided
31600 that contains the additional definitions, and a special pragma,
31601 Extend_System allows this package to be treated transparently as an
31602 extension of package System.
31605 The definitions provided by Aux_DEC are exactly compatible with those
31606 in the HP Ada 83 version of System, with one exception.
31607 HP Ada provides the following declarations:
31609 @smallexample @c ada
31610 TO_ADDRESS (INTEGER)
31611 TO_ADDRESS (UNSIGNED_LONGWORD)
31612 TO_ADDRESS (@i{universal_integer})
31616 The version of TO_ADDRESS taking a @i{universal integer} argument is in fact
31617 an extension to Ada 83 not strictly compatible with the reference manual.
31618 In GNAT, we are constrained to be exactly compatible with the standard,
31619 and this means we cannot provide this capability. In HP Ada 83, the
31620 point of this definition is to deal with a call like:
31622 @smallexample @c ada
31623 TO_ADDRESS (16#12777#);
31627 Normally, according to the Ada 83 standard, one would expect this to be
31628 ambiguous, since it matches both the INTEGER and UNSIGNED_LONGWORD forms
31629 of TO_ADDRESS@. However, in HP Ada 83, there is no ambiguity, since the
31630 definition using @i{universal_integer} takes precedence.
31632 In GNAT, since the version with @i{universal_integer} cannot be supplied, it
31633 is not possible to be 100% compatible. Since there are many programs using
31634 numeric constants for the argument to TO_ADDRESS, the decision in GNAT was
31635 to change the name of the function in the UNSIGNED_LONGWORD case, so the
31636 declarations provided in the GNAT version of AUX_Dec are:
31638 @smallexample @c ada
31639 function To_Address (X : Integer) return Address;
31640 pragma Pure_Function (To_Address);
31642 function To_Address_Long (X : Unsigned_Longword)
31644 pragma Pure_Function (To_Address_Long);
31648 This means that programs using TO_ADDRESS for UNSIGNED_LONGWORD must
31649 change the name to TO_ADDRESS_LONG@.
31651 @item Task_Id values
31652 The Task_Id values assigned will be different in the two systems, and GNAT
31653 does not provide a specified value for the Task_Id of the environment task,
31654 which in GNAT is treated like any other declared task.
31658 For full details on these and other less significant compatibility issues,
31659 see appendix E of the HP publication entitled @cite{HP Ada, Technical
31660 Overview and Comparison on HP Platforms}.
31662 For GNAT running on other than VMS systems, all the HP Ada 83 pragmas and
31663 attributes are recognized, although only a subset of them can sensibly
31664 be implemented. The description of pragmas in @ref{Implementation
31665 Defined Pragmas,,, gnat_rm, GNAT Reference Manual}
31666 indicates whether or not they are applicable to non-VMS systems.
31670 @node Transitioning to 64-Bit GNAT for OpenVMS
31671 @section Transitioning to 64-Bit @value{EDITION} for OpenVMS
31674 This section is meant to assist users of pre-2006 @value{EDITION}
31675 for Alpha OpenVMS who are transitioning to 64-bit @value{EDITION},
31676 the version of the GNAT technology supplied in 2006 and later for
31677 OpenVMS on both Alpha and I64.
31680 * Introduction to transitioning::
31681 * Migration of 32 bit code::
31682 * Taking advantage of 64 bit addressing::
31683 * Technical details::
31686 @node Introduction to transitioning
31687 @subsection Introduction
31690 64-bit @value{EDITION} for Open VMS has been designed to meet
31695 Providing a full conforming implementation of Ada 95 and Ada 2005
31698 Allowing maximum backward compatibility, thus easing migration of existing
31702 Supplying a path for exploiting the full 64-bit address range
31706 Ada's strong typing semantics has made it
31707 impractical to have different 32-bit and 64-bit modes. As soon as
31708 one object could possibly be outside the 32-bit address space, this
31709 would make it necessary for the @code{System.Address} type to be 64 bits.
31710 In particular, this would cause inconsistencies if 32-bit code is
31711 called from 64-bit code that raises an exception.
31713 This issue has been resolved by always using 64-bit addressing
31714 at the system level, but allowing for automatic conversions between
31715 32-bit and 64-bit addresses where required. Thus users who
31716 do not currently require 64-bit addressing capabilities, can
31717 recompile their code with only minimal changes (and indeed
31718 if the code is written in portable Ada, with no assumptions about
31719 the size of the @code{Address} type, then no changes at all are necessary).
31721 this approach provides a simple, gradual upgrade path to future
31722 use of larger memories than available for 32-bit systems.
31723 Also, newly written applications or libraries will by default
31724 be fully compatible with future systems exploiting 64-bit
31725 addressing capabilities.
31727 @ref{Migration of 32 bit code}, will focus on porting applications
31728 that do not require more than 2 GB of
31729 addressable memory. This code will be referred to as
31730 @emph{32-bit code}.
31731 For applications intending to exploit the full 64-bit address space,
31732 @ref{Taking advantage of 64 bit addressing},
31733 will consider further changes that may be required.
31734 Such code will be referred to below as @emph{64-bit code}.
31736 @node Migration of 32 bit code
31737 @subsection Migration of 32-bit code
31742 * Unchecked conversions::
31743 * Predefined constants::
31744 * Interfacing with C::
31745 * Experience with source compatibility::
31748 @node Address types
31749 @subsubsection Address types
31752 To solve the problem of mixing 64-bit and 32-bit addressing,
31753 while maintaining maximum backward compatibility, the following
31754 approach has been taken:
31758 @code{System.Address} always has a size of 64 bits
31761 @code{System.Short_Address} is a 32-bit subtype of @code{System.Address}
31765 Since @code{System.Short_Address} is a subtype of @code{System.Address},
31766 a @code{Short_Address}
31767 may be used where an @code{Address} is required, and vice versa, without
31768 needing explicit type conversions.
31769 By virtue of the Open VMS parameter passing conventions,
31771 and exported subprograms that have 32-bit address parameters are
31772 compatible with those that have 64-bit address parameters.
31773 (See @ref{Making code 64 bit clean} for details.)
31775 The areas that may need attention are those where record types have
31776 been defined that contain components of the type @code{System.Address}, and
31777 where objects of this type are passed to code expecting a record layout with
31780 Different compilers on different platforms cannot be
31781 expected to represent the same type in the same way,
31782 since alignment constraints
31783 and other system-dependent properties affect the compiler's decision.
31784 For that reason, Ada code
31785 generally uses representation clauses to specify the expected
31786 layout where required.
31788 If such a representation clause uses 32 bits for a component having
31789 the type @code{System.Address}, 64-bit @value{EDITION} for OpenVMS
31790 will detect that error and produce a specific diagnostic message.
31791 The developer should then determine whether the representation
31792 should be 64 bits or not and make either of two changes:
31793 change the size to 64 bits and leave the type as @code{System.Address}, or
31794 leave the size as 32 bits and change the type to @code{System.Short_Address}.
31795 Since @code{Short_Address} is a subtype of @code{Address}, no changes are
31796 required in any code setting or accessing the field; the compiler will
31797 automatically perform any needed conversions between address
31801 @subsubsection Access types
31804 By default, objects designated by access values are always
31805 allocated in the 32-bit
31806 address space. Thus legacy code will never contain
31807 any objects that are not addressable with 32-bit addresses, and
31808 the compiler will never raise exceptions as result of mixing
31809 32-bit and 64-bit addresses.
31811 However, the access values themselves are represented in 64 bits, for optimum
31812 performance and future compatibility with 64-bit code. As was
31813 the case with @code{System.Address}, the compiler will give an error message
31814 if an object or record component has a representation clause that
31815 requires the access value to fit in 32 bits. In such a situation,
31816 an explicit size clause for the access type, specifying 32 bits,
31817 will have the desired effect.
31819 General access types (declared with @code{access all}) can never be
31820 32 bits, as values of such types must be able to refer to any object
31821 of the designated type,
31822 including objects residing outside the 32-bit address range.
31823 Existing Ada 83 code will not contain such type definitions,
31824 however, since general access types were introduced in Ada 95.
31826 @node Unchecked conversions
31827 @subsubsection Unchecked conversions
31830 In the case of an @code{Unchecked_Conversion} where the source type is a
31831 64-bit access type or the type @code{System.Address}, and the target
31832 type is a 32-bit type, the compiler will generate a warning.
31833 Even though the generated code will still perform the required
31834 conversions, it is highly recommended in these cases to use
31835 respectively a 32-bit access type or @code{System.Short_Address}
31836 as the source type.
31838 @node Predefined constants
31839 @subsubsection Predefined constants
31842 The following table shows the correspondence between pre-2006 versions of
31843 @value{EDITION} on Alpha OpenVMS (``Old'') and 64-bit @value{EDITION}
31846 @multitable {@code{System.Short_Memory_Size}} {2**32} {2**64}
31847 @item @b{Constant} @tab @b{Old} @tab @b{New}
31848 @item @code{System.Word_Size} @tab 32 @tab 64
31849 @item @code{System.Memory_Size} @tab 2**32 @tab 2**64
31850 @item @code{System.Short_Memory_Size} @tab 2**32 @tab 2**32
31851 @item @code{System.Address_Size} @tab 32 @tab 64
31855 If you need to refer to the specific
31856 memory size of a 32-bit implementation, instead of the
31857 actual memory size, use @code{System.Short_Memory_Size}
31858 rather than @code{System.Memory_Size}.
31859 Similarly, references to @code{System.Address_Size} may need
31860 to be replaced by @code{System.Short_Address'Size}.
31861 The program @command{gnatfind} may be useful for locating
31862 references to the above constants, so that you can verify that they
31865 @node Interfacing with C
31866 @subsubsection Interfacing with C
31869 In order to minimize the impact of the transition to 64-bit addresses on
31870 legacy programs, some fundamental types in the @code{Interfaces.C}
31871 package hierarchy continue to be represented in 32 bits.
31872 These types are: @code{ptrdiff_t}, @code{size_t}, and @code{chars_ptr}.
31873 This eases integration with the default HP C layout choices, for example
31874 as found in the system routines in @code{DECC$SHR.EXE}.
31875 Because of this implementation choice, the type fully compatible with
31876 @code{chars_ptr} is now @code{Short_Address} and not @code{Address}.
31877 Depending on the context the compiler will issue a
31878 warning or an error when type @code{Address} is used, alerting the user to a
31879 potential problem. Otherwise 32-bit programs that use
31880 @code{Interfaces.C} should normally not require code modifications
31882 The other issue arising with C interfacing concerns pragma @code{Convention}.
31883 For VMS 64-bit systems, there is an issue of the appropriate default size
31884 of C convention pointers in the absence of an explicit size clause. The HP
31885 C compiler can choose either 32 or 64 bits depending on compiler options.
31886 GNAT chooses 32-bits rather than 64-bits in the default case where no size
31887 clause is given. This proves a better choice for porting 32-bit legacy
31888 applications. In order to have a 64-bit representation, it is necessary to
31889 specify a size representation clause. For example:
31891 @smallexample @c ada
31892 type int_star is access Interfaces.C.int;
31893 pragma Convention(C, int_star);
31894 for int_star'Size use 64; -- Necessary to get 64 and not 32 bits
31897 @node Experience with source compatibility
31898 @subsubsection Experience with source compatibility
31901 The Security Server and STARLET on I64 provide an interesting ``test case''
31902 for source compatibility issues, since it is in such system code
31903 where assumptions about @code{Address} size might be expected to occur.
31904 Indeed, there were a small number of occasions in the Security Server
31905 file @file{jibdef.ads}
31906 where a representation clause for a record type specified
31907 32 bits for a component of type @code{Address}.
31908 All of these errors were detected by the compiler.
31909 The repair was obvious and immediate; to simply replace @code{Address} by
31910 @code{Short_Address}.
31912 In the case of STARLET, there were several record types that should
31913 have had representation clauses but did not. In these record types
31914 there was an implicit assumption that an @code{Address} value occupied
31916 These compiled without error, but their usage resulted in run-time error
31917 returns from STARLET system calls.
31918 Future GNAT technology enhancements may include a tool that detects and flags
31919 these sorts of potential source code porting problems.
31921 @c ****************************************
31922 @node Taking advantage of 64 bit addressing
31923 @subsection Taking advantage of 64-bit addressing
31926 * Making code 64 bit clean::
31927 * Allocating memory from the 64 bit storage pool::
31928 * Restrictions on use of 64 bit objects::
31929 * Using 64 bit storage pools by default::
31930 * General access types::
31931 * STARLET and other predefined libraries::
31934 @node Making code 64 bit clean
31935 @subsubsection Making code 64-bit clean
31938 In order to prevent problems that may occur when (parts of) a
31939 system start using memory outside the 32-bit address range,
31940 we recommend some additional guidelines:
31944 For imported subprograms that take parameters of the
31945 type @code{System.Address}, ensure that these subprograms can
31946 indeed handle 64-bit addresses. If not, or when in doubt,
31947 change the subprogram declaration to specify
31948 @code{System.Short_Address} instead.
31951 Resolve all warnings related to size mismatches in
31952 unchecked conversions. Failing to do so causes
31953 erroneous execution if the source object is outside
31954 the 32-bit address space.
31957 (optional) Explicitly use the 32-bit storage pool
31958 for access types used in a 32-bit context, or use
31959 generic access types where possible
31960 (@pxref{Restrictions on use of 64 bit objects}).
31964 If these rules are followed, the compiler will automatically insert
31965 any necessary checks to ensure that no addresses or access values
31966 passed to 32-bit code ever refer to objects outside the 32-bit
31968 Any attempt to do this will raise @code{Constraint_Error}.
31970 @node Allocating memory from the 64 bit storage pool
31971 @subsubsection Allocating memory from the 64-bit storage pool
31974 For any access type @code{T} that potentially requires memory allocations
31975 beyond the 32-bit address space,
31976 use the following representation clause:
31978 @smallexample @c ada
31979 for T'Storage_Pool use System.Pool_64;
31982 @node Restrictions on use of 64 bit objects
31983 @subsubsection Restrictions on use of 64-bit objects
31986 Taking the address of an object allocated from a 64-bit storage pool,
31987 and then passing this address to a subprogram expecting
31988 @code{System.Short_Address},
31989 or assigning it to a variable of type @code{Short_Address}, will cause
31990 @code{Constraint_Error} to be raised. In case the code is not 64-bit clean
31991 (@pxref{Making code 64 bit clean}), or checks are suppressed,
31992 no exception is raised and execution
31993 will become erroneous.
31995 @node Using 64 bit storage pools by default
31996 @subsubsection Using 64-bit storage pools by default
31999 In some cases it may be desirable to have the compiler allocate
32000 from 64-bit storage pools by default. This may be the case for
32001 libraries that are 64-bit clean, but may be used in both 32-bit
32002 and 64-bit contexts. For these cases the following configuration
32003 pragma may be specified:
32005 @smallexample @c ada
32006 pragma Pool_64_Default;
32010 Any code compiled in the context of this pragma will by default
32011 use the @code{System.Pool_64} storage pool. This default may be overridden
32012 for a specific access type @code{T} by the representation clause:
32014 @smallexample @c ada
32015 for T'Storage_Pool use System.Pool_32;
32019 Any object whose address may be passed to a subprogram with a
32020 @code{Short_Address} argument, or assigned to a variable of type
32021 @code{Short_Address}, needs to be allocated from this pool.
32023 @node General access types
32024 @subsubsection General access types
32027 Objects designated by access values from a
32028 general access type (declared with @code{access all}) are never allocated
32029 from a 64-bit storage pool. Code that uses general access types will
32030 accept objects allocated in either 32-bit or 64-bit address spaces,
32031 but never allocate objects outside the 32-bit address space.
32032 Using general access types ensures maximum compatibility with both
32033 32-bit and 64-bit code.
32035 @node STARLET and other predefined libraries
32036 @subsubsection STARLET and other predefined libraries
32039 All code that comes as part of GNAT is 64-bit clean, but the
32040 restrictions given in @ref{Restrictions on use of 64 bit objects},
32041 still apply. Look at the package
32042 specs to see in which contexts objects allocated
32043 in 64-bit address space are acceptable.
32045 @node Technical details
32046 @subsection Technical details
32049 64-bit @value{EDITION} for Open VMS takes advantage of the freedom given in the
32050 Ada standard with respect to the type of @code{System.Address}. Previous
32051 versions of GNAT Pro have defined this type as private and implemented it as a
32054 In order to allow defining @code{System.Short_Address} as a proper subtype,
32055 and to match the implicit sign extension in parameter passing,
32056 in 64-bit @value{EDITION} for Open VMS, @code{System.Address} is defined as a
32057 visible (i.e., non-private) integer type.
32058 Standard operations on the type, such as the binary operators ``+'', ``-'',
32059 etc., that take @code{Address} operands and return an @code{Address} result,
32060 have been hidden by declaring these
32061 @code{abstract}, a feature introduced in Ada 95 that helps avoid the potential
32062 ambiguities that would otherwise result from overloading.
32063 (Note that, although @code{Address} is a visible integer type,
32064 good programming practice dictates against exploiting the type's
32065 integer properties such as literals, since this will compromise
32068 Defining @code{Address} as a visible integer type helps achieve
32069 maximum compatibility for existing Ada code,
32070 without sacrificing the capabilities of the 64-bit architecture.
32073 @c ************************************************
32075 @node Microsoft Windows Topics
32076 @appendix Microsoft Windows Topics
32082 This chapter describes topics that are specific to the Microsoft Windows
32083 platforms (NT, 2000, and XP Professional).
32086 * Using GNAT on Windows::
32087 * Using a network installation of GNAT::
32088 * CONSOLE and WINDOWS subsystems::
32089 * Temporary Files::
32090 * Mixed-Language Programming on Windows::
32091 * Windows Calling Conventions::
32092 * Introduction to Dynamic Link Libraries (DLLs)::
32093 * Using DLLs with GNAT::
32094 * Building DLLs with GNAT::
32095 * Building DLLs with GNAT Project files::
32096 * Building DLLs with gnatdll::
32097 * GNAT and Windows Resources::
32098 * Debugging a DLL::
32099 * Setting Stack Size from gnatlink::
32100 * Setting Heap Size from gnatlink::
32103 @node Using GNAT on Windows
32104 @section Using GNAT on Windows
32107 One of the strengths of the GNAT technology is that its tool set
32108 (@command{gcc}, @command{gnatbind}, @command{gnatlink}, @command{gnatmake}, the
32109 @code{gdb} debugger, etc.) is used in the same way regardless of the
32112 On Windows this tool set is complemented by a number of Microsoft-specific
32113 tools that have been provided to facilitate interoperability with Windows
32114 when this is required. With these tools:
32119 You can build applications using the @code{CONSOLE} or @code{WINDOWS}
32123 You can use any Dynamically Linked Library (DLL) in your Ada code (both
32124 relocatable and non-relocatable DLLs are supported).
32127 You can build Ada DLLs for use in other applications. These applications
32128 can be written in a language other than Ada (e.g., C, C++, etc). Again both
32129 relocatable and non-relocatable Ada DLLs are supported.
32132 You can include Windows resources in your Ada application.
32135 You can use or create COM/DCOM objects.
32139 Immediately below are listed all known general GNAT-for-Windows restrictions.
32140 Other restrictions about specific features like Windows Resources and DLLs
32141 are listed in separate sections below.
32146 It is not possible to use @code{GetLastError} and @code{SetLastError}
32147 when tasking, protected records, or exceptions are used. In these
32148 cases, in order to implement Ada semantics, the GNAT run-time system
32149 calls certain Win32 routines that set the last error variable to 0 upon
32150 success. It should be possible to use @code{GetLastError} and
32151 @code{SetLastError} when tasking, protected record, and exception
32152 features are not used, but it is not guaranteed to work.
32155 It is not possible to link against Microsoft libraries except for
32156 import libraries. The library must be built to be compatible with
32157 @file{MSVCRT.LIB} (/MD Microsoft compiler option), @file{LIBC.LIB} and
32158 @file{LIBCMT.LIB} (/ML or /MT Microsoft compiler options) are known to
32159 not be compatible with the GNAT runtime. Even if the library is
32160 compatible with @file{MSVCRT.LIB} it is not guaranteed to work.
32163 When the compilation environment is located on FAT32 drives, users may
32164 experience recompilations of the source files that have not changed if
32165 Daylight Saving Time (DST) state has changed since the last time files
32166 were compiled. NTFS drives do not have this problem.
32169 No components of the GNAT toolset use any entries in the Windows
32170 registry. The only entries that can be created are file associations and
32171 PATH settings, provided the user has chosen to create them at installation
32172 time, as well as some minimal book-keeping information needed to correctly
32173 uninstall or integrate different GNAT products.
32176 @node Using a network installation of GNAT
32177 @section Using a network installation of GNAT
32180 Make sure the system on which GNAT is installed is accessible from the
32181 current machine, i.e., the install location is shared over the network.
32182 Shared resources are accessed on Windows by means of UNC paths, which
32183 have the format @code{\\server\sharename\path}
32185 In order to use such a network installation, simply add the UNC path of the
32186 @file{bin} directory of your GNAT installation in front of your PATH. For
32187 example, if GNAT is installed in @file{\GNAT} directory of a share location
32188 called @file{c-drive} on a machine @file{LOKI}, the following command will
32191 @code{@ @ @ path \\loki\c-drive\gnat\bin;%path%}
32193 Be aware that every compilation using the network installation results in the
32194 transfer of large amounts of data across the network and will likely cause
32195 serious performance penalty.
32197 @node CONSOLE and WINDOWS subsystems
32198 @section CONSOLE and WINDOWS subsystems
32199 @cindex CONSOLE Subsystem
32200 @cindex WINDOWS Subsystem
32204 There are two main subsystems under Windows. The @code{CONSOLE} subsystem
32205 (which is the default subsystem) will always create a console when
32206 launching the application. This is not something desirable when the
32207 application has a Windows GUI. To get rid of this console the
32208 application must be using the @code{WINDOWS} subsystem. To do so
32209 the @option{-mwindows} linker option must be specified.
32212 $ gnatmake winprog -largs -mwindows
32215 @node Temporary Files
32216 @section Temporary Files
32217 @cindex Temporary files
32220 It is possible to control where temporary files gets created by setting
32221 the @env{TMP} environment variable. The file will be created:
32224 @item Under the directory pointed to by the @env{TMP} environment variable if
32225 this directory exists.
32227 @item Under @file{c:\temp}, if the @env{TMP} environment variable is not
32228 set (or not pointing to a directory) and if this directory exists.
32230 @item Under the current working directory otherwise.
32234 This allows you to determine exactly where the temporary
32235 file will be created. This is particularly useful in networked
32236 environments where you may not have write access to some
32239 @node Mixed-Language Programming on Windows
32240 @section Mixed-Language Programming on Windows
32243 Developing pure Ada applications on Windows is no different than on
32244 other GNAT-supported platforms. However, when developing or porting an
32245 application that contains a mix of Ada and C/C++, the choice of your
32246 Windows C/C++ development environment conditions your overall
32247 interoperability strategy.
32249 If you use @command{gcc} to compile the non-Ada part of your application,
32250 there are no Windows-specific restrictions that affect the overall
32251 interoperability with your Ada code. If you plan to use
32252 Microsoft tools (e.g.@: Microsoft Visual C/C++), you should be aware of
32253 the following limitations:
32257 You cannot link your Ada code with an object or library generated with
32258 Microsoft tools if these use the @code{.tls} section (Thread Local
32259 Storage section) since the GNAT linker does not yet support this section.
32262 You cannot link your Ada code with an object or library generated with
32263 Microsoft tools if these use I/O routines other than those provided in
32264 the Microsoft DLL: @code{msvcrt.dll}. This is because the GNAT run time
32265 uses the services of @code{msvcrt.dll} for its I/Os. Use of other I/O
32266 libraries can cause a conflict with @code{msvcrt.dll} services. For
32267 instance Visual C++ I/O stream routines conflict with those in
32272 If you do want to use the Microsoft tools for your non-Ada code and hit one
32273 of the above limitations, you have two choices:
32277 Encapsulate your non-Ada code in a DLL to be linked with your Ada
32278 application. In this case, use the Microsoft or whatever environment to
32279 build the DLL and use GNAT to build your executable
32280 (@pxref{Using DLLs with GNAT}).
32283 Or you can encapsulate your Ada code in a DLL to be linked with the
32284 other part of your application. In this case, use GNAT to build the DLL
32285 (@pxref{Building DLLs with GNAT}) and use the Microsoft or whatever
32286 environment to build your executable.
32289 @node Windows Calling Conventions
32290 @section Windows Calling Conventions
32295 * C Calling Convention::
32296 * Stdcall Calling Convention::
32297 * Win32 Calling Convention::
32298 * DLL Calling Convention::
32302 When a subprogram @code{F} (caller) calls a subprogram @code{G}
32303 (callee), there are several ways to push @code{G}'s parameters on the
32304 stack and there are several possible scenarios to clean up the stack
32305 upon @code{G}'s return. A calling convention is an agreed upon software
32306 protocol whereby the responsibilities between the caller (@code{F}) and
32307 the callee (@code{G}) are clearly defined. Several calling conventions
32308 are available for Windows:
32312 @code{C} (Microsoft defined)
32315 @code{Stdcall} (Microsoft defined)
32318 @code{Win32} (GNAT specific)
32321 @code{DLL} (GNAT specific)
32324 @node C Calling Convention
32325 @subsection @code{C} Calling Convention
32328 This is the default calling convention used when interfacing to C/C++
32329 routines compiled with either @command{gcc} or Microsoft Visual C++.
32331 In the @code{C} calling convention subprogram parameters are pushed on the
32332 stack by the caller from right to left. The caller itself is in charge of
32333 cleaning up the stack after the call. In addition, the name of a routine
32334 with @code{C} calling convention is mangled by adding a leading underscore.
32336 The name to use on the Ada side when importing (or exporting) a routine
32337 with @code{C} calling convention is the name of the routine. For
32338 instance the C function:
32341 int get_val (long);
32345 should be imported from Ada as follows:
32347 @smallexample @c ada
32349 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
32350 pragma Import (C, Get_Val, External_Name => "get_val");
32355 Note that in this particular case the @code{External_Name} parameter could
32356 have been omitted since, when missing, this parameter is taken to be the
32357 name of the Ada entity in lower case. When the @code{Link_Name} parameter
32358 is missing, as in the above example, this parameter is set to be the
32359 @code{External_Name} with a leading underscore.
32361 When importing a variable defined in C, you should always use the @code{C}
32362 calling convention unless the object containing the variable is part of a
32363 DLL (in which case you should use the @code{Stdcall} calling
32364 convention, @pxref{Stdcall Calling Convention}).
32366 @node Stdcall Calling Convention
32367 @subsection @code{Stdcall} Calling Convention
32370 This convention, which was the calling convention used for Pascal
32371 programs, is used by Microsoft for all the routines in the Win32 API for
32372 efficiency reasons. It must be used to import any routine for which this
32373 convention was specified.
32375 In the @code{Stdcall} calling convention subprogram parameters are pushed
32376 on the stack by the caller from right to left. The callee (and not the
32377 caller) is in charge of cleaning the stack on routine exit. In addition,
32378 the name of a routine with @code{Stdcall} calling convention is mangled by
32379 adding a leading underscore (as for the @code{C} calling convention) and a
32380 trailing @code{@@}@code{@var{nn}}, where @var{nn} is the overall size (in
32381 bytes) of the parameters passed to the routine.
32383 The name to use on the Ada side when importing a C routine with a
32384 @code{Stdcall} calling convention is the name of the C routine. The leading
32385 underscore and trailing @code{@@}@code{@var{nn}} are added automatically by
32386 the compiler. For instance the Win32 function:
32389 @b{APIENTRY} int get_val (long);
32393 should be imported from Ada as follows:
32395 @smallexample @c ada
32397 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
32398 pragma Import (Stdcall, Get_Val);
32399 -- On the x86 a long is 4 bytes, so the Link_Name is "_get_val@@4"
32404 As for the @code{C} calling convention, when the @code{External_Name}
32405 parameter is missing, it is taken to be the name of the Ada entity in lower
32406 case. If instead of writing the above import pragma you write:
32408 @smallexample @c ada
32410 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
32411 pragma Import (Stdcall, Get_Val, External_Name => "retrieve_val");
32416 then the imported routine is @code{_retrieve_val@@4}. However, if instead
32417 of specifying the @code{External_Name} parameter you specify the
32418 @code{Link_Name} as in the following example:
32420 @smallexample @c ada
32422 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
32423 pragma Import (Stdcall, Get_Val, Link_Name => "retrieve_val");
32428 then the imported routine is @code{retrieve_val}, that is, there is no
32429 decoration at all. No leading underscore and no Stdcall suffix
32430 @code{@@}@code{@var{nn}}.
32433 This is especially important as in some special cases a DLL's entry
32434 point name lacks a trailing @code{@@}@code{@var{nn}} while the exported
32435 name generated for a call has it.
32438 It is also possible to import variables defined in a DLL by using an
32439 import pragma for a variable. As an example, if a DLL contains a
32440 variable defined as:
32447 then, to access this variable from Ada you should write:
32449 @smallexample @c ada
32451 My_Var : Interfaces.C.int;
32452 pragma Import (Stdcall, My_Var);
32457 Note that to ease building cross-platform bindings this convention
32458 will be handled as a @code{C} calling convention on non-Windows platforms.
32460 @node Win32 Calling Convention
32461 @subsection @code{Win32} Calling Convention
32464 This convention, which is GNAT-specific is fully equivalent to the
32465 @code{Stdcall} calling convention described above.
32467 @node DLL Calling Convention
32468 @subsection @code{DLL} Calling Convention
32471 This convention, which is GNAT-specific is fully equivalent to the
32472 @code{Stdcall} calling convention described above.
32474 @node Introduction to Dynamic Link Libraries (DLLs)
32475 @section Introduction to Dynamic Link Libraries (DLLs)
32479 A Dynamically Linked Library (DLL) is a library that can be shared by
32480 several applications running under Windows. A DLL can contain any number of
32481 routines and variables.
32483 One advantage of DLLs is that you can change and enhance them without
32484 forcing all the applications that depend on them to be relinked or
32485 recompiled. However, you should be aware than all calls to DLL routines are
32486 slower since, as you will understand below, such calls are indirect.
32488 To illustrate the remainder of this section, suppose that an application
32489 wants to use the services of a DLL @file{API.dll}. To use the services
32490 provided by @file{API.dll} you must statically link against the DLL or
32491 an import library which contains a jump table with an entry for each
32492 routine and variable exported by the DLL. In the Microsoft world this
32493 import library is called @file{API.lib}. When using GNAT this import
32494 library is called either @file{libAPI.dll.a}, @file{libapi.dll.a},
32495 @file{libAPI.a} or @file{libapi.a} (names are case insensitive).
32497 After you have linked your application with the DLL or the import library
32498 and you run your application, here is what happens:
32502 Your application is loaded into memory.
32505 The DLL @file{API.dll} is mapped into the address space of your
32506 application. This means that:
32510 The DLL will use the stack of the calling thread.
32513 The DLL will use the virtual address space of the calling process.
32516 The DLL will allocate memory from the virtual address space of the calling
32520 Handles (pointers) can be safely exchanged between routines in the DLL
32521 routines and routines in the application using the DLL.
32525 The entries in the jump table (from the import library @file{libAPI.dll.a}
32526 or @file{API.lib} or automatically created when linking against a DLL)
32527 which is part of your application are initialized with the addresses
32528 of the routines and variables in @file{API.dll}.
32531 If present in @file{API.dll}, routines @code{DllMain} or
32532 @code{DllMainCRTStartup} are invoked. These routines typically contain
32533 the initialization code needed for the well-being of the routines and
32534 variables exported by the DLL.
32538 There is an additional point which is worth mentioning. In the Windows
32539 world there are two kind of DLLs: relocatable and non-relocatable
32540 DLLs. Non-relocatable DLLs can only be loaded at a very specific address
32541 in the target application address space. If the addresses of two
32542 non-relocatable DLLs overlap and these happen to be used by the same
32543 application, a conflict will occur and the application will run
32544 incorrectly. Hence, when possible, it is always preferable to use and
32545 build relocatable DLLs. Both relocatable and non-relocatable DLLs are
32546 supported by GNAT. Note that the @option{-s} linker option (see GNU Linker
32547 User's Guide) removes the debugging symbols from the DLL but the DLL can
32548 still be relocated.
32550 As a side note, an interesting difference between Microsoft DLLs and
32551 Unix shared libraries, is the fact that on most Unix systems all public
32552 routines are exported by default in a Unix shared library, while under
32553 Windows it is possible (but not required) to list exported routines in
32554 a definition file (@pxref{The Definition File}).
32556 @node Using DLLs with GNAT
32557 @section Using DLLs with GNAT
32560 * Creating an Ada Spec for the DLL Services::
32561 * Creating an Import Library::
32565 To use the services of a DLL, say @file{API.dll}, in your Ada application
32570 The Ada spec for the routines and/or variables you want to access in
32571 @file{API.dll}. If not available this Ada spec must be built from the C/C++
32572 header files provided with the DLL.
32575 The import library (@file{libAPI.dll.a} or @file{API.lib}). As previously
32576 mentioned an import library is a statically linked library containing the
32577 import table which will be filled at load time to point to the actual
32578 @file{API.dll} routines. Sometimes you don't have an import library for the
32579 DLL you want to use. The following sections will explain how to build
32580 one. Note that this is optional.
32583 The actual DLL, @file{API.dll}.
32587 Once you have all the above, to compile an Ada application that uses the
32588 services of @file{API.dll} and whose main subprogram is @code{My_Ada_App},
32589 you simply issue the command
32592 $ gnatmake my_ada_app -largs -lAPI
32596 The argument @option{-largs -lAPI} at the end of the @command{gnatmake} command
32597 tells the GNAT linker to look first for a library named @file{API.lib}
32598 (Microsoft-style name) and if not found for a libraries named
32599 @file{libAPI.dll.a}, @file{API.dll.a} or @file{libAPI.a}.
32600 (GNAT-style name). Note that if the Ada package spec for @file{API.dll}
32601 contains the following pragma
32603 @smallexample @c ada
32604 pragma Linker_Options ("-lAPI");
32608 you do not have to add @option{-largs -lAPI} at the end of the
32609 @command{gnatmake} command.
32611 If any one of the items above is missing you will have to create it
32612 yourself. The following sections explain how to do so using as an
32613 example a fictitious DLL called @file{API.dll}.
32615 @node Creating an Ada Spec for the DLL Services
32616 @subsection Creating an Ada Spec for the DLL Services
32619 A DLL typically comes with a C/C++ header file which provides the
32620 definitions of the routines and variables exported by the DLL. The Ada
32621 equivalent of this header file is a package spec that contains definitions
32622 for the imported entities. If the DLL you intend to use does not come with
32623 an Ada spec you have to generate one such spec yourself. For example if
32624 the header file of @file{API.dll} is a file @file{api.h} containing the
32625 following two definitions:
32637 then the equivalent Ada spec could be:
32639 @smallexample @c ada
32642 with Interfaces.C.Strings;
32647 function Get (Str : C.Strings.Chars_Ptr) return C.int;
32650 pragma Import (C, Get);
32651 pragma Import (DLL, Some_Var);
32658 Note that a variable is
32659 @strong{always imported with a Stdcall convention}. A function
32660 can have @code{C} or @code{Stdcall} convention.
32661 (@pxref{Windows Calling Conventions}).
32663 @node Creating an Import Library
32664 @subsection Creating an Import Library
32665 @cindex Import library
32668 * The Definition File::
32669 * GNAT-Style Import Library::
32670 * Microsoft-Style Import Library::
32674 If a Microsoft-style import library @file{API.lib} or a GNAT-style
32675 import library @file{libAPI.dll.a} or @file{libAPI.a} is available
32676 with @file{API.dll} you can skip this section. You can also skip this
32677 section if @file{API.dll} or @file{libAPI.dll} is built with GNU tools
32678 as in this case it is possible to link directly against the
32679 DLL. Otherwise read on.
32681 @node The Definition File
32682 @subsubsection The Definition File
32683 @cindex Definition file
32687 As previously mentioned, and unlike Unix systems, the list of symbols
32688 that are exported from a DLL must be provided explicitly in Windows.
32689 The main goal of a definition file is precisely that: list the symbols
32690 exported by a DLL. A definition file (usually a file with a @code{.def}
32691 suffix) has the following structure:
32696 @r{[}LIBRARY @var{name}@r{]}
32697 @r{[}DESCRIPTION @var{string}@r{]}
32707 @item LIBRARY @var{name}
32708 This section, which is optional, gives the name of the DLL.
32710 @item DESCRIPTION @var{string}
32711 This section, which is optional, gives a description string that will be
32712 embedded in the import library.
32715 This section gives the list of exported symbols (procedures, functions or
32716 variables). For instance in the case of @file{API.dll} the @code{EXPORTS}
32717 section of @file{API.def} looks like:
32731 Note that you must specify the correct suffix (@code{@@}@code{@var{nn}})
32732 (@pxref{Windows Calling Conventions}) for a Stdcall
32733 calling convention function in the exported symbols list.
32736 There can actually be other sections in a definition file, but these
32737 sections are not relevant to the discussion at hand.
32739 @node GNAT-Style Import Library
32740 @subsubsection GNAT-Style Import Library
32743 To create a static import library from @file{API.dll} with the GNAT tools
32744 you should proceed as follows:
32748 Create the definition file @file{API.def} (@pxref{The Definition File}).
32749 For that use the @code{dll2def} tool as follows:
32752 $ dll2def API.dll > API.def
32756 @code{dll2def} is a very simple tool: it takes as input a DLL and prints
32757 to standard output the list of entry points in the DLL. Note that if
32758 some routines in the DLL have the @code{Stdcall} convention
32759 (@pxref{Windows Calling Conventions}) with stripped @code{@@}@var{nn}
32760 suffix then you'll have to edit @file{api.def} to add it, and specify
32761 @option{-k} to @command{gnatdll} when creating the import library.
32764 Here are some hints to find the right @code{@@}@var{nn} suffix.
32768 If you have the Microsoft import library (.lib), it is possible to get
32769 the right symbols by using Microsoft @code{dumpbin} tool (see the
32770 corresponding Microsoft documentation for further details).
32773 $ dumpbin /exports api.lib
32777 If you have a message about a missing symbol at link time the compiler
32778 tells you what symbol is expected. You just have to go back to the
32779 definition file and add the right suffix.
32783 Build the import library @code{libAPI.dll.a}, using @code{gnatdll}
32784 (@pxref{Using gnatdll}) as follows:
32787 $ gnatdll -e API.def -d API.dll
32791 @code{gnatdll} takes as input a definition file @file{API.def} and the
32792 name of the DLL containing the services listed in the definition file
32793 @file{API.dll}. The name of the static import library generated is
32794 computed from the name of the definition file as follows: if the
32795 definition file name is @var{xyz}@code{.def}, the import library name will
32796 be @code{lib}@var{xyz}@code{.a}. Note that in the previous example option
32797 @option{-e} could have been removed because the name of the definition
32798 file (before the ``@code{.def}'' suffix) is the same as the name of the
32799 DLL (@pxref{Using gnatdll} for more information about @code{gnatdll}).
32802 @node Microsoft-Style Import Library
32803 @subsubsection Microsoft-Style Import Library
32806 With GNAT you can either use a GNAT-style or Microsoft-style import
32807 library. A Microsoft import library is needed only if you plan to make an
32808 Ada DLL available to applications developed with Microsoft
32809 tools (@pxref{Mixed-Language Programming on Windows}).
32811 To create a Microsoft-style import library for @file{API.dll} you
32812 should proceed as follows:
32816 Create the definition file @file{API.def} from the DLL. For this use either
32817 the @code{dll2def} tool as described above or the Microsoft @code{dumpbin}
32818 tool (see the corresponding Microsoft documentation for further details).
32821 Build the actual import library using Microsoft's @code{lib} utility:
32824 $ lib -machine:IX86 -def:API.def -out:API.lib
32828 If you use the above command the definition file @file{API.def} must
32829 contain a line giving the name of the DLL:
32836 See the Microsoft documentation for further details about the usage of
32840 @node Building DLLs with GNAT
32841 @section Building DLLs with GNAT
32842 @cindex DLLs, building
32845 This section explain how to build DLLs using the GNAT built-in DLL
32846 support. With the following procedure it is straight forward to build
32847 and use DLLs with GNAT.
32851 @item building object files
32853 The first step is to build all objects files that are to be included
32854 into the DLL. This is done by using the standard @command{gnatmake} tool.
32856 @item building the DLL
32858 To build the DLL you must use @command{gcc}'s @option{-shared}
32859 option. It is quite simple to use this method:
32862 $ gcc -shared -o api.dll obj1.o obj2.o @dots{}
32865 It is important to note that in this case all symbols found in the
32866 object files are automatically exported. It is possible to restrict
32867 the set of symbols to export by passing to @command{gcc} a definition
32868 file, @pxref{The Definition File}. For example:
32871 $ gcc -shared -o api.dll api.def obj1.o obj2.o @dots{}
32874 If you use a definition file you must export the elaboration procedures
32875 for every package that required one. Elaboration procedures are named
32876 using the package name followed by "_E".
32878 @item preparing DLL to be used
32880 For the DLL to be used by client programs the bodies must be hidden
32881 from it and the .ali set with read-only attribute. This is very important
32882 otherwise GNAT will recompile all packages and will not actually use
32883 the code in the DLL. For example:
32887 $ copy *.ads *.ali api.dll apilib
32888 $ attrib +R apilib\*.ali
32893 At this point it is possible to use the DLL by directly linking
32894 against it. Note that you must use the GNAT shared runtime when using
32895 GNAT shared libraries. This is achieved by using @option{-shared} binder's
32899 $ gnatmake main -Iapilib -bargs -shared -largs -Lapilib -lAPI
32902 @node Building DLLs with GNAT Project files
32903 @section Building DLLs with GNAT Project files
32904 @cindex DLLs, building
32907 There is nothing specific to Windows in the build process.
32908 @pxref{Library Projects}.
32911 Due to a system limitation, it is not possible under Windows to create threads
32912 when inside the @code{DllMain} routine which is used for auto-initialization
32913 of shared libraries, so it is not possible to have library level tasks in SALs.
32915 @node Building DLLs with gnatdll
32916 @section Building DLLs with gnatdll
32917 @cindex DLLs, building
32920 * Limitations When Using Ada DLLs from Ada::
32921 * Exporting Ada Entities::
32922 * Ada DLLs and Elaboration::
32923 * Ada DLLs and Finalization::
32924 * Creating a Spec for Ada DLLs::
32925 * Creating the Definition File::
32930 Note that it is preferred to use the built-in GNAT DLL support
32931 (@pxref{Building DLLs with GNAT}) or GNAT Project files
32932 (@pxref{Building DLLs with GNAT Project files}) to build DLLs.
32934 This section explains how to build DLLs containing Ada code using
32935 @code{gnatdll}. These DLLs will be referred to as Ada DLLs in the
32936 remainder of this section.
32938 The steps required to build an Ada DLL that is to be used by Ada as well as
32939 non-Ada applications are as follows:
32943 You need to mark each Ada @i{entity} exported by the DLL with a @code{C} or
32944 @code{Stdcall} calling convention to avoid any Ada name mangling for the
32945 entities exported by the DLL (@pxref{Exporting Ada Entities}). You can
32946 skip this step if you plan to use the Ada DLL only from Ada applications.
32949 Your Ada code must export an initialization routine which calls the routine
32950 @code{adainit} generated by @command{gnatbind} to perform the elaboration of
32951 the Ada code in the DLL (@pxref{Ada DLLs and Elaboration}). The initialization
32952 routine exported by the Ada DLL must be invoked by the clients of the DLL
32953 to initialize the DLL.
32956 When useful, the DLL should also export a finalization routine which calls
32957 routine @code{adafinal} generated by @command{gnatbind} to perform the
32958 finalization of the Ada code in the DLL (@pxref{Ada DLLs and Finalization}).
32959 The finalization routine exported by the Ada DLL must be invoked by the
32960 clients of the DLL when the DLL services are no further needed.
32963 You must provide a spec for the services exported by the Ada DLL in each
32964 of the programming languages to which you plan to make the DLL available.
32967 You must provide a definition file listing the exported entities
32968 (@pxref{The Definition File}).
32971 Finally you must use @code{gnatdll} to produce the DLL and the import
32972 library (@pxref{Using gnatdll}).
32976 Note that a relocatable DLL stripped using the @code{strip}
32977 binutils tool will not be relocatable anymore. To build a DLL without
32978 debug information pass @code{-largs -s} to @code{gnatdll}. This
32979 restriction does not apply to a DLL built using a Library Project.
32980 @pxref{Library Projects}.
32982 @node Limitations When Using Ada DLLs from Ada
32983 @subsection Limitations When Using Ada DLLs from Ada
32986 When using Ada DLLs from Ada applications there is a limitation users
32987 should be aware of. Because on Windows the GNAT run time is not in a DLL of
32988 its own, each Ada DLL includes a part of the GNAT run time. Specifically,
32989 each Ada DLL includes the services of the GNAT run time that are necessary
32990 to the Ada code inside the DLL. As a result, when an Ada program uses an
32991 Ada DLL there are two independent GNAT run times: one in the Ada DLL and
32992 one in the main program.
32994 It is therefore not possible to exchange GNAT run-time objects between the
32995 Ada DLL and the main Ada program. Example of GNAT run-time objects are file
32996 handles (e.g.@: @code{Text_IO.File_Type}), tasks types, protected objects
32999 It is completely safe to exchange plain elementary, array or record types,
33000 Windows object handles, etc.
33002 @node Exporting Ada Entities
33003 @subsection Exporting Ada Entities
33004 @cindex Export table
33007 Building a DLL is a way to encapsulate a set of services usable from any
33008 application. As a result, the Ada entities exported by a DLL should be
33009 exported with the @code{C} or @code{Stdcall} calling conventions to avoid
33010 any Ada name mangling. As an example here is an Ada package
33011 @code{API}, spec and body, exporting two procedures, a function, and a
33014 @smallexample @c ada
33017 with Interfaces.C; use Interfaces;
33019 Count : C.int := 0;
33020 function Factorial (Val : C.int) return C.int;
33022 procedure Initialize_API;
33023 procedure Finalize_API;
33024 -- Initialization & Finalization routines. More in the next section.
33026 pragma Export (C, Initialize_API);
33027 pragma Export (C, Finalize_API);
33028 pragma Export (C, Count);
33029 pragma Export (C, Factorial);
33035 @smallexample @c ada
33038 package body API is
33039 function Factorial (Val : C.int) return C.int is
33042 Count := Count + 1;
33043 for K in 1 .. Val loop
33049 procedure Initialize_API is
33051 pragma Import (C, Adainit);
33054 end Initialize_API;
33056 procedure Finalize_API is
33057 procedure Adafinal;
33058 pragma Import (C, Adafinal);
33068 If the Ada DLL you are building will only be used by Ada applications
33069 you do not have to export Ada entities with a @code{C} or @code{Stdcall}
33070 convention. As an example, the previous package could be written as
33073 @smallexample @c ada
33077 Count : Integer := 0;
33078 function Factorial (Val : Integer) return Integer;
33080 procedure Initialize_API;
33081 procedure Finalize_API;
33082 -- Initialization and Finalization routines.
33088 @smallexample @c ada
33091 package body API is
33092 function Factorial (Val : Integer) return Integer is
33093 Fact : Integer := 1;
33095 Count := Count + 1;
33096 for K in 1 .. Val loop
33103 -- The remainder of this package body is unchanged.
33110 Note that if you do not export the Ada entities with a @code{C} or
33111 @code{Stdcall} convention you will have to provide the mangled Ada names
33112 in the definition file of the Ada DLL
33113 (@pxref{Creating the Definition File}).
33115 @node Ada DLLs and Elaboration
33116 @subsection Ada DLLs and Elaboration
33117 @cindex DLLs and elaboration
33120 The DLL that you are building contains your Ada code as well as all the
33121 routines in the Ada library that are needed by it. The first thing a
33122 user of your DLL must do is elaborate the Ada code
33123 (@pxref{Elaboration Order Handling in GNAT}).
33125 To achieve this you must export an initialization routine
33126 (@code{Initialize_API} in the previous example), which must be invoked
33127 before using any of the DLL services. This elaboration routine must call
33128 the Ada elaboration routine @code{adainit} generated by the GNAT binder
33129 (@pxref{Binding with Non-Ada Main Programs}). See the body of
33130 @code{Initialize_Api} for an example. Note that the GNAT binder is
33131 automatically invoked during the DLL build process by the @code{gnatdll}
33132 tool (@pxref{Using gnatdll}).
33134 When a DLL is loaded, Windows systematically invokes a routine called
33135 @code{DllMain}. It would therefore be possible to call @code{adainit}
33136 directly from @code{DllMain} without having to provide an explicit
33137 initialization routine. Unfortunately, it is not possible to call
33138 @code{adainit} from the @code{DllMain} if your program has library level
33139 tasks because access to the @code{DllMain} entry point is serialized by
33140 the system (that is, only a single thread can execute ``through'' it at a
33141 time), which means that the GNAT run time will deadlock waiting for the
33142 newly created task to complete its initialization.
33144 @node Ada DLLs and Finalization
33145 @subsection Ada DLLs and Finalization
33146 @cindex DLLs and finalization
33149 When the services of an Ada DLL are no longer needed, the client code should
33150 invoke the DLL finalization routine, if available. The DLL finalization
33151 routine is in charge of releasing all resources acquired by the DLL. In the
33152 case of the Ada code contained in the DLL, this is achieved by calling
33153 routine @code{adafinal} generated by the GNAT binder
33154 (@pxref{Binding with Non-Ada Main Programs}).
33155 See the body of @code{Finalize_Api} for an
33156 example. As already pointed out the GNAT binder is automatically invoked
33157 during the DLL build process by the @code{gnatdll} tool
33158 (@pxref{Using gnatdll}).
33160 @node Creating a Spec for Ada DLLs
33161 @subsection Creating a Spec for Ada DLLs
33164 To use the services exported by the Ada DLL from another programming
33165 language (e.g.@: C), you have to translate the specs of the exported Ada
33166 entities in that language. For instance in the case of @code{API.dll},
33167 the corresponding C header file could look like:
33172 extern int *_imp__count;
33173 #define count (*_imp__count)
33174 int factorial (int);
33180 It is important to understand that when building an Ada DLL to be used by
33181 other Ada applications, you need two different specs for the packages
33182 contained in the DLL: one for building the DLL and the other for using
33183 the DLL. This is because the @code{DLL} calling convention is needed to
33184 use a variable defined in a DLL, but when building the DLL, the variable
33185 must have either the @code{Ada} or @code{C} calling convention. As an
33186 example consider a DLL comprising the following package @code{API}:
33188 @smallexample @c ada
33192 Count : Integer := 0;
33194 -- Remainder of the package omitted.
33201 After producing a DLL containing package @code{API}, the spec that
33202 must be used to import @code{API.Count} from Ada code outside of the
33205 @smallexample @c ada
33210 pragma Import (DLL, Count);
33216 @node Creating the Definition File
33217 @subsection Creating the Definition File
33220 The definition file is the last file needed to build the DLL. It lists
33221 the exported symbols. As an example, the definition file for a DLL
33222 containing only package @code{API} (where all the entities are exported
33223 with a @code{C} calling convention) is:
33238 If the @code{C} calling convention is missing from package @code{API},
33239 then the definition file contains the mangled Ada names of the above
33240 entities, which in this case are:
33249 api__initialize_api
33254 @node Using gnatdll
33255 @subsection Using @code{gnatdll}
33259 * gnatdll Example::
33260 * gnatdll behind the Scenes::
33265 @code{gnatdll} is a tool to automate the DLL build process once all the Ada
33266 and non-Ada sources that make up your DLL have been compiled.
33267 @code{gnatdll} is actually in charge of two distinct tasks: build the
33268 static import library for the DLL and the actual DLL. The form of the
33269 @code{gnatdll} command is
33273 $ gnatdll @ovar{switches} @var{list-of-files} @r{[}-largs @var{opts}@r{]}
33278 where @var{list-of-files} is a list of ALI and object files. The object
33279 file list must be the exact list of objects corresponding to the non-Ada
33280 sources whose services are to be included in the DLL. The ALI file list
33281 must be the exact list of ALI files for the corresponding Ada sources
33282 whose services are to be included in the DLL. If @var{list-of-files} is
33283 missing, only the static import library is generated.
33286 You may specify any of the following switches to @code{gnatdll}:
33289 @item -a@ovar{address}
33290 @cindex @option{-a} (@code{gnatdll})
33291 Build a non-relocatable DLL at @var{address}. If @var{address} is not
33292 specified the default address @var{0x11000000} will be used. By default,
33293 when this switch is missing, @code{gnatdll} builds relocatable DLL. We
33294 advise the reader to build relocatable DLL.
33296 @item -b @var{address}
33297 @cindex @option{-b} (@code{gnatdll})
33298 Set the relocatable DLL base address. By default the address is
33301 @item -bargs @var{opts}
33302 @cindex @option{-bargs} (@code{gnatdll})
33303 Binder options. Pass @var{opts} to the binder.
33305 @item -d @var{dllfile}
33306 @cindex @option{-d} (@code{gnatdll})
33307 @var{dllfile} is the name of the DLL. This switch must be present for
33308 @code{gnatdll} to do anything. The name of the generated import library is
33309 obtained algorithmically from @var{dllfile} as shown in the following
33310 example: if @var{dllfile} is @code{xyz.dll}, the import library name is
33311 @code{libxyz.dll.a}. The name of the definition file to use (if not specified
33312 by option @option{-e}) is obtained algorithmically from @var{dllfile}
33313 as shown in the following example:
33314 if @var{dllfile} is @code{xyz.dll}, the definition
33315 file used is @code{xyz.def}.
33317 @item -e @var{deffile}
33318 @cindex @option{-e} (@code{gnatdll})
33319 @var{deffile} is the name of the definition file.
33322 @cindex @option{-g} (@code{gnatdll})
33323 Generate debugging information. This information is stored in the object
33324 file and copied from there to the final DLL file by the linker,
33325 where it can be read by the debugger. You must use the
33326 @option{-g} switch if you plan on using the debugger or the symbolic
33330 @cindex @option{-h} (@code{gnatdll})
33331 Help mode. Displays @code{gnatdll} switch usage information.
33334 @cindex @option{-I} (@code{gnatdll})
33335 Direct @code{gnatdll} to search the @var{dir} directory for source and
33336 object files needed to build the DLL.
33337 (@pxref{Search Paths and the Run-Time Library (RTL)}).
33340 @cindex @option{-k} (@code{gnatdll})
33341 Removes the @code{@@}@var{nn} suffix from the import library's exported
33342 names, but keeps them for the link names. You must specify this
33343 option if you want to use a @code{Stdcall} function in a DLL for which
33344 the @code{@@}@var{nn} suffix has been removed. This is the case for most
33345 of the Windows NT DLL for example. This option has no effect when
33346 @option{-n} option is specified.
33348 @item -l @var{file}
33349 @cindex @option{-l} (@code{gnatdll})
33350 The list of ALI and object files used to build the DLL are listed in
33351 @var{file}, instead of being given in the command line. Each line in
33352 @var{file} contains the name of an ALI or object file.
33355 @cindex @option{-n} (@code{gnatdll})
33356 No Import. Do not create the import library.
33359 @cindex @option{-q} (@code{gnatdll})
33360 Quiet mode. Do not display unnecessary messages.
33363 @cindex @option{-v} (@code{gnatdll})
33364 Verbose mode. Display extra information.
33366 @item -largs @var{opts}
33367 @cindex @option{-largs} (@code{gnatdll})
33368 Linker options. Pass @var{opts} to the linker.
33371 @node gnatdll Example
33372 @subsubsection @code{gnatdll} Example
33375 As an example the command to build a relocatable DLL from @file{api.adb}
33376 once @file{api.adb} has been compiled and @file{api.def} created is
33379 $ gnatdll -d api.dll api.ali
33383 The above command creates two files: @file{libapi.dll.a} (the import
33384 library) and @file{api.dll} (the actual DLL). If you want to create
33385 only the DLL, just type:
33388 $ gnatdll -d api.dll -n api.ali
33392 Alternatively if you want to create just the import library, type:
33395 $ gnatdll -d api.dll
33398 @node gnatdll behind the Scenes
33399 @subsubsection @code{gnatdll} behind the Scenes
33402 This section details the steps involved in creating a DLL. @code{gnatdll}
33403 does these steps for you. Unless you are interested in understanding what
33404 goes on behind the scenes, you should skip this section.
33406 We use the previous example of a DLL containing the Ada package @code{API},
33407 to illustrate the steps necessary to build a DLL. The starting point is a
33408 set of objects that will make up the DLL and the corresponding ALI
33409 files. In the case of this example this means that @file{api.o} and
33410 @file{api.ali} are available. To build a relocatable DLL, @code{gnatdll} does
33415 @code{gnatdll} builds the base file (@file{api.base}). A base file gives
33416 the information necessary to generate relocation information for the
33422 $ gnatlink api -o api.jnk -mdll -Wl,--base-file,api.base
33427 In addition to the base file, the @command{gnatlink} command generates an
33428 output file @file{api.jnk} which can be discarded. The @option{-mdll} switch
33429 asks @command{gnatlink} to generate the routines @code{DllMain} and
33430 @code{DllMainCRTStartup} that are called by the Windows loader when the DLL
33431 is loaded into memory.
33434 @code{gnatdll} uses @code{dlltool} (@pxref{Using dlltool}) to build the
33435 export table (@file{api.exp}). The export table contains the relocation
33436 information in a form which can be used during the final link to ensure
33437 that the Windows loader is able to place the DLL anywhere in memory.
33441 $ dlltool --dllname api.dll --def api.def --base-file api.base \
33442 --output-exp api.exp
33447 @code{gnatdll} builds the base file using the new export table. Note that
33448 @command{gnatbind} must be called once again since the binder generated file
33449 has been deleted during the previous call to @command{gnatlink}.
33454 $ gnatlink api -o api.jnk api.exp -mdll
33455 -Wl,--base-file,api.base
33460 @code{gnatdll} builds the new export table using the new base file and
33461 generates the DLL import library @file{libAPI.dll.a}.
33465 $ dlltool --dllname api.dll --def api.def --base-file api.base \
33466 --output-exp api.exp --output-lib libAPI.a
33471 Finally @code{gnatdll} builds the relocatable DLL using the final export
33477 $ gnatlink api api.exp -o api.dll -mdll
33482 @node Using dlltool
33483 @subsubsection Using @code{dlltool}
33486 @code{dlltool} is the low-level tool used by @code{gnatdll} to build
33487 DLLs and static import libraries. This section summarizes the most
33488 common @code{dlltool} switches. The form of the @code{dlltool} command
33492 $ dlltool @ovar{switches}
33496 @code{dlltool} switches include:
33499 @item --base-file @var{basefile}
33500 @cindex @option{--base-file} (@command{dlltool})
33501 Read the base file @var{basefile} generated by the linker. This switch
33502 is used to create a relocatable DLL.
33504 @item --def @var{deffile}
33505 @cindex @option{--def} (@command{dlltool})
33506 Read the definition file.
33508 @item --dllname @var{name}
33509 @cindex @option{--dllname} (@command{dlltool})
33510 Gives the name of the DLL. This switch is used to embed the name of the
33511 DLL in the static import library generated by @code{dlltool} with switch
33512 @option{--output-lib}.
33515 @cindex @option{-k} (@command{dlltool})
33516 Kill @code{@@}@var{nn} from exported names
33517 (@pxref{Windows Calling Conventions}
33518 for a discussion about @code{Stdcall}-style symbols.
33521 @cindex @option{--help} (@command{dlltool})
33522 Prints the @code{dlltool} switches with a concise description.
33524 @item --output-exp @var{exportfile}
33525 @cindex @option{--output-exp} (@command{dlltool})
33526 Generate an export file @var{exportfile}. The export file contains the
33527 export table (list of symbols in the DLL) and is used to create the DLL.
33529 @item --output-lib @var{libfile}
33530 @cindex @option{--output-lib} (@command{dlltool})
33531 Generate a static import library @var{libfile}.
33534 @cindex @option{-v} (@command{dlltool})
33537 @item --as @var{assembler-name}
33538 @cindex @option{--as} (@command{dlltool})
33539 Use @var{assembler-name} as the assembler. The default is @code{as}.
33542 @node GNAT and Windows Resources
33543 @section GNAT and Windows Resources
33544 @cindex Resources, windows
33547 * Building Resources::
33548 * Compiling Resources::
33549 * Using Resources::
33553 Resources are an easy way to add Windows specific objects to your
33554 application. The objects that can be added as resources include:
33583 This section explains how to build, compile and use resources.
33585 @node Building Resources
33586 @subsection Building Resources
33587 @cindex Resources, building
33590 A resource file is an ASCII file. By convention resource files have an
33591 @file{.rc} extension.
33592 The easiest way to build a resource file is to use Microsoft tools
33593 such as @code{imagedit.exe} to build bitmaps, icons and cursors and
33594 @code{dlgedit.exe} to build dialogs.
33595 It is always possible to build an @file{.rc} file yourself by writing a
33598 It is not our objective to explain how to write a resource file. A
33599 complete description of the resource script language can be found in the
33600 Microsoft documentation.
33602 @node Compiling Resources
33603 @subsection Compiling Resources
33606 @cindex Resources, compiling
33609 This section describes how to build a GNAT-compatible (COFF) object file
33610 containing the resources. This is done using the Resource Compiler
33611 @code{windres} as follows:
33614 $ windres -i myres.rc -o myres.o
33618 By default @code{windres} will run @command{gcc} to preprocess the @file{.rc}
33619 file. You can specify an alternate preprocessor (usually named
33620 @file{cpp.exe}) using the @code{windres} @option{--preprocessor}
33621 parameter. A list of all possible options may be obtained by entering
33622 the command @code{windres} @option{--help}.
33624 It is also possible to use the Microsoft resource compiler @code{rc.exe}
33625 to produce a @file{.res} file (binary resource file). See the
33626 corresponding Microsoft documentation for further details. In this case
33627 you need to use @code{windres} to translate the @file{.res} file to a
33628 GNAT-compatible object file as follows:
33631 $ windres -i myres.res -o myres.o
33634 @node Using Resources
33635 @subsection Using Resources
33636 @cindex Resources, using
33639 To include the resource file in your program just add the
33640 GNAT-compatible object file for the resource(s) to the linker
33641 arguments. With @command{gnatmake} this is done by using the @option{-largs}
33645 $ gnatmake myprog -largs myres.o
33648 @node Debugging a DLL
33649 @section Debugging a DLL
33650 @cindex DLL debugging
33653 * Program and DLL Both Built with GCC/GNAT::
33654 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
33658 Debugging a DLL is similar to debugging a standard program. But
33659 we have to deal with two different executable parts: the DLL and the
33660 program that uses it. We have the following four possibilities:
33664 The program and the DLL are built with @code{GCC/GNAT}.
33666 The program is built with foreign tools and the DLL is built with
33669 The program is built with @code{GCC/GNAT} and the DLL is built with
33675 In this section we address only cases one and two above.
33676 There is no point in trying to debug
33677 a DLL with @code{GNU/GDB}, if there is no GDB-compatible debugging
33678 information in it. To do so you must use a debugger compatible with the
33679 tools suite used to build the DLL.
33681 @node Program and DLL Both Built with GCC/GNAT
33682 @subsection Program and DLL Both Built with GCC/GNAT
33685 This is the simplest case. Both the DLL and the program have @code{GDB}
33686 compatible debugging information. It is then possible to break anywhere in
33687 the process. Let's suppose here that the main procedure is named
33688 @code{ada_main} and that in the DLL there is an entry point named
33692 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) and
33693 program must have been built with the debugging information (see GNAT -g
33694 switch). Here are the step-by-step instructions for debugging it:
33697 @item Launch @code{GDB} on the main program.
33703 @item Start the program and stop at the beginning of the main procedure
33710 This step is required to be able to set a breakpoint inside the DLL. As long
33711 as the program is not run, the DLL is not loaded. This has the
33712 consequence that the DLL debugging information is also not loaded, so it is not
33713 possible to set a breakpoint in the DLL.
33715 @item Set a breakpoint inside the DLL
33718 (gdb) break ada_dll
33725 At this stage a breakpoint is set inside the DLL. From there on
33726 you can use the standard approach to debug the whole program
33727 (@pxref{Running and Debugging Ada Programs}).
33730 @c This used to work, probably because the DLLs were non-relocatable
33731 @c keep this section around until the problem is sorted out.
33733 To break on the @code{DllMain} routine it is not possible to follow
33734 the procedure above. At the time the program stop on @code{ada_main}
33735 the @code{DllMain} routine as already been called. Either you can use
33736 the procedure below @pxref{Debugging the DLL Directly} or this procedure:
33739 @item Launch @code{GDB} on the main program.
33745 @item Load DLL symbols
33748 (gdb) add-sym api.dll
33751 @item Set a breakpoint inside the DLL
33754 (gdb) break ada_dll.adb:45
33757 Note that at this point it is not possible to break using the routine symbol
33758 directly as the program is not yet running. The solution is to break
33759 on the proper line (break in @file{ada_dll.adb} line 45).
33761 @item Start the program
33770 @node Program Built with Foreign Tools and DLL Built with GCC/GNAT
33771 @subsection Program Built with Foreign Tools and DLL Built with GCC/GNAT
33774 * Debugging the DLL Directly::
33775 * Attaching to a Running Process::
33779 In this case things are slightly more complex because it is not possible to
33780 start the main program and then break at the beginning to load the DLL and the
33781 associated DLL debugging information. It is not possible to break at the
33782 beginning of the program because there is no @code{GDB} debugging information,
33783 and therefore there is no direct way of getting initial control. This
33784 section addresses this issue by describing some methods that can be used
33785 to break somewhere in the DLL to debug it.
33788 First suppose that the main procedure is named @code{main} (this is for
33789 example some C code built with Microsoft Visual C) and that there is a
33790 DLL named @code{test.dll} containing an Ada entry point named
33794 The DLL (@pxref{Introduction to Dynamic Link Libraries (DLLs)}) must have
33795 been built with debugging information (see GNAT -g option).
33797 @node Debugging the DLL Directly
33798 @subsubsection Debugging the DLL Directly
33802 Find out the executable starting address
33805 $ objdump --file-header main.exe
33808 The starting address is reported on the last line. For example:
33811 main.exe: file format pei-i386
33812 architecture: i386, flags 0x0000010a:
33813 EXEC_P, HAS_DEBUG, D_PAGED
33814 start address 0x00401010
33818 Launch the debugger on the executable.
33825 Set a breakpoint at the starting address, and launch the program.
33828 $ (gdb) break *0x00401010
33832 The program will stop at the given address.
33835 Set a breakpoint on a DLL subroutine.
33838 (gdb) break ada_dll.adb:45
33841 Or if you want to break using a symbol on the DLL, you need first to
33842 select the Ada language (language used by the DLL).
33845 (gdb) set language ada
33846 (gdb) break ada_dll
33850 Continue the program.
33857 This will run the program until it reaches the breakpoint that has been
33858 set. From that point you can use the standard way to debug a program
33859 as described in (@pxref{Running and Debugging Ada Programs}).
33864 It is also possible to debug the DLL by attaching to a running process.
33866 @node Attaching to a Running Process
33867 @subsubsection Attaching to a Running Process
33868 @cindex DLL debugging, attach to process
33871 With @code{GDB} it is always possible to debug a running process by
33872 attaching to it. It is possible to debug a DLL this way. The limitation
33873 of this approach is that the DLL must run long enough to perform the
33874 attach operation. It may be useful for instance to insert a time wasting
33875 loop in the code of the DLL to meet this criterion.
33879 @item Launch the main program @file{main.exe}.
33885 @item Use the Windows @i{Task Manager} to find the process ID. Let's say
33886 that the process PID for @file{main.exe} is 208.
33894 @item Attach to the running process to be debugged.
33900 @item Load the process debugging information.
33903 (gdb) symbol-file main.exe
33906 @item Break somewhere in the DLL.
33909 (gdb) break ada_dll
33912 @item Continue process execution.
33921 This last step will resume the process execution, and stop at
33922 the breakpoint we have set. From there you can use the standard
33923 approach to debug a program as described in
33924 (@pxref{Running and Debugging Ada Programs}).
33926 @node Setting Stack Size from gnatlink
33927 @section Setting Stack Size from @command{gnatlink}
33930 It is possible to specify the program stack size at link time. On modern
33931 versions of Windows, starting with XP, this is mostly useful to set the size of
33932 the main stack (environment task). The other task stacks are set with pragma
33933 Storage_Size or with the @command{gnatbind -d} command.
33935 Since older versions of Windows (2000, NT4, etc.) do not allow setting the
33936 reserve size of individual tasks, the link-time stack size applies to all
33937 tasks, and pragma Storage_Size has no effect.
33938 In particular, Stack Overflow checks are made against this
33939 link-time specified size.
33941 This setting can be done with
33942 @command{gnatlink} using either:
33946 @item using @option{-Xlinker} linker option
33949 $ gnatlink hello -Xlinker --stack=0x10000,0x1000
33952 This sets the stack reserve size to 0x10000 bytes and the stack commit
33953 size to 0x1000 bytes.
33955 @item using @option{-Wl} linker option
33958 $ gnatlink hello -Wl,--stack=0x1000000
33961 This sets the stack reserve size to 0x1000000 bytes. Note that with
33962 @option{-Wl} option it is not possible to set the stack commit size
33963 because the coma is a separator for this option.
33967 @node Setting Heap Size from gnatlink
33968 @section Setting Heap Size from @command{gnatlink}
33971 Under Windows systems, it is possible to specify the program heap size from
33972 @command{gnatlink} using either:
33976 @item using @option{-Xlinker} linker option
33979 $ gnatlink hello -Xlinker --heap=0x10000,0x1000
33982 This sets the heap reserve size to 0x10000 bytes and the heap commit
33983 size to 0x1000 bytes.
33985 @item using @option{-Wl} linker option
33988 $ gnatlink hello -Wl,--heap=0x1000000
33991 This sets the heap reserve size to 0x1000000 bytes. Note that with
33992 @option{-Wl} option it is not possible to set the heap commit size
33993 because the coma is a separator for this option.
33999 @c **********************************
34000 @c * GNU Free Documentation License *
34001 @c **********************************
34003 @c GNU Free Documentation License
34005 @node Index,,GNU Free Documentation License, Top
34011 @c Put table of contents at end, otherwise it precedes the "title page" in
34012 @c the .txt version
34013 @c Edit the pdf file to move the contents to the beginning, after the title