1 \input texinfo @c -*-texinfo-*-
3 @setfilename gnat_ugn.info
4 @documentencoding UTF-8
6 @*Generated by Sphinx 4.0.2.@*
8 @settitle GNAT User's Guide for Native Platforms
13 @dircategory GNU Ada Tools
15 * gnat_ugn: (gnat_ugn.info). gnat_ugn
18 @definfoenclose strong,`,'
19 @definfoenclose emph,`,'
24 GNAT User's Guide for Native Platforms , Jun 23, 2021
28 Copyright @copyright{} 2008-2021, Free Software Foundation
34 @title GNAT User's Guide for Native Platforms
39 @c %** start of user preamble
41 @c %** end of user preamble
45 @top GNAT User's Guide for Native Platforms
50 @anchor{gnat_ugn doc}@anchor{0}
51 @emph{GNAT, The GNU Ada Development Environment}
54 @include gcc-common.texi
55 GCC version @value{version-GCC}@*
58 Permission is granted to copy, distribute and/or modify this document
59 under the terms of the GNU Free Documentation License, Version 1.3 or
60 any later version published by the Free Software Foundation; with no
61 Invariant Sections, with the Front-Cover Texts being
62 “GNAT User’s Guide for Native Platforms”,
63 and with no Back-Cover Texts. A copy of the license is
64 included in the section entitled @ref{1,,GNU Free Documentation License}.
68 * Getting Started with GNAT::
69 * The GNAT Compilation Model::
70 * Building Executable Programs with GNAT::
71 * GNAT Utility Programs::
72 * GNAT and Program Execution::
73 * Platform-Specific Information::
74 * Example of Binder Output File::
75 * Elaboration Order Handling in GNAT::
77 * GNU Free Documentation License::
81 --- The Detailed Node Listing ---
85 * What This Guide Contains::
86 * What You Should Know before Reading This Guide::
87 * Related Information::
90 Getting Started with GNAT
92 * System Requirements::
94 * Running a Simple Ada Program::
95 * Running a Program with Multiple Units::
97 The GNAT Compilation Model
99 * Source Representation::
100 * Foreign Language Representation::
101 * File Naming Topics and Utilities::
102 * Configuration Pragmas::
103 * Generating Object Files::
104 * Source Dependencies::
105 * The Ada Library Information Files::
106 * Binding an Ada Program::
107 * GNAT and Libraries::
108 * Conditional Compilation::
109 * Mixed Language Programming::
110 * GNAT and Other Compilation Models::
111 * Using GNAT Files with External Tools::
113 Foreign Language Representation
116 * Other 8-Bit Codes::
117 * Wide_Character Encodings::
118 * Wide_Wide_Character Encodings::
120 File Naming Topics and Utilities
122 * File Naming Rules::
123 * Using Other File Names::
124 * Alternative File Naming Schemes::
125 * Handling Arbitrary File Naming Conventions with gnatname::
126 * File Name Krunching with gnatkr::
127 * Renaming Files with gnatchop::
129 Handling Arbitrary File Naming Conventions with gnatname
131 * Arbitrary File Naming Conventions::
133 * Switches for gnatname::
134 * Examples of gnatname Usage::
136 File Name Krunching with gnatkr
141 * Examples of gnatkr Usage::
143 Renaming Files with gnatchop
145 * Handling Files with Multiple Units::
146 * Operating gnatchop in Compilation Mode::
147 * Command Line for gnatchop::
148 * Switches for gnatchop::
149 * Examples of gnatchop Usage::
151 Configuration Pragmas
153 * Handling of Configuration Pragmas::
154 * The Configuration Pragmas Files::
158 * Introduction to Libraries in GNAT::
159 * General Ada Libraries::
160 * Stand-alone Ada Libraries::
161 * Rebuilding the GNAT Run-Time Library::
163 General Ada Libraries
165 * Building a library::
166 * Installing a library::
169 Stand-alone Ada Libraries
171 * Introduction to Stand-alone Libraries::
172 * Building a Stand-alone Library::
173 * Creating a Stand-alone Library to be used in a non-Ada context::
174 * Restrictions in Stand-alone Libraries::
176 Conditional Compilation
178 * Modeling Conditional Compilation in Ada::
179 * Preprocessing with gnatprep::
180 * Integrated Preprocessing::
182 Modeling Conditional Compilation in Ada
184 * Use of Boolean Constants::
185 * Debugging - A Special Case::
186 * Conditionalizing Declarations::
187 * Use of Alternative Implementations::
190 Preprocessing with gnatprep
192 * Preprocessing Symbols::
194 * Switches for gnatprep::
195 * Form of Definitions File::
196 * Form of Input Text for gnatprep::
198 Mixed Language Programming
201 * Calling Conventions::
202 * Building Mixed Ada and C++ Programs::
203 * Generating Ada Bindings for C and C++ headers::
204 * Generating C Headers for Ada Specifications::
206 Building Mixed Ada and C++ Programs
208 * Interfacing to C++::
209 * Linking a Mixed C++ & Ada Program::
211 * Interfacing with C++ constructors::
212 * Interfacing with C++ at the Class Level::
214 Generating Ada Bindings for C and C++ headers
216 * Running the Binding Generator::
217 * Generating Bindings for C++ Headers::
220 Generating C Headers for Ada Specifications
222 * Running the C Header Generator::
224 GNAT and Other Compilation Models
226 * Comparison between GNAT and C/C++ Compilation Models::
227 * Comparison between GNAT and Conventional Ada Library Models::
229 Using GNAT Files with External Tools
231 * Using Other Utility Programs with GNAT::
232 * The External Symbol Naming Scheme of GNAT::
234 Building Executable Programs with GNAT
236 * Building with gnatmake::
237 * Compiling with gcc::
238 * Compiler Switches::
240 * Binding with gnatbind::
241 * Linking with gnatlink::
242 * Using the GNU make Utility::
244 Building with gnatmake
247 * Switches for gnatmake::
248 * Mode Switches for gnatmake::
249 * Notes on the Command Line::
250 * How gnatmake Works::
251 * Examples of gnatmake Usage::
255 * Compiling Programs::
256 * Search Paths and the Run-Time Library (RTL): Search Paths and the Run-Time Library RTL.
257 * Order of Compilation Issues::
262 * Alphabetical List of All Switches::
263 * Output and Error Message Control::
264 * Warning Message Control::
265 * Debugging and Assertion Control::
266 * Validity Checking::
269 * Using gcc for Syntax Checking::
270 * Using gcc for Semantic Checking::
271 * Compiling Different Versions of Ada::
272 * Character Set Control::
273 * File Naming Control::
274 * Subprogram Inlining Control::
275 * Auxiliary Output Control::
276 * Debugging Control::
277 * Exception Handling Control::
278 * Units to Sources Mapping Files::
279 * Code Generation Control::
281 Binding with gnatbind
284 * Switches for gnatbind::
285 * Command-Line Access::
286 * Search Paths for gnatbind::
287 * Examples of gnatbind Usage::
289 Switches for gnatbind
291 * Consistency-Checking Modes::
292 * Binder Error Message Control::
293 * Elaboration Control::
295 * Dynamic Allocation Control::
296 * Binding with Non-Ada Main Programs::
297 * Binding Programs with No Main Subprogram::
299 Linking with gnatlink
302 * Switches for gnatlink::
304 Using the GNU make Utility
306 * Using gnatmake in a Makefile::
307 * Automatically Creating a List of Directories::
308 * Generating the Command Line Switches::
309 * Overcoming Command Line Length Limits::
311 GNAT Utility Programs
313 * The File Cleanup Utility gnatclean::
314 * The GNAT Library Browser gnatls::
316 The File Cleanup Utility gnatclean
318 * Running gnatclean::
319 * Switches for gnatclean::
321 The GNAT Library Browser gnatls
324 * Switches for gnatls::
325 * Example of gnatls Usage::
327 GNAT and Program Execution
329 * Running and Debugging Ada Programs::
331 * Improving Performance::
332 * Overflow Check Handling in GNAT::
333 * Performing Dimensionality Analysis in GNAT::
334 * Stack Related Facilities::
335 * Memory Management Issues::
337 Running and Debugging Ada Programs
339 * The GNAT Debugger GDB::
341 * Introduction to GDB Commands::
342 * Using Ada Expressions::
343 * Calling User-Defined Subprograms::
344 * Using the next Command in a Function::
345 * Stopping When Ada Exceptions Are Raised::
347 * Debugging Generic Units::
348 * Remote Debugging with gdbserver::
349 * GNAT Abnormal Termination or Failure to Terminate::
350 * Naming Conventions for GNAT Source Files::
351 * Getting Internal Debugging Information::
353 * Pretty-Printers for the GNAT runtime::
357 * Non-Symbolic Traceback::
358 * Symbolic Traceback::
362 * Profiling an Ada Program with gprof::
364 Profiling an Ada Program with gprof
366 * Compilation for profiling::
367 * Program execution::
369 * Interpretation of profiling results::
371 Improving Performance
373 * Performance Considerations::
374 * Text_IO Suggestions::
375 * Reducing Size of Executables with Unused Subprogram/Data Elimination::
377 Performance Considerations
379 * Controlling Run-Time Checks::
380 * Use of Restrictions::
381 * Optimization Levels::
382 * Debugging Optimized Code::
383 * Inlining of Subprograms::
384 * Floating Point Operations::
385 * Vectorization of loops::
386 * Other Optimization Switches::
387 * Optimization and Strict Aliasing::
388 * Aliased Variables and Optimization::
389 * Atomic Variables and Optimization::
390 * Passive Task Optimization::
392 Reducing Size of Executables with Unused Subprogram/Data Elimination
394 * About unused subprogram/data elimination::
395 * Compilation options::
396 * Example of unused subprogram/data elimination::
398 Overflow Check Handling in GNAT
401 * Management of Overflows in GNAT::
402 * Specifying the Desired Mode::
404 * Implementation Notes::
406 Stack Related Facilities
408 * Stack Overflow Checking::
409 * Static Stack Usage Analysis::
410 * Dynamic Stack Usage Analysis::
412 Memory Management Issues
414 * Some Useful Memory Pools::
415 * The GNAT Debug Pool Facility::
417 Platform-Specific Information
419 * Run-Time Libraries::
420 * Specifying a Run-Time Library::
422 * Microsoft Windows Topics::
427 * Summary of Run-Time Configurations::
429 Specifying a Run-Time Library
431 * Choosing the Scheduling Policy::
435 * Required Packages on GNU/Linux::
437 Microsoft Windows Topics
439 * Using GNAT on Windows::
440 * Using a network installation of GNAT::
441 * CONSOLE and WINDOWS subsystems::
443 * Disabling Command Line Argument Expansion::
444 * Windows Socket Timeouts::
445 * Mixed-Language Programming on Windows::
446 * Windows Specific Add-Ons::
448 Mixed-Language Programming on Windows
450 * Windows Calling Conventions::
451 * Introduction to Dynamic Link Libraries (DLLs): Introduction to Dynamic Link Libraries DLLs.
452 * Using DLLs with GNAT::
453 * Building DLLs with GNAT Project files::
454 * Building DLLs with GNAT::
455 * Building DLLs with gnatdll::
456 * Ada DLLs and Finalization::
457 * Creating a Spec for Ada DLLs::
458 * GNAT and Windows Resources::
459 * Using GNAT DLLs from Microsoft Visual Studio Applications::
461 * Setting Stack Size from gnatlink::
462 * Setting Heap Size from gnatlink::
464 Windows Calling Conventions
466 * C Calling Convention::
467 * Stdcall Calling Convention::
468 * Win32 Calling Convention::
469 * DLL Calling Convention::
473 * Creating an Ada Spec for the DLL Services::
474 * Creating an Import Library::
476 Building DLLs with gnatdll
478 * Limitations When Using Ada DLLs from Ada::
479 * Exporting Ada Entities::
480 * Ada DLLs and Elaboration::
482 Creating a Spec for Ada DLLs
484 * Creating the Definition File::
487 GNAT and Windows Resources
489 * Building Resources::
490 * Compiling Resources::
495 * Program and DLL Both Built with GCC/GNAT::
496 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
498 Windows Specific Add-Ons
505 * Codesigning the Debugger::
507 Elaboration Order Handling in GNAT
510 * Elaboration Order::
511 * Checking the Elaboration Order::
512 * Controlling the Elaboration Order in Ada::
513 * Controlling the Elaboration Order in GNAT::
514 * Mixing Elaboration Models::
516 * SPARK Diagnostics::
517 * Elaboration Circularities::
518 * Resolving Elaboration Circularities::
519 * Elaboration-related Compiler Switches::
520 * Summary of Procedures for Elaboration Control::
521 * Inspecting the Chosen Elaboration Order::
525 * Basic Assembler Syntax::
526 * A Simple Example of Inline Assembler::
527 * Output Variables in Inline Assembler::
528 * Input Variables in Inline Assembler::
529 * Inlining Inline Assembler Code::
530 * Other Asm Functionality::
532 Other Asm Functionality
534 * The Clobber Parameter::
535 * The Volatile Parameter::
540 @node About This Guide,Getting Started with GNAT,Top,Top
541 @anchor{gnat_ugn/about_this_guide doc}@anchor{2}@anchor{gnat_ugn/about_this_guide about-this-guide}@anchor{3}@anchor{gnat_ugn/about_this_guide gnat-user-s-guide-for-native-platforms}@anchor{4}@anchor{gnat_ugn/about_this_guide id1}@anchor{5}
542 @chapter About This Guide
546 This guide describes the use of GNAT,
547 a compiler and software development
548 toolset for the full Ada programming language.
549 It documents the features of the compiler and tools, and explains
550 how to use them to build Ada applications.
552 GNAT implements Ada 95, Ada 2005, Ada 2012, and Ada 202x, and it may also be
553 invoked in Ada 83 compatibility mode.
554 By default, GNAT assumes Ada 2012, but you can override with a
555 compiler switch (@ref{6,,Compiling Different Versions of Ada})
556 to explicitly specify the language version.
557 Throughout this manual, references to ‘Ada’ without a year suffix
558 apply to all Ada versions of the language, starting with Ada 95.
561 * What This Guide Contains::
562 * What You Should Know before Reading This Guide::
563 * Related Information::
568 @node What This Guide Contains,What You Should Know before Reading This Guide,,About This Guide
569 @anchor{gnat_ugn/about_this_guide what-this-guide-contains}@anchor{7}
570 @section What This Guide Contains
573 This guide contains the following chapters:
579 @ref{8,,Getting Started with GNAT} describes how to get started compiling
580 and running Ada programs with the GNAT Ada programming environment.
583 @ref{9,,The GNAT Compilation Model} describes the compilation model used
587 @ref{a,,Building Executable Programs with GNAT} describes how to use the
588 main GNAT tools to build executable programs, and it also gives examples of
589 using the GNU make utility with GNAT.
592 @ref{b,,GNAT Utility Programs} explains the various utility programs that
593 are included in the GNAT environment
596 @ref{c,,GNAT and Program Execution} covers a number of topics related to
597 running, debugging, and tuning the performace of programs developed
601 Appendices cover several additional topics:
607 @ref{d,,Platform-Specific Information} describes the different run-time
608 library implementations and also presents information on how to use
609 GNAT on several specific platforms
612 @ref{e,,Example of Binder Output File} shows the source code for the binder
613 output file for a sample program.
616 @ref{f,,Elaboration Order Handling in GNAT} describes how GNAT helps
617 you deal with elaboration order issues.
620 @ref{10,,Inline Assembler} shows how to use the inline assembly facility
624 @node What You Should Know before Reading This Guide,Related Information,What This Guide Contains,About This Guide
625 @anchor{gnat_ugn/about_this_guide what-you-should-know-before-reading-this-guide}@anchor{11}
626 @section What You Should Know before Reading This Guide
629 @geindex Ada 95 Language Reference Manual
631 @geindex Ada 2005 Language Reference Manual
633 This guide assumes a basic familiarity with the Ada 95 language, as
634 described in the International Standard ANSI/ISO/IEC-8652:1995, January
636 Reference manuals for Ada 95, Ada 2005, and Ada 2012 are included in
637 the GNAT documentation package.
639 @node Related Information,Conventions,What You Should Know before Reading This Guide,About This Guide
640 @anchor{gnat_ugn/about_this_guide related-information}@anchor{12}
641 @section Related Information
644 For further information about Ada and related tools, please refer to the
651 @cite{Ada 95 Reference Manual}, @cite{Ada 2005 Reference Manual}, and
652 @cite{Ada 2012 Reference Manual}, which contain reference
653 material for the several revisions of the Ada language standard.
656 @cite{GNAT Reference_Manual}, which contains all reference material for the GNAT
657 implementation of Ada.
660 @cite{Using GNAT Studio}, which describes the GNAT Studio
661 Integrated Development Environment.
664 @cite{GNAT Studio Tutorial}, which introduces the
665 main GNAT Studio features through examples.
668 @cite{Debugging with GDB},
669 for all details on the use of the GNU source-level debugger.
672 @cite{GNU Emacs Manual},
673 for full information on the extensible editor and programming
677 @node Conventions,,Related Information,About This Guide
678 @anchor{gnat_ugn/about_this_guide conventions}@anchor{13}
683 @geindex typographical
685 @geindex Typographical conventions
687 Following are examples of the typographical and graphic conventions used
694 @code{Functions}, @code{utility program names}, @code{standard names},
710 [optional information or parameters]
713 Examples are described by text
716 and then shown this way.
720 Commands that are entered by the user are shown as preceded by a prompt string
721 comprising the @code{$} character followed by a space.
724 Full file names are shown with the ‘/’ character
725 as the directory separator; e.g., @code{parent-dir/subdir/myfile.adb}.
726 If you are using GNAT on a Windows platform, please note that
727 the ‘' character should be used instead.
730 @node Getting Started with GNAT,The GNAT Compilation Model,About This Guide,Top
731 @anchor{gnat_ugn/getting_started_with_gnat doc}@anchor{14}@anchor{gnat_ugn/getting_started_with_gnat getting-started-with-gnat}@anchor{8}@anchor{gnat_ugn/getting_started_with_gnat id1}@anchor{15}
732 @chapter Getting Started with GNAT
735 This chapter describes how to use GNAT’s command line interface to build
736 executable Ada programs.
737 On most platforms a visually oriented Integrated Development Environment
738 is also available: GNAT Studio.
739 GNAT Studio offers a graphical “look and feel”, support for development in
740 other programming languages, comprehensive browsing features, and
741 many other capabilities.
742 For information on GNAT Studio please refer to the
743 @cite{GNAT Studio documentation}.
746 * System Requirements::
748 * Running a Simple Ada Program::
749 * Running a Program with Multiple Units::
753 @node System Requirements,Running GNAT,,Getting Started with GNAT
754 @anchor{gnat_ugn/getting_started_with_gnat id2}@anchor{16}@anchor{gnat_ugn/getting_started_with_gnat system-requirements}@anchor{17}
755 @section System Requirements
758 Even though any machine can run the GNAT toolset and GNAT Studio IDE, in order
759 to get the best experience, we recommend using a machine with as many cores
760 as possible since all individual compilations can run in parallel.
761 A comfortable setup for a compiler server is a machine with 24 physical cores
762 or more, with at least 48 GB of memory (2 GB per core).
764 For a desktop machine, a minimum of 4 cores is recommended (8 preferred),
765 with at least 2GB per core (so 8 to 16GB).
767 In addition, for running and navigating sources in GNAT Studio smoothly, we
768 recommend at least 1.5 GB plus 3 GB of RAM per 1 million source line of code.
769 In other words, we recommend at least 3 GB for for 500K lines of code and
770 7.5 GB for 2 million lines of code.
772 Note that using local and fast drives will also make a difference in terms of
773 build and link time. Network drives such as NFS, SMB, or worse, configuration
774 management filesystems (such as ClearCase dynamic views) should be avoided as
775 much as possible and will produce very degraded performance (typically 2 to 3
776 times slower than on local fast drives). If such slow drives cannot be avoided
777 for accessing the source code, then you should at least configure your project
778 file so that the result of the compilation is stored on a drive local to the
779 machine performing the run. This can be achieved by setting the @code{Object_Dir}
780 project file attribute.
782 @node Running GNAT,Running a Simple Ada Program,System Requirements,Getting Started with GNAT
783 @anchor{gnat_ugn/getting_started_with_gnat id3}@anchor{18}@anchor{gnat_ugn/getting_started_with_gnat running-gnat}@anchor{19}
784 @section Running GNAT
787 Three steps are needed to create an executable file from an Ada source
794 The source file(s) must be compiled.
797 The file(s) must be bound using the GNAT binder.
800 All appropriate object files must be linked to produce an executable.
803 All three steps are most commonly handled by using the @code{gnatmake}
804 utility program that, given the name of the main program, automatically
805 performs the necessary compilation, binding and linking steps.
807 @node Running a Simple Ada Program,Running a Program with Multiple Units,Running GNAT,Getting Started with GNAT
808 @anchor{gnat_ugn/getting_started_with_gnat id4}@anchor{1a}@anchor{gnat_ugn/getting_started_with_gnat running-a-simple-ada-program}@anchor{1b}
809 @section Running a Simple Ada Program
812 Any text editor may be used to prepare an Ada program.
813 (If Emacs is used, the optional Ada mode may be helpful in laying out the
815 The program text is a normal text file. We will assume in our initial
816 example that you have used your editor to prepare the following
817 standard format text file:
820 with Ada.Text_IO; use Ada.Text_IO;
823 Put_Line ("Hello WORLD!");
827 This file should be named @code{hello.adb}.
828 With the normal default file naming conventions, GNAT requires
830 contain a single compilation unit whose file name is the
832 with periods replaced by hyphens; the
833 extension is @code{ads} for a
834 spec and @code{adb} for a body.
835 You can override this default file naming convention by use of the
836 special pragma @code{Source_File_Name} (for further information please
837 see @ref{1c,,Using Other File Names}).
838 Alternatively, if you want to rename your files according to this default
839 convention, which is probably more convenient if you will be using GNAT
840 for all your compilations, then the @code{gnatchop} utility
841 can be used to generate correctly-named source files
842 (see @ref{1d,,Renaming Files with gnatchop}).
844 You can compile the program using the following command (@code{$} is used
845 as the command prompt in the examples in this document):
851 @code{gcc} is the command used to run the compiler. This compiler is
852 capable of compiling programs in several languages, including Ada and
853 C. It assumes that you have given it an Ada program if the file extension is
854 either @code{.ads} or @code{.adb}, and it will then call
855 the GNAT compiler to compile the specified file.
857 The @code{-c} switch is required. It tells @code{gcc} to only do a
858 compilation. (For C programs, @code{gcc} can also do linking, but this
859 capability is not used directly for Ada programs, so the @code{-c}
860 switch must always be present.)
862 This compile command generates a file
863 @code{hello.o}, which is the object
864 file corresponding to your Ada program. It also generates
865 an ‘Ada Library Information’ file @code{hello.ali},
866 which contains additional information used to check
867 that an Ada program is consistent.
869 To build an executable file, use either @code{gnatmake} or gprbuild with
870 the name of the main file: these tools are builders that will take care of
871 all the necessary build steps in the correct order.
872 In particular, these builders automatically recompile any sources that have
873 been modified since they were last compiled, or sources that depend
874 on such modified sources, so that ‘version skew’ is avoided.
876 @geindex Version skew (avoided by `@w{`}gnatmake`@w{`})
882 The result is an executable program called @code{hello}, which can be
889 assuming that the current directory is on the search path
890 for executable programs.
892 and, if all has gone well, you will see:
898 appear in response to this command.
900 @node Running a Program with Multiple Units,,Running a Simple Ada Program,Getting Started with GNAT
901 @anchor{gnat_ugn/getting_started_with_gnat id5}@anchor{1e}@anchor{gnat_ugn/getting_started_with_gnat running-a-program-with-multiple-units}@anchor{1f}
902 @section Running a Program with Multiple Units
905 Consider a slightly more complicated example that has three files: a
906 main program, and the spec and body of a package:
914 with Ada.Text_IO; use Ada.Text_IO;
915 package body Greetings is
918 Put_Line ("Hello WORLD!");
923 Put_Line ("Goodbye WORLD!");
935 Following the one-unit-per-file rule, place this program in the
936 following three separate files:
941 @item @emph{greetings.ads}
943 spec of package @code{Greetings}
945 @item @emph{greetings.adb}
947 body of package @code{Greetings}
949 @item @emph{gmain.adb}
954 Note that there is no required order of compilation when using GNAT.
955 In particular it is perfectly fine to compile the main program first.
956 Also, it is not necessary to compile package specs in the case where
957 there is an accompanying body; you only need to compile the body. If you want
958 to submit these files to the compiler for semantic checking and not code
959 generation, then use the @code{-gnatc} switch:
962 $ gcc -c greetings.ads -gnatc
965 Although the compilation can be done in separate steps, in practice it is
966 almost always more convenient to use the @code{gnatmake} or @code{gprbuild} tools:
972 @c -- Example: A |withing| unit has a |with| clause, it |withs| a |withed| unit
974 @node The GNAT Compilation Model,Building Executable Programs with GNAT,Getting Started with GNAT,Top
975 @anchor{gnat_ugn/the_gnat_compilation_model doc}@anchor{20}@anchor{gnat_ugn/the_gnat_compilation_model id1}@anchor{21}@anchor{gnat_ugn/the_gnat_compilation_model the-gnat-compilation-model}@anchor{9}
976 @chapter The GNAT Compilation Model
979 @geindex GNAT compilation model
981 @geindex Compilation model
983 This chapter describes the compilation model used by GNAT. Although
984 similar to that used by other languages such as C and C++, this model
985 is substantially different from the traditional Ada compilation models,
986 which are based on a centralized program library. The chapter covers
987 the following material:
993 Topics related to source file makeup and naming
999 @ref{22,,Source Representation}
1002 @ref{23,,Foreign Language Representation}
1005 @ref{24,,File Naming Topics and Utilities}
1009 @ref{25,,Configuration Pragmas}
1012 @ref{26,,Generating Object Files}
1015 @ref{27,,Source Dependencies}
1018 @ref{28,,The Ada Library Information Files}
1021 @ref{29,,Binding an Ada Program}
1024 @ref{2a,,GNAT and Libraries}
1027 @ref{2b,,Conditional Compilation}
1030 @ref{2c,,Mixed Language Programming}
1033 @ref{2d,,GNAT and Other Compilation Models}
1036 @ref{2e,,Using GNAT Files with External Tools}
1040 * Source Representation::
1041 * Foreign Language Representation::
1042 * File Naming Topics and Utilities::
1043 * Configuration Pragmas::
1044 * Generating Object Files::
1045 * Source Dependencies::
1046 * The Ada Library Information Files::
1047 * Binding an Ada Program::
1048 * GNAT and Libraries::
1049 * Conditional Compilation::
1050 * Mixed Language Programming::
1051 * GNAT and Other Compilation Models::
1052 * Using GNAT Files with External Tools::
1056 @node Source Representation,Foreign Language Representation,,The GNAT Compilation Model
1057 @anchor{gnat_ugn/the_gnat_compilation_model id2}@anchor{2f}@anchor{gnat_ugn/the_gnat_compilation_model source-representation}@anchor{22}
1058 @section Source Representation
1069 Ada source programs are represented in standard text files, using
1070 Latin-1 coding. Latin-1 is an 8-bit code that includes the familiar
1071 7-bit ASCII set, plus additional characters used for
1072 representing foreign languages (see @ref{23,,Foreign Language Representation}
1073 for support of non-USA character sets). The format effector characters
1074 are represented using their standard ASCII encodings, as follows:
1079 @multitable {xxxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxx}
1156 Source files are in standard text file format. In addition, GNAT will
1157 recognize a wide variety of stream formats, in which the end of
1158 physical lines is marked by any of the following sequences:
1159 @code{LF}, @code{CR}, @code{CR-LF}, or @code{LF-CR}. This is useful
1160 in accommodating files that are imported from other operating systems.
1162 @geindex End of source file; Source file@comma{} end
1164 @geindex SUB (control character)
1166 The end of a source file is normally represented by the physical end of
1167 file. However, the control character @code{16#1A#} (@code{SUB}) is also
1168 recognized as signalling the end of the source file. Again, this is
1169 provided for compatibility with other operating systems where this
1170 code is used to represent the end of file.
1172 @geindex spec (definition)
1173 @geindex compilation (definition)
1175 Each file contains a single Ada compilation unit, including any pragmas
1176 associated with the unit. For example, this means you must place a
1177 package declaration (a package @emph{spec}) and the corresponding body in
1178 separate files. An Ada @emph{compilation} (which is a sequence of
1179 compilation units) is represented using a sequence of files. Similarly,
1180 you will place each subunit or child unit in a separate file.
1182 @node Foreign Language Representation,File Naming Topics and Utilities,Source Representation,The GNAT Compilation Model
1183 @anchor{gnat_ugn/the_gnat_compilation_model foreign-language-representation}@anchor{23}@anchor{gnat_ugn/the_gnat_compilation_model id3}@anchor{30}
1184 @section Foreign Language Representation
1187 GNAT supports the standard character sets defined in Ada as well as
1188 several other non-standard character sets for use in localized versions
1189 of the compiler (@ref{31,,Character Set Control}).
1193 * Other 8-Bit Codes::
1194 * Wide_Character Encodings::
1195 * Wide_Wide_Character Encodings::
1199 @node Latin-1,Other 8-Bit Codes,,Foreign Language Representation
1200 @anchor{gnat_ugn/the_gnat_compilation_model id4}@anchor{32}@anchor{gnat_ugn/the_gnat_compilation_model latin-1}@anchor{33}
1206 The basic character set is Latin-1. This character set is defined by ISO
1207 standard 8859, part 1. The lower half (character codes @code{16#00#}
1208 … @code{16#7F#)} is identical to standard ASCII coding, but the upper
1209 half is used to represent additional characters. These include extended letters
1210 used by European languages, such as French accents, the vowels with umlauts
1211 used in German, and the extra letter A-ring used in Swedish.
1213 @geindex Ada.Characters.Latin_1
1215 For a complete list of Latin-1 codes and their encodings, see the source
1216 file of library unit @code{Ada.Characters.Latin_1} in file
1217 @code{a-chlat1.ads}.
1218 You may use any of these extended characters freely in character or
1219 string literals. In addition, the extended characters that represent
1220 letters can be used in identifiers.
1222 @node Other 8-Bit Codes,Wide_Character Encodings,Latin-1,Foreign Language Representation
1223 @anchor{gnat_ugn/the_gnat_compilation_model id5}@anchor{34}@anchor{gnat_ugn/the_gnat_compilation_model other-8-bit-codes}@anchor{35}
1224 @subsection Other 8-Bit Codes
1227 GNAT also supports several other 8-bit coding schemes:
1236 @item @emph{ISO 8859-2 (Latin-2)}
1238 Latin-2 letters allowed in identifiers, with uppercase and lowercase
1249 @item @emph{ISO 8859-3 (Latin-3)}
1251 Latin-3 letters allowed in identifiers, with uppercase and lowercase
1262 @item @emph{ISO 8859-4 (Latin-4)}
1264 Latin-4 letters allowed in identifiers, with uppercase and lowercase
1275 @item @emph{ISO 8859-5 (Cyrillic)}
1277 ISO 8859-5 letters (Cyrillic) allowed in identifiers, with uppercase and
1278 lowercase equivalence.
1281 @geindex ISO 8859-15
1288 @item @emph{ISO 8859-15 (Latin-9)}
1290 ISO 8859-15 (Latin-9) letters allowed in identifiers, with uppercase and
1291 lowercase equivalence
1294 @geindex code page 437 (IBM PC)
1299 @item @emph{IBM PC (code page 437)}
1301 This code page is the normal default for PCs in the U.S. It corresponds
1302 to the original IBM PC character set. This set has some, but not all, of
1303 the extended Latin-1 letters, but these letters do not have the same
1304 encoding as Latin-1. In this mode, these letters are allowed in
1305 identifiers with uppercase and lowercase equivalence.
1308 @geindex code page 850 (IBM PC)
1313 @item @emph{IBM PC (code page 850)}
1315 This code page is a modification of 437 extended to include all the
1316 Latin-1 letters, but still not with the usual Latin-1 encoding. In this
1317 mode, all these letters are allowed in identifiers with uppercase and
1318 lowercase equivalence.
1320 @item @emph{Full Upper 8-bit}
1322 Any character in the range 80-FF allowed in identifiers, and all are
1323 considered distinct. In other words, there are no uppercase and lowercase
1324 equivalences in this range. This is useful in conjunction with
1325 certain encoding schemes used for some foreign character sets (e.g.,
1326 the typical method of representing Chinese characters on the PC).
1328 @item @emph{No Upper-Half}
1330 No upper-half characters in the range 80-FF are allowed in identifiers.
1331 This gives Ada 83 compatibility for identifier names.
1334 For precise data on the encodings permitted, and the uppercase and lowercase
1335 equivalences that are recognized, see the file @code{csets.adb} in
1336 the GNAT compiler sources. You will need to obtain a full source release
1337 of GNAT to obtain this file.
1339 @node Wide_Character Encodings,Wide_Wide_Character Encodings,Other 8-Bit Codes,Foreign Language Representation
1340 @anchor{gnat_ugn/the_gnat_compilation_model id6}@anchor{36}@anchor{gnat_ugn/the_gnat_compilation_model wide-character-encodings}@anchor{37}
1341 @subsection Wide_Character Encodings
1344 GNAT allows wide character codes to appear in character and string
1345 literals, and also optionally in identifiers, by means of the following
1346 possible encoding schemes:
1351 @item @emph{Hex Coding}
1353 In this encoding, a wide character is represented by the following five
1360 where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1361 characters (using uppercase letters) of the wide character code. For
1362 example, ESC A345 is used to represent the wide character with code
1364 This scheme is compatible with use of the full Wide_Character set.
1366 @item @emph{Upper-Half Coding}
1368 @geindex Upper-Half Coding
1370 The wide character with encoding @code{16#abcd#} where the upper bit is on
1371 (in other words, ‘a’ is in the range 8-F) is represented as two bytes,
1372 @code{16#ab#} and @code{16#cd#}. The second byte cannot be a format control
1373 character, but is not required to be in the upper half. This method can
1374 be also used for shift-JIS or EUC, where the internal coding matches the
1377 @item @emph{Shift JIS Coding}
1379 @geindex Shift JIS Coding
1381 A wide character is represented by a two-character sequence,
1383 @code{16#cd#}, with the restrictions described for upper-half encoding as
1384 described above. The internal character code is the corresponding JIS
1385 character according to the standard algorithm for Shift-JIS
1386 conversion. Only characters defined in the JIS code set table can be
1387 used with this encoding method.
1389 @item @emph{EUC Coding}
1393 A wide character is represented by a two-character sequence
1395 @code{16#cd#}, with both characters being in the upper half. The internal
1396 character code is the corresponding JIS character according to the EUC
1397 encoding algorithm. Only characters defined in the JIS code set table
1398 can be used with this encoding method.
1400 @item @emph{UTF-8 Coding}
1402 A wide character is represented using
1403 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
1404 10646-1/Am.2. Depending on the character value, the representation
1405 is a one, two, or three byte sequence:
1408 16#0000#-16#007f#: 2#0xxxxxxx#
1409 16#0080#-16#07ff#: 2#110xxxxx# 2#10xxxxxx#
1410 16#0800#-16#ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
1413 where the @code{xxx} bits correspond to the left-padded bits of the
1414 16-bit character value. Note that all lower half ASCII characters
1415 are represented as ASCII bytes and all upper half characters and
1416 other wide characters are represented as sequences of upper-half
1417 (The full UTF-8 scheme allows for encoding 31-bit characters as
1418 6-byte sequences, and in the following section on wide wide
1419 characters, the use of these sequences is documented).
1421 @item @emph{Brackets Coding}
1423 In this encoding, a wide character is represented by the following eight
1430 where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
1431 characters (using uppercase letters) of the wide character code. For
1432 example, [‘A345’] is used to represent the wide character with code
1433 @code{16#A345#}. It is also possible (though not required) to use the
1434 Brackets coding for upper half characters. For example, the code
1435 @code{16#A3#} can be represented as @code{['A3']}.
1437 This scheme is compatible with use of the full Wide_Character set,
1438 and is also the method used for wide character encoding in some standard
1439 ACATS (Ada Conformity Assessment Test Suite) test suite distributions.
1444 Some of these coding schemes do not permit the full use of the
1445 Ada character set. For example, neither Shift JIS nor EUC allow the
1446 use of the upper half of the Latin-1 set.
1450 @node Wide_Wide_Character Encodings,,Wide_Character Encodings,Foreign Language Representation
1451 @anchor{gnat_ugn/the_gnat_compilation_model id7}@anchor{38}@anchor{gnat_ugn/the_gnat_compilation_model wide-wide-character-encodings}@anchor{39}
1452 @subsection Wide_Wide_Character Encodings
1455 GNAT allows wide wide character codes to appear in character and string
1456 literals, and also optionally in identifiers, by means of the following
1457 possible encoding schemes:
1462 @item @emph{UTF-8 Coding}
1464 A wide character is represented using
1465 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
1466 10646-1/Am.2. Depending on the character value, the representation
1467 of character codes with values greater than 16#FFFF# is a
1468 is a four, five, or six byte sequence:
1471 16#01_0000#-16#10_FFFF#: 11110xxx 10xxxxxx 10xxxxxx
1473 16#0020_0000#-16#03FF_FFFF#: 111110xx 10xxxxxx 10xxxxxx
1475 16#0400_0000#-16#7FFF_FFFF#: 1111110x 10xxxxxx 10xxxxxx
1476 10xxxxxx 10xxxxxx 10xxxxxx
1479 where the @code{xxx} bits correspond to the left-padded bits of the
1480 32-bit character value.
1482 @item @emph{Brackets Coding}
1484 In this encoding, a wide wide character is represented by the following ten or
1485 twelve byte character sequence:
1489 [ " a b c d e f g h " ]
1492 where @code{a-h} are the six or eight hexadecimal
1493 characters (using uppercase letters) of the wide wide character code. For
1494 example, [“1F4567”] is used to represent the wide wide character with code
1495 @code{16#001F_4567#}.
1497 This scheme is compatible with use of the full Wide_Wide_Character set,
1498 and is also the method used for wide wide character encoding in some standard
1499 ACATS (Ada Conformity Assessment Test Suite) test suite distributions.
1502 @node File Naming Topics and Utilities,Configuration Pragmas,Foreign Language Representation,The GNAT Compilation Model
1503 @anchor{gnat_ugn/the_gnat_compilation_model file-naming-topics-and-utilities}@anchor{24}@anchor{gnat_ugn/the_gnat_compilation_model id8}@anchor{3a}
1504 @section File Naming Topics and Utilities
1507 GNAT has a default file naming scheme and also provides the user with
1508 a high degree of control over how the names and extensions of the
1509 source files correspond to the Ada compilation units that they contain.
1512 * File Naming Rules::
1513 * Using Other File Names::
1514 * Alternative File Naming Schemes::
1515 * Handling Arbitrary File Naming Conventions with gnatname::
1516 * File Name Krunching with gnatkr::
1517 * Renaming Files with gnatchop::
1521 @node File Naming Rules,Using Other File Names,,File Naming Topics and Utilities
1522 @anchor{gnat_ugn/the_gnat_compilation_model file-naming-rules}@anchor{3b}@anchor{gnat_ugn/the_gnat_compilation_model id9}@anchor{3c}
1523 @subsection File Naming Rules
1526 The default file name is determined by the name of the unit that the
1527 file contains. The name is formed by taking the full expanded name of
1528 the unit and replacing the separating dots with hyphens and using
1529 lowercase for all letters.
1531 An exception arises if the file name generated by the above rules starts
1532 with one of the characters
1533 @code{a}, @code{g}, @code{i}, or @code{s}, and the second character is a
1534 minus. In this case, the character tilde is used in place
1535 of the minus. The reason for this special rule is to avoid clashes with
1536 the standard names for child units of the packages System, Ada,
1537 Interfaces, and GNAT, which use the prefixes
1538 @code{s-}, @code{a-}, @code{i-}, and @code{g-},
1541 The file extension is @code{.ads} for a spec and
1542 @code{.adb} for a body. The following table shows some
1543 examples of these rules.
1548 @multitable {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
1555 Ada Compilation Unit
1575 @code{arith_functions.ads}
1579 Arith_Functions (package spec)
1583 @code{arith_functions.adb}
1587 Arith_Functions (package body)
1591 @code{func-spec.ads}
1595 Func.Spec (child package spec)
1599 @code{func-spec.adb}
1603 Func.Spec (child package body)
1611 Sub (subunit of Main)
1619 A.Bad (child package body)
1625 Following these rules can result in excessively long
1626 file names if corresponding
1627 unit names are long (for example, if child units or subunits are
1628 heavily nested). An option is available to shorten such long file names
1629 (called file name ‘krunching’). This may be particularly useful when
1630 programs being developed with GNAT are to be used on operating systems
1631 with limited file name lengths. @ref{3d,,Using gnatkr}.
1633 Of course, no file shortening algorithm can guarantee uniqueness over
1634 all possible unit names; if file name krunching is used, it is your
1635 responsibility to ensure no name clashes occur. Alternatively you
1636 can specify the exact file names that you want used, as described
1637 in the next section. Finally, if your Ada programs are migrating from a
1638 compiler with a different naming convention, you can use the gnatchop
1639 utility to produce source files that follow the GNAT naming conventions.
1640 (For details see @ref{1d,,Renaming Files with gnatchop}.)
1642 Note: in the case of Windows or Mac OS operating systems, case is not
1643 significant. So for example on Windows if the canonical name is
1644 @code{main-sub.adb}, you can use the file name @code{Main-Sub.adb} instead.
1645 However, case is significant for other operating systems, so for example,
1646 if you want to use other than canonically cased file names on a Unix system,
1647 you need to follow the procedures described in the next section.
1649 @node Using Other File Names,Alternative File Naming Schemes,File Naming Rules,File Naming Topics and Utilities
1650 @anchor{gnat_ugn/the_gnat_compilation_model id10}@anchor{3e}@anchor{gnat_ugn/the_gnat_compilation_model using-other-file-names}@anchor{1c}
1651 @subsection Using Other File Names
1656 In the previous section, we have described the default rules used by
1657 GNAT to determine the file name in which a given unit resides. It is
1658 often convenient to follow these default rules, and if you follow them,
1659 the compiler knows without being explicitly told where to find all
1662 @geindex Source_File_Name pragma
1664 However, in some cases, particularly when a program is imported from
1665 another Ada compiler environment, it may be more convenient for the
1666 programmer to specify which file names contain which units. GNAT allows
1667 arbitrary file names to be used by means of the Source_File_Name pragma.
1668 The form of this pragma is as shown in the following examples:
1671 pragma Source_File_Name (My_Utilities.Stacks,
1672 Spec_File_Name => "myutilst_a.ada");
1673 pragma Source_File_name (My_Utilities.Stacks,
1674 Body_File_Name => "myutilst.ada");
1677 As shown in this example, the first argument for the pragma is the unit
1678 name (in this example a child unit). The second argument has the form
1679 of a named association. The identifier
1680 indicates whether the file name is for a spec or a body;
1681 the file name itself is given by a string literal.
1683 The source file name pragma is a configuration pragma, which means that
1684 normally it will be placed in the @code{gnat.adc}
1685 file used to hold configuration
1686 pragmas that apply to a complete compilation environment.
1687 For more details on how the @code{gnat.adc} file is created and used
1688 see @ref{3f,,Handling of Configuration Pragmas}.
1692 GNAT allows completely arbitrary file names to be specified using the
1693 source file name pragma. However, if the file name specified has an
1694 extension other than @code{.ads} or @code{.adb} it is necessary to use
1695 a special syntax when compiling the file. The name in this case must be
1696 preceded by the special sequence @code{-x} followed by a space and the name
1697 of the language, here @code{ada}, as in:
1700 $ gcc -c -x ada peculiar_file_name.sim
1703 @code{gnatmake} handles non-standard file names in the usual manner (the
1704 non-standard file name for the main program is simply used as the
1705 argument to gnatmake). Note that if the extension is also non-standard,
1706 then it must be included in the @code{gnatmake} command, it may not
1709 @node Alternative File Naming Schemes,Handling Arbitrary File Naming Conventions with gnatname,Using Other File Names,File Naming Topics and Utilities
1710 @anchor{gnat_ugn/the_gnat_compilation_model alternative-file-naming-schemes}@anchor{40}@anchor{gnat_ugn/the_gnat_compilation_model id11}@anchor{41}
1711 @subsection Alternative File Naming Schemes
1714 @geindex File naming schemes
1715 @geindex alternative
1719 The previous section described the use of the @code{Source_File_Name}
1720 pragma to allow arbitrary names to be assigned to individual source files.
1721 However, this approach requires one pragma for each file, and especially in
1722 large systems can result in very long @code{gnat.adc} files, and also create
1723 a maintenance problem.
1725 @geindex Source_File_Name pragma
1727 GNAT also provides a facility for specifying systematic file naming schemes
1728 other than the standard default naming scheme previously described. An
1729 alternative scheme for naming is specified by the use of
1730 @code{Source_File_Name} pragmas having the following format:
1733 pragma Source_File_Name (
1734 Spec_File_Name => FILE_NAME_PATTERN
1735 [ , Casing => CASING_SPEC]
1736 [ , Dot_Replacement => STRING_LITERAL ] );
1738 pragma Source_File_Name (
1739 Body_File_Name => FILE_NAME_PATTERN
1740 [ , Casing => CASING_SPEC ]
1741 [ , Dot_Replacement => STRING_LITERAL ] ) ;
1743 pragma Source_File_Name (
1744 Subunit_File_Name => FILE_NAME_PATTERN
1745 [ , Casing => CASING_SPEC ]
1746 [ , Dot_Replacement => STRING_LITERAL ] ) ;
1748 FILE_NAME_PATTERN ::= STRING_LITERAL
1749 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
1752 The @code{FILE_NAME_PATTERN} string shows how the file name is constructed.
1753 It contains a single asterisk character, and the unit name is substituted
1754 systematically for this asterisk. The optional parameter
1755 @code{Casing} indicates
1756 whether the unit name is to be all upper-case letters, all lower-case letters,
1757 or mixed-case. If no
1758 @code{Casing} parameter is used, then the default is all
1761 The optional @code{Dot_Replacement} string is used to replace any periods
1762 that occur in subunit or child unit names. If no @code{Dot_Replacement}
1763 argument is used then separating dots appear unchanged in the resulting
1765 Although the above syntax indicates that the
1766 @code{Casing} argument must appear
1767 before the @code{Dot_Replacement} argument, but it
1768 is also permissible to write these arguments in the opposite order.
1770 As indicated, it is possible to specify different naming schemes for
1771 bodies, specs, and subunits. Quite often the rule for subunits is the
1772 same as the rule for bodies, in which case, there is no need to give
1773 a separate @code{Subunit_File_Name} rule, and in this case the
1774 @code{Body_File_name} rule is used for subunits as well.
1776 The separate rule for subunits can also be used to implement the rather
1777 unusual case of a compilation environment (e.g., a single directory) which
1778 contains a subunit and a child unit with the same unit name. Although
1779 both units cannot appear in the same partition, the Ada Reference Manual
1780 allows (but does not require) the possibility of the two units coexisting
1781 in the same environment.
1783 The file name translation works in the following steps:
1789 If there is a specific @code{Source_File_Name} pragma for the given unit,
1790 then this is always used, and any general pattern rules are ignored.
1793 If there is a pattern type @code{Source_File_Name} pragma that applies to
1794 the unit, then the resulting file name will be used if the file exists. If
1795 more than one pattern matches, the latest one will be tried first, and the
1796 first attempt resulting in a reference to a file that exists will be used.
1799 If no pattern type @code{Source_File_Name} pragma that applies to the unit
1800 for which the corresponding file exists, then the standard GNAT default
1801 naming rules are used.
1804 As an example of the use of this mechanism, consider a commonly used scheme
1805 in which file names are all lower case, with separating periods copied
1806 unchanged to the resulting file name, and specs end with @code{.1.ada}, and
1807 bodies end with @code{.2.ada}. GNAT will follow this scheme if the following
1811 pragma Source_File_Name
1812 (Spec_File_Name => ".1.ada");
1813 pragma Source_File_Name
1814 (Body_File_Name => ".2.ada");
1817 The default GNAT scheme is actually implemented by providing the following
1818 default pragmas internally:
1821 pragma Source_File_Name
1822 (Spec_File_Name => ".ads", Dot_Replacement => "-");
1823 pragma Source_File_Name
1824 (Body_File_Name => ".adb", Dot_Replacement => "-");
1827 Our final example implements a scheme typically used with one of the
1828 Ada 83 compilers, where the separator character for subunits was ‘__’
1829 (two underscores), specs were identified by adding @code{_.ADA}, bodies
1830 by adding @code{.ADA}, and subunits by
1831 adding @code{.SEP}. All file names were
1832 upper case. Child units were not present of course since this was an
1833 Ada 83 compiler, but it seems reasonable to extend this scheme to use
1834 the same double underscore separator for child units.
1837 pragma Source_File_Name
1838 (Spec_File_Name => "_.ADA",
1839 Dot_Replacement => "__",
1840 Casing = Uppercase);
1841 pragma Source_File_Name
1842 (Body_File_Name => ".ADA",
1843 Dot_Replacement => "__",
1844 Casing = Uppercase);
1845 pragma Source_File_Name
1846 (Subunit_File_Name => ".SEP",
1847 Dot_Replacement => "__",
1848 Casing = Uppercase);
1853 @node Handling Arbitrary File Naming Conventions with gnatname,File Name Krunching with gnatkr,Alternative File Naming Schemes,File Naming Topics and Utilities
1854 @anchor{gnat_ugn/the_gnat_compilation_model handling-arbitrary-file-naming-conventions-with-gnatname}@anchor{42}@anchor{gnat_ugn/the_gnat_compilation_model id12}@anchor{43}
1855 @subsection Handling Arbitrary File Naming Conventions with @code{gnatname}
1858 @geindex File Naming Conventions
1861 * Arbitrary File Naming Conventions::
1862 * Running gnatname::
1863 * Switches for gnatname::
1864 * Examples of gnatname Usage::
1868 @node Arbitrary File Naming Conventions,Running gnatname,,Handling Arbitrary File Naming Conventions with gnatname
1869 @anchor{gnat_ugn/the_gnat_compilation_model arbitrary-file-naming-conventions}@anchor{44}@anchor{gnat_ugn/the_gnat_compilation_model id13}@anchor{45}
1870 @subsubsection Arbitrary File Naming Conventions
1873 The GNAT compiler must be able to know the source file name of a compilation
1874 unit. When using the standard GNAT default file naming conventions
1875 (@code{.ads} for specs, @code{.adb} for bodies), the GNAT compiler
1876 does not need additional information.
1878 When the source file names do not follow the standard GNAT default file naming
1879 conventions, the GNAT compiler must be given additional information through
1880 a configuration pragmas file (@ref{25,,Configuration Pragmas})
1882 When the non-standard file naming conventions are well-defined,
1883 a small number of pragmas @code{Source_File_Name} specifying a naming pattern
1884 (@ref{40,,Alternative File Naming Schemes}) may be sufficient. However,
1885 if the file naming conventions are irregular or arbitrary, a number
1886 of pragma @code{Source_File_Name} for individual compilation units
1888 To help maintain the correspondence between compilation unit names and
1889 source file names within the compiler,
1890 GNAT provides a tool @code{gnatname} to generate the required pragmas for a
1893 @node Running gnatname,Switches for gnatname,Arbitrary File Naming Conventions,Handling Arbitrary File Naming Conventions with gnatname
1894 @anchor{gnat_ugn/the_gnat_compilation_model id14}@anchor{46}@anchor{gnat_ugn/the_gnat_compilation_model running-gnatname}@anchor{47}
1895 @subsubsection Running @code{gnatname}
1898 The usual form of the @code{gnatname} command is:
1901 $ gnatname [ switches ] naming_pattern [ naming_patterns ]
1902 [--and [ switches ] naming_pattern [ naming_patterns ]]
1905 All of the arguments are optional. If invoked without any argument,
1906 @code{gnatname} will display its usage.
1908 When used with at least one naming pattern, @code{gnatname} will attempt to
1909 find all the compilation units in files that follow at least one of the
1910 naming patterns. To find these compilation units,
1911 @code{gnatname} will use the GNAT compiler in syntax-check-only mode on all
1914 One or several Naming Patterns may be given as arguments to @code{gnatname}.
1915 Each Naming Pattern is enclosed between double quotes (or single
1917 A Naming Pattern is a regular expression similar to the wildcard patterns
1918 used in file names by the Unix shells or the DOS prompt.
1920 @code{gnatname} may be called with several sections of directories/patterns.
1921 Sections are separated by the switch @code{--and}. In each section, there must be
1922 at least one pattern. If no directory is specified in a section, the current
1923 directory (or the project directory if @code{-P} is used) is implied.
1924 The options other that the directory switches and the patterns apply globally
1925 even if they are in different sections.
1927 Examples of Naming Patterns are:
1935 For a more complete description of the syntax of Naming Patterns,
1936 see the second kind of regular expressions described in @code{g-regexp.ads}
1937 (the ‘Glob’ regular expressions).
1939 When invoked without the switch @code{-P}, @code{gnatname} will create a
1940 configuration pragmas file @code{gnat.adc} in the current working directory,
1941 with pragmas @code{Source_File_Name} for each file that contains a valid Ada
1944 @node Switches for gnatname,Examples of gnatname Usage,Running gnatname,Handling Arbitrary File Naming Conventions with gnatname
1945 @anchor{gnat_ugn/the_gnat_compilation_model id15}@anchor{48}@anchor{gnat_ugn/the_gnat_compilation_model switches-for-gnatname}@anchor{49}
1946 @subsubsection Switches for @code{gnatname}
1949 Switches for @code{gnatname} must precede any specified Naming Pattern.
1951 You may specify any of the following switches to @code{gnatname}:
1953 @geindex --version (gnatname)
1958 @item @code{--version}
1960 Display Copyright and version, then exit disregarding all other options.
1963 @geindex --help (gnatname)
1970 If @code{--version} was not used, display usage, then exit disregarding
1973 @item @code{--subdirs=@emph{dir}}
1975 Real object, library or exec directories are subdirectories <dir> of the
1978 @item @code{--no-backup}
1980 Do not create a backup copy of an existing project file.
1984 Start another section of directories/patterns.
1987 @geindex -c (gnatname)
1992 @item @code{-c@emph{filename}}
1994 Create a configuration pragmas file @code{filename} (instead of the default
1996 There may be zero, one or more space between @code{-c} and
1998 @code{filename} may include directory information. @code{filename} must be
1999 writable. There may be only one switch @code{-c}.
2000 When a switch @code{-c} is
2001 specified, no switch @code{-P} may be specified (see below).
2004 @geindex -d (gnatname)
2009 @item @code{-d@emph{dir}}
2011 Look for source files in directory @code{dir}. There may be zero, one or more
2012 spaces between @code{-d} and @code{dir}.
2013 @code{dir} may end with @code{/**}, that is it may be of the form
2014 @code{root_dir/**}. In this case, the directory @code{root_dir} and all of its
2015 subdirectories, recursively, have to be searched for sources.
2016 When a switch @code{-d}
2017 is specified, the current working directory will not be searched for source
2018 files, unless it is explicitly specified with a @code{-d}
2019 or @code{-D} switch.
2020 Several switches @code{-d} may be specified.
2021 If @code{dir} is a relative path, it is relative to the directory of
2022 the configuration pragmas file specified with switch
2024 or to the directory of the project file specified with switch
2026 if neither switch @code{-c}
2027 nor switch @code{-P} are specified, it is relative to the
2028 current working directory. The directory
2029 specified with switch @code{-d} must exist and be readable.
2032 @geindex -D (gnatname)
2037 @item @code{-D@emph{filename}}
2039 Look for source files in all directories listed in text file @code{filename}.
2040 There may be zero, one or more spaces between @code{-D}
2041 and @code{filename}.
2042 @code{filename} must be an existing, readable text file.
2043 Each nonempty line in @code{filename} must be a directory.
2044 Specifying switch @code{-D} is equivalent to specifying as many
2045 switches @code{-d} as there are nonempty lines in
2050 Follow symbolic links when processing project files.
2052 @geindex -f (gnatname)
2054 @item @code{-f@emph{pattern}}
2056 Foreign patterns. Using this switch, it is possible to add sources of languages
2057 other than Ada to the list of sources of a project file.
2058 It is only useful if a -P switch is used.
2062 gnatname -Pprj -f"*.c" "*.ada"
2065 will look for Ada units in all files with the @code{.ada} extension,
2066 and will add to the list of file for project @code{prj.gpr} the C files
2067 with extension @code{.c}.
2069 @geindex -h (gnatname)
2073 Output usage (help) information. The output is written to @code{stdout}.
2075 @geindex -P (gnatname)
2077 @item @code{-P@emph{proj}}
2079 Create or update project file @code{proj}. There may be zero, one or more space
2080 between @code{-P} and @code{proj}. @code{proj} may include directory
2081 information. @code{proj} must be writable.
2082 There may be only one switch @code{-P}.
2083 When a switch @code{-P} is specified,
2084 no switch @code{-c} may be specified.
2085 On all platforms, except on VMS, when @code{gnatname} is invoked for an
2086 existing project file <proj>.gpr, a backup copy of the project file is created
2087 in the project directory with file name <proj>.gpr.saved_x. ‘x’ is the first
2088 non negative number that makes this backup copy a new file.
2090 @geindex -v (gnatname)
2094 Verbose mode. Output detailed explanation of behavior to @code{stdout}.
2095 This includes name of the file written, the name of the directories to search
2096 and, for each file in those directories whose name matches at least one of
2097 the Naming Patterns, an indication of whether the file contains a unit,
2098 and if so the name of the unit.
2101 @geindex -v -v (gnatname)
2108 Very Verbose mode. In addition to the output produced in verbose mode,
2109 for each file in the searched directories whose name matches none of
2110 the Naming Patterns, an indication is given that there is no match.
2112 @geindex -x (gnatname)
2114 @item @code{-x@emph{pattern}}
2116 Excluded patterns. Using this switch, it is possible to exclude some files
2117 that would match the name patterns. For example,
2120 gnatname -x "*_nt.ada" "*.ada"
2123 will look for Ada units in all files with the @code{.ada} extension,
2124 except those whose names end with @code{_nt.ada}.
2127 @node Examples of gnatname Usage,,Switches for gnatname,Handling Arbitrary File Naming Conventions with gnatname
2128 @anchor{gnat_ugn/the_gnat_compilation_model examples-of-gnatname-usage}@anchor{4a}@anchor{gnat_ugn/the_gnat_compilation_model id16}@anchor{4b}
2129 @subsubsection Examples of @code{gnatname} Usage
2133 $ gnatname -c /home/me/names.adc -d sources "[a-z]*.ada*"
2136 In this example, the directory @code{/home/me} must already exist
2137 and be writable. In addition, the directory
2138 @code{/home/me/sources} (specified by
2139 @code{-d sources}) must exist and be readable.
2141 Note the optional spaces after @code{-c} and @code{-d}.
2144 $ gnatname -P/home/me/proj -x "*_nt_body.ada"
2145 -dsources -dsources/plus -Dcommon_dirs.txt "body_*" "spec_*"
2148 Note that several switches @code{-d} may be used,
2149 even in conjunction with one or several switches
2150 @code{-D}. Several Naming Patterns and one excluded pattern
2151 are used in this example.
2153 @node File Name Krunching with gnatkr,Renaming Files with gnatchop,Handling Arbitrary File Naming Conventions with gnatname,File Naming Topics and Utilities
2154 @anchor{gnat_ugn/the_gnat_compilation_model file-name-krunching-with-gnatkr}@anchor{4c}@anchor{gnat_ugn/the_gnat_compilation_model id17}@anchor{4d}
2155 @subsection File Name Krunching with @code{gnatkr}
2160 This section discusses the method used by the compiler to shorten
2161 the default file names chosen for Ada units so that they do not
2162 exceed the maximum length permitted. It also describes the
2163 @code{gnatkr} utility that can be used to determine the result of
2164 applying this shortening.
2169 * Krunching Method::
2170 * Examples of gnatkr Usage::
2174 @node About gnatkr,Using gnatkr,,File Name Krunching with gnatkr
2175 @anchor{gnat_ugn/the_gnat_compilation_model about-gnatkr}@anchor{4e}@anchor{gnat_ugn/the_gnat_compilation_model id18}@anchor{4f}
2176 @subsubsection About @code{gnatkr}
2179 The default file naming rule in GNAT
2180 is that the file name must be derived from
2181 the unit name. The exact default rule is as follows:
2187 Take the unit name and replace all dots by hyphens.
2190 If such a replacement occurs in the
2191 second character position of a name, and the first character is
2192 @code{a}, @code{g}, @code{s}, or @code{i},
2193 then replace the dot by the character
2197 The reason for this exception is to avoid clashes
2198 with the standard names for children of System, Ada, Interfaces,
2199 and GNAT, which use the prefixes
2200 @code{s-}, @code{a-}, @code{i-}, and @code{g-},
2204 The @code{-gnatk@emph{nn}}
2205 switch of the compiler activates a ‘krunching’
2206 circuit that limits file names to nn characters (where nn is a decimal
2209 The @code{gnatkr} utility can be used to determine the krunched name for
2210 a given file, when krunched to a specified maximum length.
2212 @node Using gnatkr,Krunching Method,About gnatkr,File Name Krunching with gnatkr
2213 @anchor{gnat_ugn/the_gnat_compilation_model id19}@anchor{50}@anchor{gnat_ugn/the_gnat_compilation_model using-gnatkr}@anchor{3d}
2214 @subsubsection Using @code{gnatkr}
2217 The @code{gnatkr} command has the form:
2220 $ gnatkr name [ length ]
2223 @code{name} is the uncrunched file name, derived from the name of the unit
2224 in the standard manner described in the previous section (i.e., in particular
2225 all dots are replaced by hyphens). The file name may or may not have an
2226 extension (defined as a suffix of the form period followed by arbitrary
2227 characters other than period). If an extension is present then it will
2228 be preserved in the output. For example, when krunching @code{hellofile.ads}
2229 to eight characters, the result will be hellofil.ads.
2231 Note: for compatibility with previous versions of @code{gnatkr} dots may
2232 appear in the name instead of hyphens, but the last dot will always be
2233 taken as the start of an extension. So if @code{gnatkr} is given an argument
2234 such as @code{Hello.World.adb} it will be treated exactly as if the first
2235 period had been a hyphen, and for example krunching to eight characters
2236 gives the result @code{hellworl.adb}.
2238 Note that the result is always all lower case.
2239 Characters of the other case are folded as required.
2241 @code{length} represents the length of the krunched name. The default
2242 when no argument is given is 8 characters. A length of zero stands for
2243 unlimited, in other words do not chop except for system files where the
2244 implied crunching length is always eight characters.
2246 The output is the krunched name. The output has an extension only if the
2247 original argument was a file name with an extension.
2249 @node Krunching Method,Examples of gnatkr Usage,Using gnatkr,File Name Krunching with gnatkr
2250 @anchor{gnat_ugn/the_gnat_compilation_model id20}@anchor{51}@anchor{gnat_ugn/the_gnat_compilation_model krunching-method}@anchor{52}
2251 @subsubsection Krunching Method
2254 The initial file name is determined by the name of the unit that the file
2255 contains. The name is formed by taking the full expanded name of the
2256 unit and replacing the separating dots with hyphens and
2258 for all letters, except that a hyphen in the second character position is
2259 replaced by a tilde if the first character is
2260 @code{a}, @code{i}, @code{g}, or @code{s}.
2261 The extension is @code{.ads} for a
2262 spec and @code{.adb} for a body.
2263 Krunching does not affect the extension, but the file name is shortened to
2264 the specified length by following these rules:
2270 The name is divided into segments separated by hyphens, tildes or
2271 underscores and all hyphens, tildes, and underscores are
2272 eliminated. If this leaves the name short enough, we are done.
2275 If the name is too long, the longest segment is located (left-most
2276 if there are two of equal length), and shortened by dropping
2277 its last character. This is repeated until the name is short enough.
2279 As an example, consider the krunching of @code{our-strings-wide_fixed.adb}
2280 to fit the name into 8 characters as required by some operating systems:
2283 our-strings-wide_fixed 22
2284 our strings wide fixed 19
2285 our string wide fixed 18
2286 our strin wide fixed 17
2287 our stri wide fixed 16
2288 our stri wide fixe 15
2289 our str wide fixe 14
2296 Final file name: oustwifi.adb
2300 The file names for all predefined units are always krunched to eight
2301 characters. The krunching of these predefined units uses the following
2302 special prefix replacements:
2305 @multitable {xxxxxxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxxxx}
2349 These system files have a hyphen in the second character position. That
2350 is why normal user files replace such a character with a
2351 tilde, to avoid confusion with system file names.
2353 As an example of this special rule, consider
2354 @code{ada-strings-wide_fixed.adb}, which gets krunched as follows:
2357 ada-strings-wide_fixed 22
2358 a- strings wide fixed 18
2359 a- string wide fixed 17
2360 a- strin wide fixed 16
2361 a- stri wide fixed 15
2362 a- stri wide fixe 14
2369 Final file name: a-stwifi.adb
2373 Of course no file shortening algorithm can guarantee uniqueness over all
2374 possible unit names, and if file name krunching is used then it is your
2375 responsibility to ensure that no name clashes occur. The utility
2376 program @code{gnatkr} is supplied for conveniently determining the
2377 krunched name of a file.
2379 @node Examples of gnatkr Usage,,Krunching Method,File Name Krunching with gnatkr
2380 @anchor{gnat_ugn/the_gnat_compilation_model examples-of-gnatkr-usage}@anchor{53}@anchor{gnat_ugn/the_gnat_compilation_model id21}@anchor{54}
2381 @subsubsection Examples of @code{gnatkr} Usage
2385 $ gnatkr very_long_unit_name.ads --> velounna.ads
2386 $ gnatkr grandparent-parent-child.ads --> grparchi.ads
2387 $ gnatkr Grandparent.Parent.Child.ads --> grparchi.ads
2388 $ gnatkr grandparent-parent-child --> grparchi
2389 $ gnatkr very_long_unit_name.ads/count=6 --> vlunna.ads
2390 $ gnatkr very_long_unit_name.ads/count=0 --> very_long_unit_name.ads
2393 @node Renaming Files with gnatchop,,File Name Krunching with gnatkr,File Naming Topics and Utilities
2394 @anchor{gnat_ugn/the_gnat_compilation_model id22}@anchor{55}@anchor{gnat_ugn/the_gnat_compilation_model renaming-files-with-gnatchop}@anchor{1d}
2395 @subsection Renaming Files with @code{gnatchop}
2400 This section discusses how to handle files with multiple units by using
2401 the @code{gnatchop} utility. This utility is also useful in renaming
2402 files to meet the standard GNAT default file naming conventions.
2405 * Handling Files with Multiple Units::
2406 * Operating gnatchop in Compilation Mode::
2407 * Command Line for gnatchop::
2408 * Switches for gnatchop::
2409 * Examples of gnatchop Usage::
2413 @node Handling Files with Multiple Units,Operating gnatchop in Compilation Mode,,Renaming Files with gnatchop
2414 @anchor{gnat_ugn/the_gnat_compilation_model handling-files-with-multiple-units}@anchor{56}@anchor{gnat_ugn/the_gnat_compilation_model id23}@anchor{57}
2415 @subsubsection Handling Files with Multiple Units
2418 The basic compilation model of GNAT requires that a file submitted to the
2419 compiler have only one unit and there be a strict correspondence
2420 between the file name and the unit name.
2422 If you want to keep your files with multiple units,
2423 perhaps to maintain compatibility with some other Ada compilation system,
2424 you can use @code{gnatname} to generate or update your project files.
2425 Generated or modified project files can be processed by GNAT.
2427 See @ref{42,,Handling Arbitrary File Naming Conventions with gnatname}
2428 for more details on how to use @cite{gnatname}.
2430 Alternatively, if you want to permanently restructure a set of ‘foreign’
2431 files so that they match the GNAT rules, and do the remaining development
2432 using the GNAT structure, you can simply use @code{gnatchop} once, generate the
2433 new set of files and work with them from that point on.
2435 Note that if your file containing multiple units starts with a byte order
2436 mark (BOM) specifying UTF-8 encoding, then the files generated by gnatchop
2437 will each start with a copy of this BOM, meaning that they can be compiled
2438 automatically in UTF-8 mode without needing to specify an explicit encoding.
2440 @node Operating gnatchop in Compilation Mode,Command Line for gnatchop,Handling Files with Multiple Units,Renaming Files with gnatchop
2441 @anchor{gnat_ugn/the_gnat_compilation_model id24}@anchor{58}@anchor{gnat_ugn/the_gnat_compilation_model operating-gnatchop-in-compilation-mode}@anchor{59}
2442 @subsubsection Operating gnatchop in Compilation Mode
2445 The basic function of @code{gnatchop} is to take a file with multiple units
2446 and split it into separate files. The boundary between files is reasonably
2447 clear, except for the issue of comments and pragmas. In default mode, the
2448 rule is that any pragmas between units belong to the previous unit, except
2449 that configuration pragmas always belong to the following unit. Any comments
2450 belong to the following unit. These rules
2451 almost always result in the right choice of
2452 the split point without needing to mark it explicitly and most users will
2453 find this default to be what they want. In this default mode it is incorrect to
2454 submit a file containing only configuration pragmas, or one that ends in
2455 configuration pragmas, to @code{gnatchop}.
2457 However, using a special option to activate ‘compilation mode’,
2459 can perform another function, which is to provide exactly the semantics
2460 required by the RM for handling of configuration pragmas in a compilation.
2461 In the absence of configuration pragmas (at the main file level), this
2462 option has no effect, but it causes such configuration pragmas to be handled
2463 in a quite different manner.
2465 First, in compilation mode, if @code{gnatchop} is given a file that consists of
2466 only configuration pragmas, then this file is appended to the
2467 @code{gnat.adc} file in the current directory. This behavior provides
2468 the required behavior described in the RM for the actions to be taken
2469 on submitting such a file to the compiler, namely that these pragmas
2470 should apply to all subsequent compilations in the same compilation
2471 environment. Using GNAT, the current directory, possibly containing a
2472 @code{gnat.adc} file is the representation
2473 of a compilation environment. For more information on the
2474 @code{gnat.adc} file, see @ref{3f,,Handling of Configuration Pragmas}.
2476 Second, in compilation mode, if @code{gnatchop}
2477 is given a file that starts with
2478 configuration pragmas, and contains one or more units, then these
2479 configuration pragmas are prepended to each of the chopped files. This
2480 behavior provides the required behavior described in the RM for the
2481 actions to be taken on compiling such a file, namely that the pragmas
2482 apply to all units in the compilation, but not to subsequently compiled
2485 Finally, if configuration pragmas appear between units, they are appended
2486 to the previous unit. This results in the previous unit being illegal,
2487 since the compiler does not accept configuration pragmas that follow
2488 a unit. This provides the required RM behavior that forbids configuration
2489 pragmas other than those preceding the first compilation unit of a
2492 For most purposes, @code{gnatchop} will be used in default mode. The
2493 compilation mode described above is used only if you need exactly
2494 accurate behavior with respect to compilations, and you have files
2495 that contain multiple units and configuration pragmas. In this
2496 circumstance the use of @code{gnatchop} with the compilation mode
2497 switch provides the required behavior, and is for example the mode
2498 in which GNAT processes the ACVC tests.
2500 @node Command Line for gnatchop,Switches for gnatchop,Operating gnatchop in Compilation Mode,Renaming Files with gnatchop
2501 @anchor{gnat_ugn/the_gnat_compilation_model command-line-for-gnatchop}@anchor{5a}@anchor{gnat_ugn/the_gnat_compilation_model id25}@anchor{5b}
2502 @subsubsection Command Line for @code{gnatchop}
2505 The @code{gnatchop} command has the form:
2508 $ gnatchop switches file_name [file_name ...]
2512 The only required argument is the file name of the file to be chopped.
2513 There are no restrictions on the form of this file name. The file itself
2514 contains one or more Ada units, in normal GNAT format, concatenated
2515 together. As shown, more than one file may be presented to be chopped.
2517 When run in default mode, @code{gnatchop} generates one output file in
2518 the current directory for each unit in each of the files.
2520 @code{directory}, if specified, gives the name of the directory to which
2521 the output files will be written. If it is not specified, all files are
2522 written to the current directory.
2524 For example, given a
2525 file called @code{hellofiles} containing
2530 with Ada.Text_IO; use Ada.Text_IO;
2540 $ gnatchop hellofiles
2543 generates two files in the current directory, one called
2544 @code{hello.ads} containing the single line that is the procedure spec,
2545 and the other called @code{hello.adb} containing the remaining text. The
2546 original file is not affected. The generated files can be compiled in
2549 When gnatchop is invoked on a file that is empty or that contains only empty
2550 lines and/or comments, gnatchop will not fail, but will not produce any
2553 For example, given a
2554 file called @code{toto.txt} containing
2566 will not produce any new file and will result in the following warnings:
2569 toto.txt:1:01: warning: empty file, contains no compilation units
2570 no compilation units found
2571 no source files written
2574 @node Switches for gnatchop,Examples of gnatchop Usage,Command Line for gnatchop,Renaming Files with gnatchop
2575 @anchor{gnat_ugn/the_gnat_compilation_model id26}@anchor{5c}@anchor{gnat_ugn/the_gnat_compilation_model switches-for-gnatchop}@anchor{5d}
2576 @subsubsection Switches for @code{gnatchop}
2579 @code{gnatchop} recognizes the following switches:
2581 @geindex --version (gnatchop)
2586 @item @code{--version}
2588 Display Copyright and version, then exit disregarding all other options.
2591 @geindex --help (gnatchop)
2598 If @code{--version} was not used, display usage, then exit disregarding
2602 @geindex -c (gnatchop)
2609 Causes @code{gnatchop} to operate in compilation mode, in which
2610 configuration pragmas are handled according to strict RM rules. See
2611 previous section for a full description of this mode.
2613 @item @code{-gnat@emph{xxx}}
2615 This passes the given @code{-gnat@emph{xxx}} switch to @code{gnat} which is
2616 used to parse the given file. Not all @emph{xxx} options make sense,
2617 but for example, the use of @code{-gnati2} allows @code{gnatchop} to
2618 process a source file that uses Latin-2 coding for identifiers.
2622 Causes @code{gnatchop} to generate a brief help summary to the standard
2623 output file showing usage information.
2626 @geindex -k (gnatchop)
2631 @item @code{-k@emph{mm}}
2633 Limit generated file names to the specified number @code{mm}
2635 This is useful if the
2636 resulting set of files is required to be interoperable with systems
2637 which limit the length of file names.
2638 No space is allowed between the @code{-k} and the numeric value. The numeric
2639 value may be omitted in which case a default of @code{-k8},
2641 with DOS-like file systems, is used. If no @code{-k} switch
2643 there is no limit on the length of file names.
2646 @geindex -p (gnatchop)
2653 Causes the file modification time stamp of the input file to be
2654 preserved and used for the time stamp of the output file(s). This may be
2655 useful for preserving coherency of time stamps in an environment where
2656 @code{gnatchop} is used as part of a standard build process.
2659 @geindex -q (gnatchop)
2666 Causes output of informational messages indicating the set of generated
2667 files to be suppressed. Warnings and error messages are unaffected.
2670 @geindex -r (gnatchop)
2672 @geindex Source_Reference pragmas
2679 Generate @code{Source_Reference} pragmas. Use this switch if the output
2680 files are regarded as temporary and development is to be done in terms
2681 of the original unchopped file. This switch causes
2682 @code{Source_Reference} pragmas to be inserted into each of the
2683 generated files to refers back to the original file name and line number.
2684 The result is that all error messages refer back to the original
2686 In addition, the debugging information placed into the object file (when
2687 the @code{-g} switch of @code{gcc} or @code{gnatmake} is
2689 also refers back to this original file so that tools like profilers and
2690 debuggers will give information in terms of the original unchopped file.
2692 If the original file to be chopped itself contains
2693 a @code{Source_Reference}
2694 pragma referencing a third file, then gnatchop respects
2695 this pragma, and the generated @code{Source_Reference} pragmas
2696 in the chopped file refer to the original file, with appropriate
2697 line numbers. This is particularly useful when @code{gnatchop}
2698 is used in conjunction with @code{gnatprep} to compile files that
2699 contain preprocessing statements and multiple units.
2702 @geindex -v (gnatchop)
2709 Causes @code{gnatchop} to operate in verbose mode. The version
2710 number and copyright notice are output, as well as exact copies of
2711 the gnat1 commands spawned to obtain the chop control information.
2714 @geindex -w (gnatchop)
2721 Overwrite existing file names. Normally @code{gnatchop} regards it as a
2722 fatal error if there is already a file with the same name as a
2723 file it would otherwise output, in other words if the files to be
2724 chopped contain duplicated units. This switch bypasses this
2725 check, and causes all but the last instance of such duplicated
2726 units to be skipped.
2729 @geindex --GCC= (gnatchop)
2734 @item @code{--GCC=@emph{xxxx}}
2736 Specify the path of the GNAT parser to be used. When this switch is used,
2737 no attempt is made to add the prefix to the GNAT parser executable.
2740 @node Examples of gnatchop Usage,,Switches for gnatchop,Renaming Files with gnatchop
2741 @anchor{gnat_ugn/the_gnat_compilation_model examples-of-gnatchop-usage}@anchor{5e}@anchor{gnat_ugn/the_gnat_compilation_model id27}@anchor{5f}
2742 @subsubsection Examples of @code{gnatchop} Usage
2746 $ gnatchop -w hello_s.ada prerelease/files
2749 Chops the source file @code{hello_s.ada}. The output files will be
2750 placed in the directory @code{prerelease/files},
2752 files with matching names in that directory (no files in the current
2753 directory are modified).
2759 Chops the source file @code{archive}
2760 into the current directory. One
2761 useful application of @code{gnatchop} is in sending sets of sources
2762 around, for example in email messages. The required sources are simply
2763 concatenated (for example, using a Unix @code{cat}
2765 @code{gnatchop} is used at the other end to reconstitute the original
2769 $ gnatchop file1 file2 file3 direc
2772 Chops all units in files @code{file1}, @code{file2}, @code{file3}, placing
2773 the resulting files in the directory @code{direc}. Note that if any units
2774 occur more than once anywhere within this set of files, an error message
2775 is generated, and no files are written. To override this check, use the
2777 in which case the last occurrence in the last file will
2778 be the one that is output, and earlier duplicate occurrences for a given
2779 unit will be skipped.
2781 @node Configuration Pragmas,Generating Object Files,File Naming Topics and Utilities,The GNAT Compilation Model
2782 @anchor{gnat_ugn/the_gnat_compilation_model configuration-pragmas}@anchor{25}@anchor{gnat_ugn/the_gnat_compilation_model id28}@anchor{60}
2783 @section Configuration Pragmas
2786 @geindex Configuration pragmas
2789 @geindex configuration
2791 Configuration pragmas include those pragmas described as
2792 such in the Ada Reference Manual, as well as
2793 implementation-dependent pragmas that are configuration pragmas.
2794 See the @code{Implementation_Defined_Pragmas} chapter in the
2795 @cite{GNAT_Reference_Manual} for details on these
2796 additional GNAT-specific configuration pragmas.
2797 Most notably, the pragma @code{Source_File_Name}, which allows
2798 specifying non-default names for source files, is a configuration
2799 pragma. The following is a complete list of configuration pragmas
2809 Allow_Integer_Address
2812 Assume_No_Invalid_Values
2814 Check_Float_Overflow
2818 Compile_Time_Warning
2820 Compiler_Unit_Warning
2822 Convention_Identifier
2825 Default_Scalar_Storage_Order
2826 Default_Storage_Pool
2827 Disable_Atomic_Synchronization
2831 Enable_Atomic_Synchronization
2834 External_Name_Casing
2843 No_Component_Reordering
2844 No_Heap_Finalization
2850 Overriding_Renamings
2851 Partition_Elaboration_Policy
2853 Prefix_Exception_Messages
2854 Priority_Specific_Dispatching
2857 Propagate_Exceptions
2864 Restrictions_Warnings
2866 Short_Circuit_And_Or
2869 Source_File_Name_Project
2873 Suppress_Exception_Locations
2874 Task_Dispatching_Policy
2875 Unevaluated_Use_Of_Old
2882 Wide_Character_Encoding
2886 * Handling of Configuration Pragmas::
2887 * The Configuration Pragmas Files::
2891 @node Handling of Configuration Pragmas,The Configuration Pragmas Files,,Configuration Pragmas
2892 @anchor{gnat_ugn/the_gnat_compilation_model handling-of-configuration-pragmas}@anchor{3f}@anchor{gnat_ugn/the_gnat_compilation_model id29}@anchor{61}
2893 @subsection Handling of Configuration Pragmas
2896 Configuration pragmas may either appear at the start of a compilation
2897 unit, or they can appear in a configuration pragma file to apply to
2898 all compilations performed in a given compilation environment.
2900 GNAT also provides the @code{gnatchop} utility to provide an automatic
2901 way to handle configuration pragmas following the semantics for
2902 compilations (that is, files with multiple units), described in the RM.
2903 See @ref{59,,Operating gnatchop in Compilation Mode} for details.
2904 However, for most purposes, it will be more convenient to edit the
2905 @code{gnat.adc} file that contains configuration pragmas directly,
2906 as described in the following section.
2908 In the case of @code{Restrictions} pragmas appearing as configuration
2909 pragmas in individual compilation units, the exact handling depends on
2910 the type of restriction.
2912 Restrictions that require partition-wide consistency (like
2913 @code{No_Tasking}) are
2914 recognized wherever they appear
2915 and can be freely inherited, e.g. from a @emph{with}ed unit to the @emph{with}ing
2916 unit. This makes sense since the binder will in any case insist on seeing
2917 consistent use, so any unit not conforming to any restrictions that are
2918 anywhere in the partition will be rejected, and you might as well find
2919 that out at compile time rather than at bind time.
2921 For restrictions that do not require partition-wide consistency, e.g.
2922 SPARK or No_Implementation_Attributes, in general the restriction applies
2923 only to the unit in which the pragma appears, and not to any other units.
2925 The exception is No_Elaboration_Code which always applies to the entire
2926 object file from a compilation, i.e. to the body, spec, and all subunits.
2927 This restriction can be specified in a configuration pragma file, or it
2928 can be on the body and/or the spec (in eithe case it applies to all the
2929 relevant units). It can appear on a subunit only if it has previously
2930 appeared in the body of spec.
2932 @node The Configuration Pragmas Files,,Handling of Configuration Pragmas,Configuration Pragmas
2933 @anchor{gnat_ugn/the_gnat_compilation_model id30}@anchor{62}@anchor{gnat_ugn/the_gnat_compilation_model the-configuration-pragmas-files}@anchor{63}
2934 @subsection The Configuration Pragmas Files
2939 In GNAT a compilation environment is defined by the current
2940 directory at the time that a compile command is given. This current
2941 directory is searched for a file whose name is @code{gnat.adc}. If
2942 this file is present, it is expected to contain one or more
2943 configuration pragmas that will be applied to the current compilation.
2944 However, if the switch @code{-gnatA} is used, @code{gnat.adc} is not
2945 considered. When taken into account, @code{gnat.adc} is added to the
2946 dependencies, so that if @code{gnat.adc} is modified later, an invocation of
2947 @code{gnatmake} will recompile the source.
2949 Configuration pragmas may be entered into the @code{gnat.adc} file
2950 either by running @code{gnatchop} on a source file that consists only of
2951 configuration pragmas, or more conveniently by direct editing of the
2952 @code{gnat.adc} file, which is a standard format source file.
2954 Besides @code{gnat.adc}, additional files containing configuration
2955 pragmas may be applied to the current compilation using the switch
2956 @code{-gnatec=@emph{path}} where @code{path} must designate an existing file that
2957 contains only configuration pragmas. These configuration pragmas are
2958 in addition to those found in @code{gnat.adc} (provided @code{gnat.adc}
2959 is present and switch @code{-gnatA} is not used).
2961 It is allowable to specify several switches @code{-gnatec=}, all of which
2962 will be taken into account.
2964 Files containing configuration pragmas specified with switches
2965 @code{-gnatec=} are added to the dependencies, unless they are
2966 temporary files. A file is considered temporary if its name ends in
2967 @code{.tmp} or @code{.TMP}. Certain tools follow this naming
2968 convention because they pass information to @code{gcc} via
2969 temporary files that are immediately deleted; it doesn’t make sense to
2970 depend on a file that no longer exists. Such tools include
2971 @code{gprbuild}, @code{gnatmake}, and @code{gnatcheck}.
2973 By default, configuration pragma files are stored by their absolute paths in
2974 ALI files. You can use the @code{-gnateb} switch in order to store them by
2975 their basename instead.
2977 If you are using project file, a separate mechanism is provided using
2981 @c See :ref:`Specifying_Configuration_Pragmas` for more details.
2983 @node Generating Object Files,Source Dependencies,Configuration Pragmas,The GNAT Compilation Model
2984 @anchor{gnat_ugn/the_gnat_compilation_model generating-object-files}@anchor{26}@anchor{gnat_ugn/the_gnat_compilation_model id31}@anchor{64}
2985 @section Generating Object Files
2988 An Ada program consists of a set of source files, and the first step in
2989 compiling the program is to generate the corresponding object files.
2990 These are generated by compiling a subset of these source files.
2991 The files you need to compile are the following:
2997 If a package spec has no body, compile the package spec to produce the
2998 object file for the package.
3001 If a package has both a spec and a body, compile the body to produce the
3002 object file for the package. The source file for the package spec need
3003 not be compiled in this case because there is only one object file, which
3004 contains the code for both the spec and body of the package.
3007 For a subprogram, compile the subprogram body to produce the object file
3008 for the subprogram. The spec, if one is present, is as usual in a
3009 separate file, and need not be compiled.
3018 In the case of subunits, only compile the parent unit. A single object
3019 file is generated for the entire subunit tree, which includes all the
3023 Compile child units independently of their parent units
3024 (though, of course, the spec of all the ancestor unit must be present in order
3025 to compile a child unit).
3030 Compile generic units in the same manner as any other units. The object
3031 files in this case are small dummy files that contain at most the
3032 flag used for elaboration checking. This is because GNAT always handles generic
3033 instantiation by means of macro expansion. However, it is still necessary to
3034 compile generic units, for dependency checking and elaboration purposes.
3037 The preceding rules describe the set of files that must be compiled to
3038 generate the object files for a program. Each object file has the same
3039 name as the corresponding source file, except that the extension is
3042 You may wish to compile other files for the purpose of checking their
3043 syntactic and semantic correctness. For example, in the case where a
3044 package has a separate spec and body, you would not normally compile the
3045 spec. However, it is convenient in practice to compile the spec to make
3046 sure it is error-free before compiling clients of this spec, because such
3047 compilations will fail if there is an error in the spec.
3049 GNAT provides an option for compiling such files purely for the
3050 purposes of checking correctness; such compilations are not required as
3051 part of the process of building a program. To compile a file in this
3052 checking mode, use the @code{-gnatc} switch.
3054 @node Source Dependencies,The Ada Library Information Files,Generating Object Files,The GNAT Compilation Model
3055 @anchor{gnat_ugn/the_gnat_compilation_model id32}@anchor{65}@anchor{gnat_ugn/the_gnat_compilation_model source-dependencies}@anchor{27}
3056 @section Source Dependencies
3059 A given object file clearly depends on the source file which is compiled
3060 to produce it. Here we are using “depends” in the sense of a typical
3061 @code{make} utility; in other words, an object file depends on a source
3062 file if changes to the source file require the object file to be
3064 In addition to this basic dependency, a given object may depend on
3065 additional source files as follows:
3071 If a file being compiled @emph{with}s a unit @code{X}, the object file
3072 depends on the file containing the spec of unit @code{X}. This includes
3073 files that are @emph{with}ed implicitly either because they are parents
3074 of @emph{with}ed child units or they are run-time units required by the
3075 language constructs used in a particular unit.
3078 If a file being compiled instantiates a library level generic unit, the
3079 object file depends on both the spec and body files for this generic
3083 If a file being compiled instantiates a generic unit defined within a
3084 package, the object file depends on the body file for the package as
3085 well as the spec file.
3090 @geindex -gnatn switch
3096 If a file being compiled contains a call to a subprogram for which
3097 pragma @code{Inline} applies and inlining is activated with the
3098 @code{-gnatn} switch, the object file depends on the file containing the
3099 body of this subprogram as well as on the file containing the spec. Note
3100 that for inlining to actually occur as a result of the use of this switch,
3101 it is necessary to compile in optimizing mode.
3103 @geindex -gnatN switch
3105 The use of @code{-gnatN} activates inlining optimization
3106 that is performed by the front end of the compiler. This inlining does
3107 not require that the code generation be optimized. Like @code{-gnatn},
3108 the use of this switch generates additional dependencies.
3110 When using a gcc-based back end, then the use of
3111 @code{-gnatN} is deprecated, and the use of @code{-gnatn} is preferred.
3112 Historically front end inlining was more extensive than the gcc back end
3113 inlining, but that is no longer the case.
3116 If an object file @code{O} depends on the proper body of a subunit through
3117 inlining or instantiation, it depends on the parent unit of the subunit.
3118 This means that any modification of the parent unit or one of its subunits
3119 affects the compilation of @code{O}.
3122 The object file for a parent unit depends on all its subunit body files.
3125 The previous two rules meant that for purposes of computing dependencies and
3126 recompilation, a body and all its subunits are treated as an indivisible whole.
3128 These rules are applied transitively: if unit @code{A} @emph{with}s
3129 unit @code{B}, whose elaboration calls an inlined procedure in package
3130 @code{C}, the object file for unit @code{A} will depend on the body of
3131 @code{C}, in file @code{c.adb}.
3133 The set of dependent files described by these rules includes all the
3134 files on which the unit is semantically dependent, as dictated by the
3135 Ada language standard. However, it is a superset of what the
3136 standard describes, because it includes generic, inline, and subunit
3139 An object file must be recreated by recompiling the corresponding source
3140 file if any of the source files on which it depends are modified. For
3141 example, if the @code{make} utility is used to control compilation,
3142 the rule for an Ada object file must mention all the source files on
3143 which the object file depends, according to the above definition.
3144 The determination of the necessary
3145 recompilations is done automatically when one uses @code{gnatmake}.
3148 @node The Ada Library Information Files,Binding an Ada Program,Source Dependencies,The GNAT Compilation Model
3149 @anchor{gnat_ugn/the_gnat_compilation_model id33}@anchor{66}@anchor{gnat_ugn/the_gnat_compilation_model the-ada-library-information-files}@anchor{28}
3150 @section The Ada Library Information Files
3153 @geindex Ada Library Information files
3157 Each compilation actually generates two output files. The first of these
3158 is the normal object file that has a @code{.o} extension. The second is a
3159 text file containing full dependency information. It has the same
3160 name as the source file, but an @code{.ali} extension.
3161 This file is known as the Ada Library Information (@code{ALI}) file.
3162 The following information is contained in the @code{ALI} file.
3168 Version information (indicates which version of GNAT was used to compile
3169 the unit(s) in question)
3172 Main program information (including priority and time slice settings,
3173 as well as the wide character encoding used during compilation).
3176 List of arguments used in the @code{gcc} command for the compilation
3179 Attributes of the unit, including configuration pragmas used, an indication
3180 of whether the compilation was successful, exception model used etc.
3183 A list of relevant restrictions applying to the unit (used for consistency)
3187 Categorization information (e.g., use of pragma @code{Pure}).
3190 Information on all @emph{with}ed units, including presence of
3191 @code{Elaborate} or @code{Elaborate_All} pragmas.
3194 Information from any @code{Linker_Options} pragmas used in the unit
3197 Information on the use of @code{Body_Version} or @code{Version}
3198 attributes in the unit.
3201 Dependency information. This is a list of files, together with
3202 time stamp and checksum information. These are files on which
3203 the unit depends in the sense that recompilation is required
3204 if any of these units are modified.
3207 Cross-reference data. Contains information on all entities referenced
3208 in the unit. Used by tools like @code{gnatxref} and @code{gnatfind} to
3209 provide cross-reference information.
3212 For a full detailed description of the format of the @code{ALI} file,
3213 see the source of the body of unit @code{Lib.Writ}, contained in file
3214 @code{lib-writ.adb} in the GNAT compiler sources.
3216 @node Binding an Ada Program,GNAT and Libraries,The Ada Library Information Files,The GNAT Compilation Model
3217 @anchor{gnat_ugn/the_gnat_compilation_model binding-an-ada-program}@anchor{29}@anchor{gnat_ugn/the_gnat_compilation_model id34}@anchor{67}
3218 @section Binding an Ada Program
3221 When using languages such as C and C++, once the source files have been
3222 compiled the only remaining step in building an executable program
3223 is linking the object modules together. This means that it is possible to
3224 link an inconsistent version of a program, in which two units have
3225 included different versions of the same header.
3227 The rules of Ada do not permit such an inconsistent program to be built.
3228 For example, if two clients have different versions of the same package,
3229 it is illegal to build a program containing these two clients.
3230 These rules are enforced by the GNAT binder, which also determines an
3231 elaboration order consistent with the Ada rules.
3233 The GNAT binder is run after all the object files for a program have
3234 been created. It is given the name of the main program unit, and from
3235 this it determines the set of units required by the program, by reading the
3236 corresponding ALI files. It generates error messages if the program is
3237 inconsistent or if no valid order of elaboration exists.
3239 If no errors are detected, the binder produces a main program, in Ada by
3240 default, that contains calls to the elaboration procedures of those
3241 compilation unit that require them, followed by
3242 a call to the main program. This Ada program is compiled to generate the
3243 object file for the main program. The name of
3244 the Ada file is @code{b~xxx}.adb` (with the corresponding spec
3245 @code{b~xxx}.ads`) where @code{xxx} is the name of the
3248 Finally, the linker is used to build the resulting executable program,
3249 using the object from the main program from the bind step as well as the
3250 object files for the Ada units of the program.
3252 @node GNAT and Libraries,Conditional Compilation,Binding an Ada Program,The GNAT Compilation Model
3253 @anchor{gnat_ugn/the_gnat_compilation_model gnat-and-libraries}@anchor{2a}@anchor{gnat_ugn/the_gnat_compilation_model id35}@anchor{68}
3254 @section GNAT and Libraries
3257 @geindex Library building and using
3259 This section describes how to build and use libraries with GNAT, and also shows
3260 how to recompile the GNAT run-time library. You should be familiar with the
3261 Project Manager facility (see the @emph{GNAT_Project_Manager} chapter of the
3262 @emph{GPRbuild User’s Guide}) before reading this chapter.
3265 * Introduction to Libraries in GNAT::
3266 * General Ada Libraries::
3267 * Stand-alone Ada Libraries::
3268 * Rebuilding the GNAT Run-Time Library::
3272 @node Introduction to Libraries in GNAT,General Ada Libraries,,GNAT and Libraries
3273 @anchor{gnat_ugn/the_gnat_compilation_model id36}@anchor{69}@anchor{gnat_ugn/the_gnat_compilation_model introduction-to-libraries-in-gnat}@anchor{6a}
3274 @subsection Introduction to Libraries in GNAT
3277 A library is, conceptually, a collection of objects which does not have its
3278 own main thread of execution, but rather provides certain services to the
3279 applications that use it. A library can be either statically linked with the
3280 application, in which case its code is directly included in the application,
3281 or, on platforms that support it, be dynamically linked, in which case
3282 its code is shared by all applications making use of this library.
3284 GNAT supports both types of libraries.
3285 In the static case, the compiled code can be provided in different ways. The
3286 simplest approach is to provide directly the set of objects resulting from
3287 compilation of the library source files. Alternatively, you can group the
3288 objects into an archive using whatever commands are provided by the operating
3289 system. For the latter case, the objects are grouped into a shared library.
3291 In the GNAT environment, a library has three types of components:
3300 @code{ALI} files (see @ref{28,,The Ada Library Information Files}), and
3303 Object files, an archive or a shared library.
3306 A GNAT library may expose all its source files, which is useful for
3307 documentation purposes. Alternatively, it may expose only the units needed by
3308 an external user to make use of the library. That is to say, the specs
3309 reflecting the library services along with all the units needed to compile
3310 those specs, which can include generic bodies or any body implementing an
3311 inlined routine. In the case of @emph{stand-alone libraries} those exposed
3312 units are called @emph{interface units} (@ref{6b,,Stand-alone Ada Libraries}).
3314 All compilation units comprising an application, including those in a library,
3315 need to be elaborated in an order partially defined by Ada’s semantics. GNAT
3316 computes the elaboration order from the @code{ALI} files and this is why they
3317 constitute a mandatory part of GNAT libraries.
3318 @emph{Stand-alone libraries} are the exception to this rule because a specific
3319 library elaboration routine is produced independently of the application(s)
3322 @node General Ada Libraries,Stand-alone Ada Libraries,Introduction to Libraries in GNAT,GNAT and Libraries
3323 @anchor{gnat_ugn/the_gnat_compilation_model general-ada-libraries}@anchor{6c}@anchor{gnat_ugn/the_gnat_compilation_model id37}@anchor{6d}
3324 @subsection General Ada Libraries
3328 * Building a library::
3329 * Installing a library::
3334 @node Building a library,Installing a library,,General Ada Libraries
3335 @anchor{gnat_ugn/the_gnat_compilation_model building-a-library}@anchor{6e}@anchor{gnat_ugn/the_gnat_compilation_model id38}@anchor{6f}
3336 @subsubsection Building a library
3339 The easiest way to build a library is to use the Project Manager,
3340 which supports a special type of project called a @emph{Library Project}
3341 (see the @emph{Library Projects} section in the @emph{GNAT Project Manager}
3342 chapter of the @emph{GPRbuild User’s Guide}).
3344 A project is considered a library project, when two project-level attributes
3345 are defined in it: @code{Library_Name} and @code{Library_Dir}. In order to
3346 control different aspects of library configuration, additional optional
3347 project-level attributes can be specified:
3356 @item @code{Library_Kind}
3358 This attribute controls whether the library is to be static or dynamic
3365 @item @code{Library_Version}
3367 This attribute specifies the library version; this value is used
3368 during dynamic linking of shared libraries to determine if the currently
3369 installed versions of the binaries are compatible.
3373 @code{Library_Options}
3379 @item @code{Library_GCC}
3381 These attributes specify additional low-level options to be used during
3382 library generation, and redefine the actual application used to generate
3387 The GNAT Project Manager takes full care of the library maintenance task,
3388 including recompilation of the source files for which objects do not exist
3389 or are not up to date, assembly of the library archive, and installation of
3390 the library (i.e., copying associated source, object and @code{ALI} files
3391 to the specified location).
3393 Here is a simple library project file:
3397 for Source_Dirs use ("src1", "src2");
3398 for Object_Dir use "obj";
3399 for Library_Name use "mylib";
3400 for Library_Dir use "lib";
3401 for Library_Kind use "dynamic";
3405 and the compilation command to build and install the library:
3411 It is not entirely trivial to perform manually all the steps required to
3412 produce a library. We recommend that you use the GNAT Project Manager
3413 for this task. In special cases where this is not desired, the necessary
3414 steps are discussed below.
3416 There are various possibilities for compiling the units that make up the
3417 library: for example with a Makefile (@ref{70,,Using the GNU make Utility}) or
3418 with a conventional script. For simple libraries, it is also possible to create
3419 a dummy main program which depends upon all the packages that comprise the
3420 interface of the library. This dummy main program can then be given to
3421 @code{gnatmake}, which will ensure that all necessary objects are built.
3423 After this task is accomplished, you should follow the standard procedure
3424 of the underlying operating system to produce the static or shared library.
3426 Here is an example of such a dummy program:
3429 with My_Lib.Service1;
3430 with My_Lib.Service2;
3431 with My_Lib.Service3;
3432 procedure My_Lib_Dummy is
3438 Here are the generic commands that will build an archive or a shared library.
3441 # compiling the library
3442 $ gnatmake -c my_lib_dummy.adb
3444 # we don't need the dummy object itself
3445 $ rm my_lib_dummy.o my_lib_dummy.ali
3447 # create an archive with the remaining objects
3448 $ ar rc libmy_lib.a *.o
3449 # some systems may require "ranlib" to be run as well
3451 # or create a shared library
3452 $ gcc -shared -o libmy_lib.so *.o
3453 # some systems may require the code to have been compiled with -fPIC
3455 # remove the object files that are now in the library
3458 # Make the ALI files read-only so that gnatmake will not try to
3459 # regenerate the objects that are in the library
3463 Please note that the library must have a name of the form @code{lib@emph{xxx}.a}
3464 or @code{lib@emph{xxx}.so} (or @code{lib@emph{xxx}.dll} on Windows) in order to
3465 be accessed by the directive @code{-l@emph{xxx}} at link time.
3467 @node Installing a library,Using a library,Building a library,General Ada Libraries
3468 @anchor{gnat_ugn/the_gnat_compilation_model id39}@anchor{71}@anchor{gnat_ugn/the_gnat_compilation_model installing-a-library}@anchor{72}
3469 @subsubsection Installing a library
3472 @geindex ADA_PROJECT_PATH
3474 @geindex GPR_PROJECT_PATH
3476 If you use project files, library installation is part of the library build
3477 process (see the @emph{Installing a Library with Project Files} section of the
3478 @emph{GNAT Project Manager} chapter of the @emph{GPRbuild User’s Guide}).
3480 When project files are not an option, it is also possible, but not recommended,
3481 to install the library so that the sources needed to use the library are on the
3482 Ada source path and the ALI files & libraries be on the Ada Object path (see
3483 @ref{73,,Search Paths and the Run-Time Library (RTL)}. Alternatively, the system
3484 administrator can place general-purpose libraries in the default compiler
3485 paths, by specifying the libraries’ location in the configuration files
3486 @code{ada_source_path} and @code{ada_object_path}. These configuration files
3487 must be located in the GNAT installation tree at the same place as the gcc spec
3488 file. The location of the gcc spec file can be determined as follows:
3494 The configuration files mentioned above have a simple format: each line
3495 must contain one unique directory name.
3496 Those names are added to the corresponding path
3497 in their order of appearance in the file. The names can be either absolute
3498 or relative; in the latter case, they are relative to where theses files
3501 The files @code{ada_source_path} and @code{ada_object_path} might not be
3503 GNAT installation, in which case, GNAT will look for its run-time library in
3504 the directories @code{adainclude} (for the sources) and @code{adalib} (for the
3505 objects and @code{ALI} files). When the files exist, the compiler does not
3506 look in @code{adainclude} and @code{adalib}, and thus the
3507 @code{ada_source_path} file
3508 must contain the location for the GNAT run-time sources (which can simply
3509 be @code{adainclude}). In the same way, the @code{ada_object_path} file must
3510 contain the location for the GNAT run-time objects (which can simply
3513 You can also specify a new default path to the run-time library at compilation
3514 time with the switch @code{--RTS=rts-path}. You can thus choose / change
3515 the run-time library you want your program to be compiled with. This switch is
3516 recognized by @code{gcc}, @code{gnatmake}, @code{gnatbind},
3517 @code{gnatls}, @code{gnatfind} and @code{gnatxref}.
3519 It is possible to install a library before or after the standard GNAT
3520 library, by reordering the lines in the configuration files. In general, a
3521 library must be installed before the GNAT library if it redefines
3524 @node Using a library,,Installing a library,General Ada Libraries
3525 @anchor{gnat_ugn/the_gnat_compilation_model id40}@anchor{74}@anchor{gnat_ugn/the_gnat_compilation_model using-a-library}@anchor{75}
3526 @subsubsection Using a library
3529 Once again, the project facility greatly simplifies the use of
3530 libraries. In this context, using a library is just a matter of adding a
3531 @emph{with} clause in the user project. For instance, to make use of the
3532 library @code{My_Lib} shown in examples in earlier sections, you can
3542 Even if you have a third-party, non-Ada library, you can still use GNAT’s
3543 Project Manager facility to provide a wrapper for it. For example, the
3544 following project, when @emph{with}ed by your main project, will link with the
3545 third-party library @code{liba.a}:
3549 for Externally_Built use "true";
3550 for Source_Files use ();
3551 for Library_Dir use "lib";
3552 for Library_Name use "a";
3553 for Library_Kind use "static";
3557 This is an alternative to the use of @code{pragma Linker_Options}. It is
3558 especially interesting in the context of systems with several interdependent
3559 static libraries where finding a proper linker order is not easy and best be
3560 left to the tools having visibility over project dependence information.
3562 In order to use an Ada library manually, you need to make sure that this
3563 library is on both your source and object path
3564 (see @ref{73,,Search Paths and the Run-Time Library (RTL)}
3565 and @ref{76,,Search Paths for gnatbind}). Furthermore, when the objects are grouped
3566 in an archive or a shared library, you need to specify the desired
3567 library at link time.
3569 For example, you can use the library @code{mylib} installed in
3570 @code{/dir/my_lib_src} and @code{/dir/my_lib_obj} with the following commands:
3573 $ gnatmake -aI/dir/my_lib_src -aO/dir/my_lib_obj my_appl \\
3577 This can be expressed more simply:
3583 when the following conditions are met:
3589 @code{/dir/my_lib_src} has been added by the user to the environment
3591 @geindex ADA_INCLUDE_PATH
3592 @geindex environment variable; ADA_INCLUDE_PATH
3593 @code{ADA_INCLUDE_PATH}, or by the administrator to the file
3594 @code{ada_source_path}
3597 @code{/dir/my_lib_obj} has been added by the user to the environment
3599 @geindex ADA_OBJECTS_PATH
3600 @geindex environment variable; ADA_OBJECTS_PATH
3601 @code{ADA_OBJECTS_PATH}, or by the administrator to the file
3602 @code{ada_object_path}
3605 a pragma @code{Linker_Options} has been added to one of the sources.
3609 pragma Linker_Options ("-lmy_lib");
3613 Note that you may also load a library dynamically at
3614 run time given its filename, as illustrated in the GNAT @code{plugins} example
3615 in the directory @code{share/examples/gnat/plugins} within the GNAT
3618 @node Stand-alone Ada Libraries,Rebuilding the GNAT Run-Time Library,General Ada Libraries,GNAT and Libraries
3619 @anchor{gnat_ugn/the_gnat_compilation_model id41}@anchor{77}@anchor{gnat_ugn/the_gnat_compilation_model stand-alone-ada-libraries}@anchor{6b}
3620 @subsection Stand-alone Ada Libraries
3623 @geindex Stand-alone libraries
3626 * Introduction to Stand-alone Libraries::
3627 * Building a Stand-alone Library::
3628 * Creating a Stand-alone Library to be used in a non-Ada context::
3629 * Restrictions in Stand-alone Libraries::
3633 @node Introduction to Stand-alone Libraries,Building a Stand-alone Library,,Stand-alone Ada Libraries
3634 @anchor{gnat_ugn/the_gnat_compilation_model id42}@anchor{78}@anchor{gnat_ugn/the_gnat_compilation_model introduction-to-stand-alone-libraries}@anchor{79}
3635 @subsubsection Introduction to Stand-alone Libraries
3638 A Stand-alone Library (abbreviated ‘SAL’) is a library that contains the
3640 elaborate the Ada units that are included in the library. In contrast with
3641 an ordinary library, which consists of all sources, objects and @code{ALI}
3643 library, a SAL may specify a restricted subset of compilation units
3644 to serve as a library interface. In this case, the fully
3645 self-sufficient set of files will normally consist of an objects
3646 archive, the sources of interface units’ specs, and the @code{ALI}
3647 files of interface units.
3648 If an interface spec contains a generic unit or an inlined subprogram,
3650 source must also be provided; if the units that must be provided in the source
3651 form depend on other units, the source and @code{ALI} files of those must
3654 The main purpose of a SAL is to minimize the recompilation overhead of client
3655 applications when a new version of the library is installed. Specifically,
3656 if the interface sources have not changed, client applications do not need to
3657 be recompiled. If, furthermore, a SAL is provided in the shared form and its
3658 version, controlled by @code{Library_Version} attribute, is not changed,
3659 then the clients do not need to be relinked.
3661 SALs also allow the library providers to minimize the amount of library source
3662 text exposed to the clients. Such ‘information hiding’ might be useful or
3663 necessary for various reasons.
3665 Stand-alone libraries are also well suited to be used in an executable whose
3666 main routine is not written in Ada.
3668 @node Building a Stand-alone Library,Creating a Stand-alone Library to be used in a non-Ada context,Introduction to Stand-alone Libraries,Stand-alone Ada Libraries
3669 @anchor{gnat_ugn/the_gnat_compilation_model building-a-stand-alone-library}@anchor{7a}@anchor{gnat_ugn/the_gnat_compilation_model id43}@anchor{7b}
3670 @subsubsection Building a Stand-alone Library
3673 GNAT’s Project facility provides a simple way of building and installing
3674 stand-alone libraries; see the @emph{Stand-alone Library Projects} section
3675 in the @emph{GNAT Project Manager} chapter of the @emph{GPRbuild User’s Guide}.
3676 To be a Stand-alone Library Project, in addition to the two attributes
3677 that make a project a Library Project (@code{Library_Name} and
3678 @code{Library_Dir}; see the @emph{Library Projects} section in the
3679 @emph{GNAT Project Manager} chapter of the @emph{GPRbuild User’s Guide}),
3680 the attribute @code{Library_Interface} must be defined. For example:
3683 for Library_Dir use "lib_dir";
3684 for Library_Name use "dummy";
3685 for Library_Interface use ("int1", "int1.child");
3688 Attribute @code{Library_Interface} has a non-empty string list value,
3689 each string in the list designating a unit contained in an immediate source
3690 of the project file.
3692 When a Stand-alone Library is built, first the binder is invoked to build
3693 a package whose name depends on the library name
3694 (@code{b~dummy.ads/b} in the example above).
3695 This binder-generated package includes initialization and
3696 finalization procedures whose
3697 names depend on the library name (@code{dummyinit} and @code{dummyfinal}
3699 above). The object corresponding to this package is included in the library.
3701 You must ensure timely (e.g., prior to any use of interfaces in the SAL)
3702 calling of these procedures if a static SAL is built, or if a shared SAL
3704 with the project-level attribute @code{Library_Auto_Init} set to
3707 For a Stand-Alone Library, only the @code{ALI} files of the Interface Units
3708 (those that are listed in attribute @code{Library_Interface}) are copied to
3709 the Library Directory. As a consequence, only the Interface Units may be
3710 imported from Ada units outside of the library. If other units are imported,
3711 the binding phase will fail.
3713 It is also possible to build an encapsulated library where not only
3714 the code to elaborate and finalize the library is embedded but also
3715 ensuring that the library is linked only against static
3716 libraries. So an encapsulated library only depends on system
3717 libraries, all other code, including the GNAT runtime, is embedded. To
3718 build an encapsulated library the attribute
3719 @code{Library_Standalone} must be set to @code{encapsulated}:
3722 for Library_Dir use "lib_dir";
3723 for Library_Name use "dummy";
3724 for Library_Kind use "dynamic";
3725 for Library_Interface use ("int1", "int1.child");
3726 for Library_Standalone use "encapsulated";
3729 The default value for this attribute is @code{standard} in which case
3730 a stand-alone library is built.
3732 The attribute @code{Library_Src_Dir} may be specified for a
3733 Stand-Alone Library. @code{Library_Src_Dir} is a simple attribute that has a
3734 single string value. Its value must be the path (absolute or relative to the
3735 project directory) of an existing directory. This directory cannot be the
3736 object directory or one of the source directories, but it can be the same as
3737 the library directory. The sources of the Interface
3738 Units of the library that are needed by an Ada client of the library will be
3739 copied to the designated directory, called the Interface Copy directory.
3740 These sources include the specs of the Interface Units, but they may also
3741 include bodies and subunits, when pragmas @code{Inline} or @code{Inline_Always}
3742 are used, or when there is a generic unit in the spec. Before the sources
3743 are copied to the Interface Copy directory, an attempt is made to delete all
3744 files in the Interface Copy directory.
3746 Building stand-alone libraries by hand is somewhat tedious, but for those
3747 occasions when it is necessary here are the steps that you need to perform:
3753 Compile all library sources.
3756 Invoke the binder with the switch @code{-n} (No Ada main program),
3757 with all the @code{ALI} files of the interfaces, and
3758 with the switch @code{-L} to give specific names to the @code{init}
3759 and @code{final} procedures. For example:
3762 $ gnatbind -n int1.ali int2.ali -Lsal1
3766 Compile the binder generated file:
3773 Link the dynamic library with all the necessary object files,
3774 indicating to the linker the names of the @code{init} (and possibly
3775 @code{final}) procedures for automatic initialization (and finalization).
3776 The built library should be placed in a directory different from
3777 the object directory.
3780 Copy the @code{ALI} files of the interface to the library directory,
3781 add in this copy an indication that it is an interface to a SAL
3782 (i.e., add a word @code{SL} on the line in the @code{ALI} file that starts
3783 with letter ‘P’) and make the modified copy of the @code{ALI} file
3787 Using SALs is not different from using other libraries
3788 (see @ref{75,,Using a library}).
3790 @node Creating a Stand-alone Library to be used in a non-Ada context,Restrictions in Stand-alone Libraries,Building a Stand-alone Library,Stand-alone Ada Libraries
3791 @anchor{gnat_ugn/the_gnat_compilation_model creating-a-stand-alone-library-to-be-used-in-a-non-ada-context}@anchor{7c}@anchor{gnat_ugn/the_gnat_compilation_model id44}@anchor{7d}
3792 @subsubsection Creating a Stand-alone Library to be used in a non-Ada context
3795 It is easy to adapt the SAL build procedure discussed above for use of a SAL in
3798 The only extra step required is to ensure that library interface subprograms
3799 are compatible with the main program, by means of @code{pragma Export}
3800 or @code{pragma Convention}.
3802 Here is an example of simple library interface for use with C main program:
3805 package My_Package is
3807 procedure Do_Something;
3808 pragma Export (C, Do_Something, "do_something");
3810 procedure Do_Something_Else;
3811 pragma Export (C, Do_Something_Else, "do_something_else");
3816 On the foreign language side, you must provide a ‘foreign’ view of the
3817 library interface; remember that it should contain elaboration routines in
3818 addition to interface subprograms.
3820 The example below shows the content of @code{mylib_interface.h} (note
3821 that there is no rule for the naming of this file, any name can be used)
3824 /* the library elaboration procedure */
3825 extern void mylibinit (void);
3827 /* the library finalization procedure */
3828 extern void mylibfinal (void);
3830 /* the interface exported by the library */
3831 extern void do_something (void);
3832 extern void do_something_else (void);
3835 Libraries built as explained above can be used from any program, provided
3836 that the elaboration procedures (named @code{mylibinit} in the previous
3837 example) are called before the library services are used. Any number of
3838 libraries can be used simultaneously, as long as the elaboration
3839 procedure of each library is called.
3841 Below is an example of a C program that uses the @code{mylib} library.
3844 #include "mylib_interface.h"
3849 /* First, elaborate the library before using it */
3852 /* Main program, using the library exported entities */
3854 do_something_else ();
3856 /* Library finalization at the end of the program */
3862 Note that invoking any library finalization procedure generated by
3863 @code{gnatbind} shuts down the Ada run-time environment.
3865 finalization of all Ada libraries must be performed at the end of the program.
3866 No call to these libraries or to the Ada run-time library should be made
3867 after the finalization phase.
3869 Note also that special care must be taken with multi-tasks
3870 applications. The initialization and finalization routines are not
3871 protected against concurrent access. If such requirement is needed it
3872 must be ensured at the application level using a specific operating
3873 system services like a mutex or a critical-section.
3875 @node Restrictions in Stand-alone Libraries,,Creating a Stand-alone Library to be used in a non-Ada context,Stand-alone Ada Libraries
3876 @anchor{gnat_ugn/the_gnat_compilation_model id45}@anchor{7e}@anchor{gnat_ugn/the_gnat_compilation_model restrictions-in-stand-alone-libraries}@anchor{7f}
3877 @subsubsection Restrictions in Stand-alone Libraries
3880 The pragmas listed below should be used with caution inside libraries,
3881 as they can create incompatibilities with other Ada libraries:
3887 pragma @code{Locking_Policy}
3890 pragma @code{Partition_Elaboration_Policy}
3893 pragma @code{Queuing_Policy}
3896 pragma @code{Task_Dispatching_Policy}
3899 pragma @code{Unreserve_All_Interrupts}
3902 When using a library that contains such pragmas, the user must make sure
3903 that all libraries use the same pragmas with the same values. Otherwise,
3904 @code{Program_Error} will
3905 be raised during the elaboration of the conflicting
3906 libraries. The usage of these pragmas and its consequences for the user
3907 should therefore be well documented.
3909 Similarly, the traceback in the exception occurrence mechanism should be
3910 enabled or disabled in a consistent manner across all libraries.
3911 Otherwise, Program_Error will be raised during the elaboration of the
3912 conflicting libraries.
3914 If the @code{Version} or @code{Body_Version}
3915 attributes are used inside a library, then you need to
3916 perform a @code{gnatbind} step that specifies all @code{ALI} files in all
3917 libraries, so that version identifiers can be properly computed.
3918 In practice these attributes are rarely used, so this is unlikely
3919 to be a consideration.
3921 @node Rebuilding the GNAT Run-Time Library,,Stand-alone Ada Libraries,GNAT and Libraries
3922 @anchor{gnat_ugn/the_gnat_compilation_model id46}@anchor{80}@anchor{gnat_ugn/the_gnat_compilation_model rebuilding-the-gnat-run-time-library}@anchor{81}
3923 @subsection Rebuilding the GNAT Run-Time Library
3926 @geindex GNAT Run-Time Library
3929 @geindex Building the GNAT Run-Time Library
3931 @geindex Rebuilding the GNAT Run-Time Library
3933 @geindex Run-Time Library
3936 It may be useful to recompile the GNAT library in various debugging or
3937 experimentation contexts. A project file called
3938 @code{libada.gpr} is provided to that effect and can be found in
3939 the directory containing the GNAT library. The location of this
3940 directory depends on the way the GNAT environment has been installed and can
3941 be determined by means of the command:
3947 The last entry in the source search path usually contains the
3948 gnat library (the @code{adainclude} directory). This project file contains its
3949 own documentation and in particular the set of instructions needed to rebuild a
3950 new library and to use it.
3952 Note that rebuilding the GNAT Run-Time is only recommended for temporary
3953 experiments or debugging, and is not supported.
3955 @geindex Conditional compilation
3957 @node Conditional Compilation,Mixed Language Programming,GNAT and Libraries,The GNAT Compilation Model
3958 @anchor{gnat_ugn/the_gnat_compilation_model conditional-compilation}@anchor{2b}@anchor{gnat_ugn/the_gnat_compilation_model id47}@anchor{82}
3959 @section Conditional Compilation
3962 This section presents some guidelines for modeling conditional compilation in Ada and describes the
3963 gnatprep preprocessor utility.
3965 @geindex Conditional compilation
3968 * Modeling Conditional Compilation in Ada::
3969 * Preprocessing with gnatprep::
3970 * Integrated Preprocessing::
3974 @node Modeling Conditional Compilation in Ada,Preprocessing with gnatprep,,Conditional Compilation
3975 @anchor{gnat_ugn/the_gnat_compilation_model id48}@anchor{83}@anchor{gnat_ugn/the_gnat_compilation_model modeling-conditional-compilation-in-ada}@anchor{84}
3976 @subsection Modeling Conditional Compilation in Ada
3979 It is often necessary to arrange for a single source program
3980 to serve multiple purposes, where it is compiled in different
3981 ways to achieve these different goals. Some examples of the
3982 need for this feature are
3988 Adapting a program to a different hardware environment
3991 Adapting a program to a different target architecture
3994 Turning debugging features on and off
3997 Arranging for a program to compile with different compilers
4000 In C, or C++, the typical approach would be to use the preprocessor
4001 that is defined as part of the language. The Ada language does not
4002 contain such a feature. This is not an oversight, but rather a very
4003 deliberate design decision, based on the experience that overuse of
4004 the preprocessing features in C and C++ can result in programs that
4005 are extremely difficult to maintain. For example, if we have ten
4006 switches that can be on or off, this means that there are a thousand
4007 separate programs, any one of which might not even be syntactically
4008 correct, and even if syntactically correct, the resulting program
4009 might not work correctly. Testing all combinations can quickly become
4012 Nevertheless, the need to tailor programs certainly exists, and in
4013 this section we will discuss how this can
4014 be achieved using Ada in general, and GNAT in particular.
4017 * Use of Boolean Constants::
4018 * Debugging - A Special Case::
4019 * Conditionalizing Declarations::
4020 * Use of Alternative Implementations::
4025 @node Use of Boolean Constants,Debugging - A Special Case,,Modeling Conditional Compilation in Ada
4026 @anchor{gnat_ugn/the_gnat_compilation_model id49}@anchor{85}@anchor{gnat_ugn/the_gnat_compilation_model use-of-boolean-constants}@anchor{86}
4027 @subsubsection Use of Boolean Constants
4030 In the case where the difference is simply which code
4031 sequence is executed, the cleanest solution is to use Boolean
4032 constants to control which code is executed.
4035 FP_Initialize_Required : constant Boolean := True;
4037 if FP_Initialize_Required then
4042 Not only will the code inside the @code{if} statement not be executed if
4043 the constant Boolean is @code{False}, but it will also be completely
4044 deleted from the program.
4045 However, the code is only deleted after the @code{if} statement
4046 has been checked for syntactic and semantic correctness.
4047 (In contrast, with preprocessors the code is deleted before the
4048 compiler ever gets to see it, so it is not checked until the switch
4051 @geindex Preprocessors (contrasted with conditional compilation)
4053 Typically the Boolean constants will be in a separate package,
4058 FP_Initialize_Required : constant Boolean := True;
4059 Reset_Available : constant Boolean := False;
4064 The @code{Config} package exists in multiple forms for the various targets,
4065 with an appropriate script selecting the version of @code{Config} needed.
4066 Then any other unit requiring conditional compilation can do a @emph{with}
4067 of @code{Config} to make the constants visible.
4069 @node Debugging - A Special Case,Conditionalizing Declarations,Use of Boolean Constants,Modeling Conditional Compilation in Ada
4070 @anchor{gnat_ugn/the_gnat_compilation_model debugging-a-special-case}@anchor{87}@anchor{gnat_ugn/the_gnat_compilation_model id50}@anchor{88}
4071 @subsubsection Debugging - A Special Case
4074 A common use of conditional code is to execute statements (for example
4075 dynamic checks, or output of intermediate results) under control of a
4076 debug switch, so that the debugging behavior can be turned on and off.
4077 This can be done using a Boolean constant to control whether the code
4082 Put_Line ("got to the first stage!");
4089 if Debugging and then Temperature > 999.0 then
4090 raise Temperature_Crazy;
4094 @geindex pragma Assert
4096 Since this is a common case, there are special features to deal with
4097 this in a convenient manner. For the case of tests, Ada 2005 has added
4098 a pragma @code{Assert} that can be used for such tests. This pragma is modeled
4099 on the @code{Assert} pragma that has always been available in GNAT, so this
4100 feature may be used with GNAT even if you are not using Ada 2005 features.
4101 The use of pragma @code{Assert} is described in the
4102 @cite{GNAT_Reference_Manual}, but as an
4103 example, the last test could be written:
4106 pragma Assert (Temperature <= 999.0, "Temperature Crazy");
4112 pragma Assert (Temperature <= 999.0);
4115 In both cases, if assertions are active and the temperature is excessive,
4116 the exception @code{Assert_Failure} will be raised, with the given string in
4117 the first case or a string indicating the location of the pragma in the second
4118 case used as the exception message.
4120 @geindex pragma Assertion_Policy
4122 You can turn assertions on and off by using the @code{Assertion_Policy}
4125 @geindex -gnata switch
4127 This is an Ada 2005 pragma which is implemented in all modes by
4128 GNAT. Alternatively, you can use the @code{-gnata} switch
4129 to enable assertions from the command line, which applies to
4130 all versions of Ada.
4132 @geindex pragma Debug
4134 For the example above with the @code{Put_Line}, the GNAT-specific pragma
4135 @code{Debug} can be used:
4138 pragma Debug (Put_Line ("got to the first stage!"));
4141 If debug pragmas are enabled, the argument, which must be of the form of
4142 a procedure call, is executed (in this case, @code{Put_Line} will be called).
4143 Only one call can be present, but of course a special debugging procedure
4144 containing any code you like can be included in the program and then
4145 called in a pragma @code{Debug} argument as needed.
4147 One advantage of pragma @code{Debug} over the @code{if Debugging then}
4148 construct is that pragma @code{Debug} can appear in declarative contexts,
4149 such as at the very beginning of a procedure, before local declarations have
4152 @geindex pragma Debug_Policy
4154 Debug pragmas are enabled using either the @code{-gnata} switch that also
4155 controls assertions, or with a separate Debug_Policy pragma.
4157 The latter pragma is new in the Ada 2005 versions of GNAT (but it can be used
4158 in Ada 95 and Ada 83 programs as well), and is analogous to
4159 pragma @code{Assertion_Policy} to control assertions.
4161 @code{Assertion_Policy} and @code{Debug_Policy} are configuration pragmas,
4162 and thus they can appear in @code{gnat.adc} if you are not using a
4163 project file, or in the file designated to contain configuration pragmas
4165 They then apply to all subsequent compilations. In practice the use of
4166 the @code{-gnata} switch is often the most convenient method of controlling
4167 the status of these pragmas.
4169 Note that a pragma is not a statement, so in contexts where a statement
4170 sequence is required, you can’t just write a pragma on its own. You have
4171 to add a @code{null} statement.
4175 ... -- some statements
4177 pragma Assert (Num_Cases < 10);
4182 @node Conditionalizing Declarations,Use of Alternative Implementations,Debugging - A Special Case,Modeling Conditional Compilation in Ada
4183 @anchor{gnat_ugn/the_gnat_compilation_model conditionalizing-declarations}@anchor{89}@anchor{gnat_ugn/the_gnat_compilation_model id51}@anchor{8a}
4184 @subsubsection Conditionalizing Declarations
4187 In some cases it may be necessary to conditionalize declarations to meet
4188 different requirements. For example we might want a bit string whose length
4189 is set to meet some hardware message requirement.
4191 This may be possible using declare blocks controlled
4192 by conditional constants:
4195 if Small_Machine then
4197 X : Bit_String (1 .. 10);
4203 X : Large_Bit_String (1 .. 1000);
4210 Note that in this approach, both declarations are analyzed by the
4211 compiler so this can only be used where both declarations are legal,
4212 even though one of them will not be used.
4214 Another approach is to define integer constants, e.g., @code{Bits_Per_Word},
4215 or Boolean constants, e.g., @code{Little_Endian}, and then write declarations
4216 that are parameterized by these constants. For example
4220 Field1 at 0 range Boolean'Pos (Little_Endian) * 10 .. Bits_Per_Word;
4224 If @code{Bits_Per_Word} is set to 32, this generates either
4228 Field1 at 0 range 0 .. 32;
4232 for the big endian case, or
4236 Field1 at 0 range 10 .. 32;
4240 for the little endian case. Since a powerful subset of Ada expression
4241 notation is usable for creating static constants, clever use of this
4242 feature can often solve quite difficult problems in conditionalizing
4243 compilation (note incidentally that in Ada 95, the little endian
4244 constant was introduced as @code{System.Default_Bit_Order}, so you do not
4245 need to define this one yourself).
4247 @node Use of Alternative Implementations,Preprocessing,Conditionalizing Declarations,Modeling Conditional Compilation in Ada
4248 @anchor{gnat_ugn/the_gnat_compilation_model id52}@anchor{8b}@anchor{gnat_ugn/the_gnat_compilation_model use-of-alternative-implementations}@anchor{8c}
4249 @subsubsection Use of Alternative Implementations
4252 In some cases, none of the approaches described above are adequate. This
4253 can occur for example if the set of declarations required is radically
4254 different for two different configurations.
4256 In this situation, the official Ada way of dealing with conditionalizing
4257 such code is to write separate units for the different cases. As long as
4258 this does not result in excessive duplication of code, this can be done
4259 without creating maintenance problems. The approach is to share common
4260 code as far as possible, and then isolate the code and declarations
4261 that are different. Subunits are often a convenient method for breaking
4262 out a piece of a unit that is to be conditionalized, with separate files
4263 for different versions of the subunit for different targets, where the
4264 build script selects the right one to give to the compiler.
4266 @geindex Subunits (and conditional compilation)
4268 As an example, consider a situation where a new feature in Ada 2005
4269 allows something to be done in a really nice way. But your code must be able
4270 to compile with an Ada 95 compiler. Conceptually you want to say:
4274 ... neat Ada 2005 code
4276 ... not quite as neat Ada 95 code
4280 where @code{Ada_2005} is a Boolean constant.
4282 But this won’t work when @code{Ada_2005} is set to @code{False},
4283 since the @code{then} clause will be illegal for an Ada 95 compiler.
4284 (Recall that although such unreachable code would eventually be deleted
4285 by the compiler, it still needs to be legal. If it uses features
4286 introduced in Ada 2005, it will be illegal in Ada 95.)
4291 procedure Insert is separate;
4294 Then we have two files for the subunit @code{Insert}, with the two sets of
4296 If the package containing this is called @code{File_Queries}, then we might
4303 @code{file_queries-insert-2005.adb}
4306 @code{file_queries-insert-95.adb}
4309 and the build script renames the appropriate file to @code{file_queries-insert.adb} and then carries out the compilation.
4311 This can also be done with project files’ naming schemes. For example:
4314 for body ("File_Queries.Insert") use "file_queries-insert-2005.ada";
4317 Note also that with project files it is desirable to use a different extension
4318 than @code{ads} / @code{adb} for alternative versions. Otherwise a naming
4319 conflict may arise through another commonly used feature: to declare as part
4320 of the project a set of directories containing all the sources obeying the
4321 default naming scheme.
4323 The use of alternative units is certainly feasible in all situations,
4324 and for example the Ada part of the GNAT run-time is conditionalized
4325 based on the target architecture using this approach. As a specific example,
4326 consider the implementation of the AST feature in VMS. There is one
4327 spec: @code{s-asthan.ads} which is the same for all architectures, and three
4337 @item @code{s-asthan.adb}
4339 used for all non-VMS operating systems
4346 @item @code{s-asthan-vms-alpha.adb}
4348 used for VMS on the Alpha
4355 @item @code{s-asthan-vms-ia64.adb}
4357 used for VMS on the ia64
4361 The dummy version @code{s-asthan.adb} simply raises exceptions noting that
4362 this operating system feature is not available, and the two remaining
4363 versions interface with the corresponding versions of VMS to provide
4364 VMS-compatible AST handling. The GNAT build script knows the architecture
4365 and operating system, and automatically selects the right version,
4366 renaming it if necessary to @code{s-asthan.adb} before the run-time build.
4368 Another style for arranging alternative implementations is through Ada’s
4369 access-to-subprogram facility.
4370 In case some functionality is to be conditionally included,
4371 you can declare an access-to-procedure variable @code{Ref} that is initialized
4372 to designate a ‘do nothing’ procedure, and then invoke @code{Ref.all}
4374 In some library package, set @code{Ref} to @code{Proc'Access} for some
4375 procedure @code{Proc} that performs the relevant processing.
4376 The initialization only occurs if the library package is included in the
4378 The same idea can also be implemented using tagged types and dispatching
4381 @node Preprocessing,,Use of Alternative Implementations,Modeling Conditional Compilation in Ada
4382 @anchor{gnat_ugn/the_gnat_compilation_model id53}@anchor{8d}@anchor{gnat_ugn/the_gnat_compilation_model preprocessing}@anchor{8e}
4383 @subsubsection Preprocessing
4386 @geindex Preprocessing
4388 Although it is quite possible to conditionalize code without the use of
4389 C-style preprocessing, as described earlier in this section, it is
4390 nevertheless convenient in some cases to use the C approach. Moreover,
4391 older Ada compilers have often provided some preprocessing capability,
4392 so legacy code may depend on this approach, even though it is not
4395 To accommodate such use, GNAT provides a preprocessor (modeled to a large
4396 extent on the various preprocessors that have been used
4397 with legacy code on other compilers, to enable easier transition).
4401 The preprocessor may be used in two separate modes. It can be used quite
4402 separately from the compiler, to generate a separate output source file
4403 that is then fed to the compiler as a separate step. This is the
4404 @code{gnatprep} utility, whose use is fully described in
4405 @ref{8f,,Preprocessing with gnatprep}.
4407 The preprocessing language allows such constructs as
4410 #if DEBUG or else (PRIORITY > 4) then
4411 sequence of declarations
4413 completely different sequence of declarations
4417 The values of the symbols @code{DEBUG} and @code{PRIORITY} can be
4418 defined either on the command line or in a separate file.
4420 The other way of running the preprocessor is even closer to the C style and
4421 often more convenient. In this approach the preprocessing is integrated into
4422 the compilation process. The compiler is given the preprocessor input which
4423 includes @code{#if} lines etc, and then the compiler carries out the
4424 preprocessing internally and processes the resulting output.
4425 For more details on this approach, see @ref{90,,Integrated Preprocessing}.
4427 @node Preprocessing with gnatprep,Integrated Preprocessing,Modeling Conditional Compilation in Ada,Conditional Compilation
4428 @anchor{gnat_ugn/the_gnat_compilation_model id54}@anchor{91}@anchor{gnat_ugn/the_gnat_compilation_model preprocessing-with-gnatprep}@anchor{8f}
4429 @subsection Preprocessing with @code{gnatprep}
4434 @geindex Preprocessing (gnatprep)
4436 This section discusses how to use GNAT’s @code{gnatprep} utility for simple
4438 Although designed for use with GNAT, @code{gnatprep} does not depend on any
4439 special GNAT features.
4440 For further discussion of conditional compilation in general, see
4441 @ref{2b,,Conditional Compilation}.
4444 * Preprocessing Symbols::
4446 * Switches for gnatprep::
4447 * Form of Definitions File::
4448 * Form of Input Text for gnatprep::
4452 @node Preprocessing Symbols,Using gnatprep,,Preprocessing with gnatprep
4453 @anchor{gnat_ugn/the_gnat_compilation_model id55}@anchor{92}@anchor{gnat_ugn/the_gnat_compilation_model preprocessing-symbols}@anchor{93}
4454 @subsubsection Preprocessing Symbols
4457 Preprocessing symbols are defined in @emph{definition files} and referenced in the
4458 sources to be preprocessed. A preprocessing symbol is an identifier, following
4459 normal Ada (case-insensitive) rules for its syntax, with the restriction that
4460 all characters need to be in the ASCII set (no accented letters).
4462 @node Using gnatprep,Switches for gnatprep,Preprocessing Symbols,Preprocessing with gnatprep
4463 @anchor{gnat_ugn/the_gnat_compilation_model id56}@anchor{94}@anchor{gnat_ugn/the_gnat_compilation_model using-gnatprep}@anchor{95}
4464 @subsubsection Using @code{gnatprep}
4467 To call @code{gnatprep} use:
4470 $ gnatprep [ switches ] infile outfile [ deffile ]
4482 @item @emph{switches}
4484 is an optional sequence of switches as described in the next section.
4493 is the full name of the input file, which is an Ada source
4494 file containing preprocessor directives.
4501 @item @emph{outfile}
4503 is the full name of the output file, which is an Ada source
4504 in standard Ada form. When used with GNAT, this file name will
4505 normally have an @code{ads} or @code{adb} suffix.
4512 @item @code{deffile}
4514 is the full name of a text file containing definitions of
4515 preprocessing symbols to be referenced by the preprocessor. This argument is
4516 optional, and can be replaced by the use of the @code{-D} switch.
4520 @node Switches for gnatprep,Form of Definitions File,Using gnatprep,Preprocessing with gnatprep
4521 @anchor{gnat_ugn/the_gnat_compilation_model id57}@anchor{96}@anchor{gnat_ugn/the_gnat_compilation_model switches-for-gnatprep}@anchor{97}
4522 @subsubsection Switches for @code{gnatprep}
4525 @geindex --version (gnatprep)
4530 @item @code{--version}
4532 Display Copyright and version, then exit disregarding all other options.
4535 @geindex --help (gnatprep)
4542 If @code{--version} was not used, display usage and then exit disregarding
4546 @geindex -b (gnatprep)
4553 Causes both preprocessor lines and the lines deleted by
4554 preprocessing to be replaced by blank lines in the output source file,
4555 preserving line numbers in the output file.
4558 @geindex -c (gnatprep)
4565 Causes both preprocessor lines and the lines deleted
4566 by preprocessing to be retained in the output source as comments marked
4567 with the special string @code{"--! "}. This option will result in line numbers
4568 being preserved in the output file.
4571 @geindex -C (gnatprep)
4578 Causes comments to be scanned. Normally comments are ignored by gnatprep.
4579 If this option is specified, then comments are scanned and any $symbol
4580 substitutions performed as in program text. This is particularly useful
4581 when structured comments are used (e.g., for programs written in a
4582 pre-2014 version of the SPARK Ada subset). Note that this switch is not
4583 available when doing integrated preprocessing (it would be useless in
4584 this context since comments are ignored by the compiler in any case).
4587 @geindex -D (gnatprep)
4592 @item @code{-D@emph{symbol}[=@emph{value}]}
4594 Defines a new preprocessing symbol with the specified value. If no value is given
4595 on the command line, then symbol is considered to be @code{True}. This switch
4596 can be used in place of a definition file.
4599 @geindex -r (gnatprep)
4606 Causes a @code{Source_Reference} pragma to be generated that
4607 references the original input file, so that error messages will use
4608 the file name of this original file. The use of this switch implies
4609 that preprocessor lines are not to be removed from the file, so its
4610 use will force @code{-b} mode if @code{-c}
4611 has not been specified explicitly.
4613 Note that if the file to be preprocessed contains multiple units, then
4614 it will be necessary to @code{gnatchop} the output file from
4615 @code{gnatprep}. If a @code{Source_Reference} pragma is present
4616 in the preprocessed file, it will be respected by
4618 so that the final chopped files will correctly refer to the original
4619 input source file for @code{gnatprep}.
4622 @geindex -s (gnatprep)
4629 Causes a sorted list of symbol names and values to be
4630 listed on the standard output file.
4633 @geindex -T (gnatprep)
4640 Use LF as line terminators when writing files. By default the line terminator
4641 of the host (LF under unix, CR/LF under Windows) is used.
4644 @geindex -u (gnatprep)
4651 Causes undefined symbols to be treated as having the value FALSE in the context
4652 of a preprocessor test. In the absence of this option, an undefined symbol in
4653 a @code{#if} or @code{#elsif} test will be treated as an error.
4656 @geindex -v (gnatprep)
4663 Verbose mode: generates more output about work done.
4666 Note: if neither @code{-b} nor @code{-c} is present,
4667 then preprocessor lines and
4668 deleted lines are completely removed from the output, unless -r is
4669 specified, in which case -b is assumed.
4671 @node Form of Definitions File,Form of Input Text for gnatprep,Switches for gnatprep,Preprocessing with gnatprep
4672 @anchor{gnat_ugn/the_gnat_compilation_model form-of-definitions-file}@anchor{98}@anchor{gnat_ugn/the_gnat_compilation_model id58}@anchor{99}
4673 @subsubsection Form of Definitions File
4676 The definitions file contains lines of the form:
4682 where @code{symbol} is a preprocessing symbol, and @code{value} is one of the following:
4688 Empty, corresponding to a null substitution,
4691 A string literal using normal Ada syntax, or
4694 Any sequence of characters from the set @{letters, digits, period, underline@}.
4697 Comment lines may also appear in the definitions file, starting with
4698 the usual @code{--},
4699 and comments may be added to the definitions lines.
4701 @node Form of Input Text for gnatprep,,Form of Definitions File,Preprocessing with gnatprep
4702 @anchor{gnat_ugn/the_gnat_compilation_model form-of-input-text-for-gnatprep}@anchor{9a}@anchor{gnat_ugn/the_gnat_compilation_model id59}@anchor{9b}
4703 @subsubsection Form of Input Text for @code{gnatprep}
4706 The input text may contain preprocessor conditional inclusion lines,
4707 as well as general symbol substitution sequences.
4709 The preprocessor conditional inclusion commands have the form:
4712 #if <expression> [then]
4714 #elsif <expression> [then]
4716 #elsif <expression> [then]
4724 In this example, <expression> is defined by the following grammar:
4727 <expression> ::= <symbol>
4728 <expression> ::= <symbol> = "<value>"
4729 <expression> ::= <symbol> = <symbol>
4730 <expression> ::= <symbol> = <integer>
4731 <expression> ::= <symbol> > <integer>
4732 <expression> ::= <symbol> >= <integer>
4733 <expression> ::= <symbol> < <integer>
4734 <expression> ::= <symbol> <= <integer>
4735 <expression> ::= <symbol> 'Defined
4736 <expression> ::= not <expression>
4737 <expression> ::= <expression> and <expression>
4738 <expression> ::= <expression> or <expression>
4739 <expression> ::= <expression> and then <expression>
4740 <expression> ::= <expression> or else <expression>
4741 <expression> ::= ( <expression> )
4744 Note the following restriction: it is not allowed to have “and” or “or”
4745 following “not” in the same expression without parentheses. For example, this
4752 This can be expressed instead as one of the following forms:
4759 For the first test (<expression> ::= <symbol>) the symbol must have
4760 either the value true or false, that is to say the right-hand of the
4761 symbol definition must be one of the (case-insensitive) literals
4762 @code{True} or @code{False}. If the value is true, then the
4763 corresponding lines are included, and if the value is false, they are
4766 When comparing a symbol to an integer, the integer is any non negative
4767 literal integer as defined in the Ada Reference Manual, such as 3, 16#FF# or
4768 2#11#. The symbol value must also be a non negative integer. Integer values
4769 in the range 0 .. 2**31-1 are supported.
4771 The test (<expression> ::= <symbol>’Defined) is true only if
4772 the symbol has been defined in the definition file or by a @code{-D}
4773 switch on the command line. Otherwise, the test is false.
4775 The equality tests are case insensitive, as are all the preprocessor lines.
4777 If the symbol referenced is not defined in the symbol definitions file,
4778 then the effect depends on whether or not switch @code{-u}
4779 is specified. If so, then the symbol is treated as if it had the value
4780 false and the test fails. If this switch is not specified, then
4781 it is an error to reference an undefined symbol. It is also an error to
4782 reference a symbol that is defined with a value other than @code{True}
4785 The use of the @code{not} operator inverts the sense of this logical test.
4786 The @code{not} operator cannot be combined with the @code{or} or @code{and}
4787 operators, without parentheses. For example, “if not X or Y then” is not
4788 allowed, but “if (not X) or Y then” and “if not (X or Y) then” are.
4790 The @code{then} keyword is optional as shown
4792 The @code{#} must be the first non-blank character on a line, but
4793 otherwise the format is free form. Spaces or tabs may appear between
4794 the @code{#} and the keyword. The keywords and the symbols are case
4795 insensitive as in normal Ada code. Comments may be used on a
4796 preprocessor line, but other than that, no other tokens may appear on a
4797 preprocessor line. Any number of @code{elsif} clauses can be present,
4798 including none at all. The @code{else} is optional, as in Ada.
4800 The @code{#} marking the start of a preprocessor line must be the first
4801 non-blank character on the line, i.e., it must be preceded only by
4802 spaces or horizontal tabs.
4804 Symbol substitution outside of preprocessor lines is obtained by using
4811 anywhere within a source line, except in a comment or within a
4812 string literal. The identifier
4813 following the @code{$} must match one of the symbols defined in the symbol
4814 definition file, and the result is to substitute the value of the
4815 symbol in place of @code{$symbol} in the output file.
4817 Note that although the substitution of strings within a string literal
4818 is not possible, it is possible to have a symbol whose defined value is
4819 a string literal. So instead of setting XYZ to @code{hello} and writing:
4822 Header : String := "$XYZ";
4825 you should set XYZ to @code{"hello"} and write:
4828 Header : String := $XYZ;
4831 and then the substitution will occur as desired.
4833 @node Integrated Preprocessing,,Preprocessing with gnatprep,Conditional Compilation
4834 @anchor{gnat_ugn/the_gnat_compilation_model id60}@anchor{9c}@anchor{gnat_ugn/the_gnat_compilation_model integrated-preprocessing}@anchor{90}
4835 @subsection Integrated Preprocessing
4838 As noted above, a file to be preprocessed consists of Ada source code
4839 in which preprocessing lines have been inserted. However,
4840 instead of using @code{gnatprep} to explicitly preprocess a file as a separate
4841 step before compilation, you can carry out the preprocessing implicitly
4842 as part of compilation. Such @emph{integrated preprocessing}, which is the common
4843 style with C, is performed when either or both of the following switches
4844 are passed to the compiler:
4852 @code{-gnatep}, which specifies the @emph{preprocessor data file}.
4853 This file dictates how the source files will be preprocessed (e.g., which
4854 symbol definition files apply to which sources).
4857 @code{-gnateD}, which defines values for preprocessing symbols.
4861 Integrated preprocessing applies only to Ada source files, it is
4862 not available for configuration pragma files.
4864 With integrated preprocessing, the output from the preprocessor is not,
4865 by default, written to any external file. Instead it is passed
4866 internally to the compiler. To preserve the result of
4867 preprocessing in a file, either run @code{gnatprep}
4868 in standalone mode or else supply the @code{-gnateG} switch
4869 (described below) to the compiler.
4871 When using project files:
4879 the builder switch @code{-x} should be used if any Ada source is
4880 compiled with @code{gnatep=}, so that the compiler finds the
4881 @emph{preprocessor data file}.
4884 the preprocessing data file and the symbol definition files should be
4885 located in the source directories of the project.
4889 Note that the @code{gnatmake} switch @code{-m} will almost
4890 always trigger recompilation for sources that are preprocessed,
4891 because @code{gnatmake} cannot compute the checksum of the source after
4894 The actual preprocessing function is described in detail in
4895 @ref{8f,,Preprocessing with gnatprep}. This section explains the switches
4896 that relate to integrated preprocessing.
4898 @geindex -gnatep (gcc)
4903 @item @code{-gnatep=@emph{preprocessor_data_file}}
4905 This switch specifies the file name (without directory
4906 information) of the preprocessor data file. Either place this file
4907 in one of the source directories, or, when using project
4908 files, reference the project file’s directory via the
4909 @code{project_name'Project_Dir} project attribute; e.g:
4916 for Switches ("Ada") use
4917 ("-gnatep=" & Prj'Project_Dir & "prep.def");
4923 A preprocessor data file is a text file that contains @emph{preprocessor
4924 control lines}. A preprocessor control line directs the preprocessing of
4925 either a particular source file, or, analogous to @code{others} in Ada,
4926 all sources not specified elsewhere in the preprocessor data file.
4927 A preprocessor control line
4928 can optionally identify a @emph{definition file} that assigns values to
4929 preprocessor symbols, as well as a list of switches that relate to
4931 Empty lines and comments (using Ada syntax) are also permitted, with no
4934 Here’s an example of a preprocessor data file:
4939 "toto.adb" "prep.def" -u
4940 -- Preprocess toto.adb, using definition file prep.def
4941 -- Undefined symbols are treated as False
4944 -- Preprocess all other sources without using a definition file
4945 -- Suppressed lined are commented
4946 -- Symbol VERSION has the value V101
4948 "tata.adb" "prep2.def" -s
4949 -- Preprocess tata.adb, using definition file prep2.def
4950 -- List all symbols with their values
4954 A preprocessor control line has the following syntax:
4959 <preprocessor_control_line> ::=
4960 <preprocessor_input> [ <definition_file_name> ] @{ <switch> @}
4962 <preprocessor_input> ::= <source_file_name> | '*'
4964 <definition_file_name> ::= <string_literal>
4966 <source_file_name> := <string_literal>
4968 <switch> := (See below for list)
4972 Thus each preprocessor control line starts with either a literal string or
4979 A literal string is the file name (without directory information) of the source
4980 file that will be input to the preprocessor.
4983 The character ‘*’ is a wild-card indicator; the additional parameters on the line
4984 indicate the preprocessing for all the sources
4985 that are not specified explicitly on other lines (the order of the lines is not
4989 It is an error to have two lines with the same file name or two
4990 lines starting with the character ‘*’.
4992 After the file name or ‘*’, an optional literal string specifies the name of
4993 the definition file to be used for preprocessing
4994 (@ref{98,,Form of Definitions File}). The definition files are found by the
4995 compiler in one of the source directories. In some cases, when compiling
4996 a source in a directory other than the current directory, if the definition
4997 file is in the current directory, it may be necessary to add the current
4998 directory as a source directory through the @code{-I} switch; otherwise
4999 the compiler would not find the definition file.
5001 Finally, switches similar to those of @code{gnatprep} may optionally appear:
5008 Causes both preprocessor lines and the lines deleted by
5009 preprocessing to be replaced by blank lines, preserving the line number.
5010 This switch is always implied; however, if specified after @code{-c}
5011 it cancels the effect of @code{-c}.
5015 Causes both preprocessor lines and the lines deleted
5016 by preprocessing to be retained as comments marked
5017 with the special string ‘@cite{–!}’.
5019 @item @code{-D@emph{symbol}=@emph{new_value}}
5021 Define or redefine @code{symbol} to have @code{new_value} as its value.
5022 The permitted form for @code{symbol} is either an Ada identifier, or any Ada reserved word
5023 aside from @code{if},
5024 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
5025 The permitted form for @code{new_value} is a literal string, an Ada identifier or any Ada reserved
5026 word. A symbol declared with this switch replaces a symbol with the
5027 same name defined in a definition file.
5031 Causes a sorted list of symbol names and values to be
5032 listed on the standard output file.
5036 Causes undefined symbols to be treated as having the value @code{FALSE}
5038 of a preprocessor test. In the absence of this option, an undefined symbol in
5039 a @code{#if} or @code{#elsif} test will be treated as an error.
5043 @geindex -gnateD (gcc)
5048 @item @code{-gnateD@emph{symbol}[=@emph{new_value}]}
5050 Define or redefine @code{symbol} to have @code{new_value} as its value. If no value
5051 is supplied, then the value of @code{symbol} is @code{True}.
5052 The form of @code{symbol} is an identifier, following normal Ada (case-insensitive)
5053 rules for its syntax, and @code{new_value} is either an arbitrary string between double
5054 quotes or any sequence (including an empty sequence) of characters from the
5055 set (letters, digits, period, underline).
5056 Ada reserved words may be used as symbols, with the exceptions of @code{if},
5057 @code{else}, @code{elsif}, @code{end}, @code{and}, @code{or} and @code{then}.
5066 -gnateDFoo=\"Foo-Bar\"
5070 A symbol declared with this switch on the command line replaces a
5071 symbol with the same name either in a definition file or specified with a
5072 switch @code{-D} in the preprocessor data file.
5074 This switch is similar to switch @code{-D} of @code{gnatprep}.
5076 @item @code{-gnateG}
5078 When integrated preprocessing is performed on source file @code{filename.extension},
5079 create or overwrite @code{filename.extension.prep} to contain
5080 the result of the preprocessing.
5081 For example if the source file is @code{foo.adb} then
5082 the output file will be @code{foo.adb.prep}.
5085 @node Mixed Language Programming,GNAT and Other Compilation Models,Conditional Compilation,The GNAT Compilation Model
5086 @anchor{gnat_ugn/the_gnat_compilation_model id61}@anchor{9d}@anchor{gnat_ugn/the_gnat_compilation_model mixed-language-programming}@anchor{2c}
5087 @section Mixed Language Programming
5090 @geindex Mixed Language Programming
5092 This section describes how to develop a mixed-language program,
5093 with a focus on combining Ada with C or C++.
5096 * Interfacing to C::
5097 * Calling Conventions::
5098 * Building Mixed Ada and C++ Programs::
5099 * Generating Ada Bindings for C and C++ headers::
5100 * Generating C Headers for Ada Specifications::
5104 @node Interfacing to C,Calling Conventions,,Mixed Language Programming
5105 @anchor{gnat_ugn/the_gnat_compilation_model id62}@anchor{9e}@anchor{gnat_ugn/the_gnat_compilation_model interfacing-to-c}@anchor{9f}
5106 @subsection Interfacing to C
5109 Interfacing Ada with a foreign language such as C involves using
5110 compiler directives to import and/or export entity definitions in each
5111 language – using @code{extern} statements in C, for instance, and the
5112 @code{Import}, @code{Export}, and @code{Convention} pragmas in Ada.
5113 A full treatment of these topics is provided in Appendix B, section 1
5114 of the Ada Reference Manual.
5116 There are two ways to build a program using GNAT that contains some Ada
5117 sources and some foreign language sources, depending on whether or not
5118 the main subprogram is written in Ada. Here is a source example with
5119 the main subprogram in Ada:
5125 void print_num (int num)
5127 printf ("num is %d.\\n", num);
5135 /* num_from_Ada is declared in my_main.adb */
5136 extern int num_from_Ada;
5140 return num_from_Ada;
5146 procedure My_Main is
5148 -- Declare then export an Integer entity called num_from_Ada
5149 My_Num : Integer := 10;
5150 pragma Export (C, My_Num, "num_from_Ada");
5152 -- Declare an Ada function spec for Get_Num, then use
5153 -- C function get_num for the implementation.
5154 function Get_Num return Integer;
5155 pragma Import (C, Get_Num, "get_num");
5157 -- Declare an Ada procedure spec for Print_Num, then use
5158 -- C function print_num for the implementation.
5159 procedure Print_Num (Num : Integer);
5160 pragma Import (C, Print_Num, "print_num");
5163 Print_Num (Get_Num);
5167 To build this example:
5173 First compile the foreign language files to
5174 generate object files:
5182 Then, compile the Ada units to produce a set of object files and ALI
5186 $ gnatmake -c my_main.adb
5190 Run the Ada binder on the Ada main program:
5193 $ gnatbind my_main.ali
5197 Link the Ada main program, the Ada objects and the other language
5201 $ gnatlink my_main.ali file1.o file2.o
5205 The last three steps can be grouped in a single command:
5208 $ gnatmake my_main.adb -largs file1.o file2.o
5211 @geindex Binder output file
5213 If the main program is in a language other than Ada, then you may have
5214 more than one entry point into the Ada subsystem. You must use a special
5215 binder option to generate callable routines that initialize and
5216 finalize the Ada units (@ref{a0,,Binding with Non-Ada Main Programs}).
5217 Calls to the initialization and finalization routines must be inserted
5218 in the main program, or some other appropriate point in the code. The
5219 call to initialize the Ada units must occur before the first Ada
5220 subprogram is called, and the call to finalize the Ada units must occur
5221 after the last Ada subprogram returns. The binder will place the
5222 initialization and finalization subprograms into the
5223 @code{b~xxx.adb} file where they can be accessed by your C
5224 sources. To illustrate, we have the following example:
5228 extern void adainit (void);
5229 extern void adafinal (void);
5230 extern int add (int, int);
5231 extern int sub (int, int);
5233 int main (int argc, char *argv[])
5239 /* Should print "21 + 7 = 28" */
5240 printf ("%d + %d = %d\\n", a, b, add (a, b));
5242 /* Should print "21 - 7 = 14" */
5243 printf ("%d - %d = %d\\n", a, b, sub (a, b));
5252 function Add (A, B : Integer) return Integer;
5253 pragma Export (C, Add, "add");
5259 package body Unit1 is
5260 function Add (A, B : Integer) return Integer is
5270 function Sub (A, B : Integer) return Integer;
5271 pragma Export (C, Sub, "sub");
5277 package body Unit2 is
5278 function Sub (A, B : Integer) return Integer is
5285 The build procedure for this application is similar to the last
5292 First, compile the foreign language files to generate object files:
5299 Next, compile the Ada units to produce a set of object files and ALI
5303 $ gnatmake -c unit1.adb
5304 $ gnatmake -c unit2.adb
5308 Run the Ada binder on every generated ALI file. Make sure to use the
5309 @code{-n} option to specify a foreign main program:
5312 $ gnatbind -n unit1.ali unit2.ali
5316 Link the Ada main program, the Ada objects and the foreign language
5317 objects. You need only list the last ALI file here:
5320 $ gnatlink unit2.ali main.o -o exec_file
5323 This procedure yields a binary executable called @code{exec_file}.
5326 Depending on the circumstances (for example when your non-Ada main object
5327 does not provide symbol @code{main}), you may also need to instruct the
5328 GNAT linker not to include the standard startup objects by passing the
5329 @code{-nostartfiles} switch to @code{gnatlink}.
5331 @node Calling Conventions,Building Mixed Ada and C++ Programs,Interfacing to C,Mixed Language Programming
5332 @anchor{gnat_ugn/the_gnat_compilation_model calling-conventions}@anchor{a1}@anchor{gnat_ugn/the_gnat_compilation_model id63}@anchor{a2}
5333 @subsection Calling Conventions
5336 @geindex Foreign Languages
5338 @geindex Calling Conventions
5340 GNAT follows standard calling sequence conventions and will thus interface
5341 to any other language that also follows these conventions. The following
5342 Convention identifiers are recognized by GNAT:
5344 @geindex Interfacing to Ada
5346 @geindex Other Ada compilers
5348 @geindex Convention Ada
5355 This indicates that the standard Ada calling sequence will be
5356 used and all Ada data items may be passed without any limitations in the
5357 case where GNAT is used to generate both the caller and callee. It is also
5358 possible to mix GNAT generated code and code generated by another Ada
5359 compiler. In this case, the data types should be restricted to simple
5360 cases, including primitive types. Whether complex data types can be passed
5361 depends on the situation. Probably it is safe to pass simple arrays, such
5362 as arrays of integers or floats. Records may or may not work, depending
5363 on whether both compilers lay them out identically. Complex structures
5364 involving variant records, access parameters, tasks, or protected types,
5365 are unlikely to be able to be passed.
5367 Note that in the case of GNAT running
5368 on a platform that supports HP Ada 83, a higher degree of compatibility
5369 can be guaranteed, and in particular records are laid out in an identical
5370 manner in the two compilers. Note also that if output from two different
5371 compilers is mixed, the program is responsible for dealing with elaboration
5372 issues. Probably the safest approach is to write the main program in the
5373 version of Ada other than GNAT, so that it takes care of its own elaboration
5374 requirements, and then call the GNAT-generated adainit procedure to ensure
5375 elaboration of the GNAT components. Consult the documentation of the other
5376 Ada compiler for further details on elaboration.
5378 However, it is not possible to mix the tasking run time of GNAT and
5379 HP Ada 83, All the tasking operations must either be entirely within
5380 GNAT compiled sections of the program, or entirely within HP Ada 83
5381 compiled sections of the program.
5384 @geindex Interfacing to Assembly
5386 @geindex Convention Assembler
5391 @item @code{Assembler}
5393 Specifies assembler as the convention. In practice this has the
5394 same effect as convention Ada (but is not equivalent in the sense of being
5395 considered the same convention).
5398 @geindex Convention Asm
5407 Equivalent to Assembler.
5409 @geindex Interfacing to COBOL
5411 @geindex Convention COBOL
5421 Data will be passed according to the conventions described
5422 in section B.4 of the Ada Reference Manual.
5427 @geindex Interfacing to C
5429 @geindex Convention C
5436 Data will be passed according to the conventions described
5437 in section B.3 of the Ada Reference Manual.
5439 A note on interfacing to a C ‘varargs’ function:
5443 @geindex C varargs function
5445 @geindex Interfacing to C varargs function
5447 @geindex varargs function interfaces
5449 In C, @code{varargs} allows a function to take a variable number of
5450 arguments. There is no direct equivalent in this to Ada. One
5451 approach that can be used is to create a C wrapper for each
5452 different profile and then interface to this C wrapper. For
5453 example, to print an @code{int} value using @code{printf},
5454 create a C function @code{printfi} that takes two arguments, a
5455 pointer to a string and an int, and calls @code{printf}.
5456 Then in the Ada program, use pragma @code{Import} to
5457 interface to @code{printfi}.
5459 It may work on some platforms to directly interface to
5460 a @code{varargs} function by providing a specific Ada profile
5461 for a particular call. However, this does not work on
5462 all platforms, since there is no guarantee that the
5463 calling sequence for a two argument normal C function
5464 is the same as for calling a @code{varargs} C function with
5465 the same two arguments.
5469 @geindex Convention Default
5476 @item @code{Default}
5481 @geindex Convention External
5488 @item @code{External}
5495 @geindex Interfacing to C++
5497 @geindex Convention C++
5502 @item @code{C_Plus_Plus} (or @code{CPP})
5504 This stands for C++. For most purposes this is identical to C.
5505 See the separate description of the specialized GNAT pragmas relating to
5506 C++ interfacing for further details.
5511 @geindex Interfacing to Fortran
5513 @geindex Convention Fortran
5518 @item @code{Fortran}
5520 Data will be passed according to the conventions described
5521 in section B.5 of the Ada Reference Manual.
5523 @item @code{Intrinsic}
5525 This applies to an intrinsic operation, as defined in the Ada
5526 Reference Manual. If a pragma Import (Intrinsic) applies to a subprogram,
5527 this means that the body of the subprogram is provided by the compiler itself,
5528 usually by means of an efficient code sequence, and that the user does not
5529 supply an explicit body for it. In an application program, the pragma may
5530 be applied to the following sets of names:
5536 Rotate_Left, Rotate_Right, Shift_Left, Shift_Right, Shift_Right_Arithmetic.
5537 The corresponding subprogram declaration must have
5538 two formal parameters. The
5539 first one must be a signed integer type or a modular type with a binary
5540 modulus, and the second parameter must be of type Natural.
5541 The return type must be the same as the type of the first argument. The size
5542 of this type can only be 8, 16, 32, or 64.
5545 Binary arithmetic operators: ‘+’, ‘-‘, ‘*’, ‘/’.
5546 The corresponding operator declaration must have parameters and result type
5547 that have the same root numeric type (for example, all three are long_float
5548 types). This simplifies the definition of operations that use type checking
5549 to perform dimensional checks:
5553 type Distance is new Long_Float;
5554 type Time is new Long_Float;
5555 type Velocity is new Long_Float;
5556 function "/" (D : Distance; T : Time)
5558 pragma Import (Intrinsic, "/");
5560 This common idiom is often programmed with a generic definition and an
5561 explicit body. The pragma makes it simpler to introduce such declarations.
5562 It incurs no overhead in compilation time or code size, because it is
5563 implemented as a single machine instruction.
5570 General subprogram entities. This is used to bind an Ada subprogram
5572 a compiler builtin by name with back-ends where such interfaces are
5573 available. A typical example is the set of @code{__builtin} functions
5574 exposed by the GCC back-end, as in the following example:
5577 function builtin_sqrt (F : Float) return Float;
5578 pragma Import (Intrinsic, builtin_sqrt, "__builtin_sqrtf");
5581 Most of the GCC builtins are accessible this way, and as for other
5582 import conventions (e.g. C), it is the user’s responsibility to ensure
5583 that the Ada subprogram profile matches the underlying builtin
5590 @geindex Convention Stdcall
5595 @item @code{Stdcall}
5597 This is relevant only to Windows implementations of GNAT,
5598 and specifies that the @code{Stdcall} calling sequence will be used,
5599 as defined by the NT API. Nevertheless, to ease building
5600 cross-platform bindings this convention will be handled as a @code{C} calling
5601 convention on non-Windows platforms.
5606 @geindex Convention DLL
5613 This is equivalent to @code{Stdcall}.
5618 @geindex Convention Win32
5625 This is equivalent to @code{Stdcall}.
5630 @geindex Convention Stubbed
5635 @item @code{Stubbed}
5637 This is a special convention that indicates that the compiler
5638 should provide a stub body that raises @code{Program_Error}.
5641 GNAT additionally provides a useful pragma @code{Convention_Identifier}
5642 that can be used to parameterize conventions and allow additional synonyms
5643 to be specified. For example if you have legacy code in which the convention
5644 identifier Fortran77 was used for Fortran, you can use the configuration
5648 pragma Convention_Identifier (Fortran77, Fortran);
5651 And from now on the identifier Fortran77 may be used as a convention
5652 identifier (for example in an @code{Import} pragma) with the same
5655 @node Building Mixed Ada and C++ Programs,Generating Ada Bindings for C and C++ headers,Calling Conventions,Mixed Language Programming
5656 @anchor{gnat_ugn/the_gnat_compilation_model building-mixed-ada-and-c-programs}@anchor{a3}@anchor{gnat_ugn/the_gnat_compilation_model id64}@anchor{a4}
5657 @subsection Building Mixed Ada and C++ Programs
5660 A programmer inexperienced with mixed-language development may find that
5661 building an application containing both Ada and C++ code can be a
5662 challenge. This section gives a few hints that should make this task easier.
5665 * Interfacing to C++::
5666 * Linking a Mixed C++ & Ada Program::
5667 * A Simple Example::
5668 * Interfacing with C++ constructors::
5669 * Interfacing with C++ at the Class Level::
5673 @node Interfacing to C++,Linking a Mixed C++ & Ada Program,,Building Mixed Ada and C++ Programs
5674 @anchor{gnat_ugn/the_gnat_compilation_model id65}@anchor{a5}@anchor{gnat_ugn/the_gnat_compilation_model id66}@anchor{a6}
5675 @subsubsection Interfacing to C++
5678 GNAT supports interfacing with the G++ compiler (or any C++ compiler
5679 generating code that is compatible with the G++ Application Binary
5680 Interface —see @indicateurl{http://www.codesourcery.com/archives/cxx-abi}).
5682 Interfacing can be done at 3 levels: simple data, subprograms, and
5683 classes. In the first two cases, GNAT offers a specific @code{Convention C_Plus_Plus}
5684 (or @code{CPP}) that behaves exactly like @code{Convention C}.
5685 Usually, C++ mangles the names of subprograms. To generate proper mangled
5686 names automatically, see @ref{a7,,Generating Ada Bindings for C and C++ headers}).
5687 This problem can also be addressed manually in two ways:
5693 by modifying the C++ code in order to force a C convention using
5694 the @code{extern "C"} syntax.
5697 by figuring out the mangled name (using e.g. @code{nm}) and using it as the
5698 Link_Name argument of the pragma import.
5701 Interfacing at the class level can be achieved by using the GNAT specific
5702 pragmas such as @code{CPP_Constructor}. See the @cite{GNAT_Reference_Manual} for additional information.
5704 @node Linking a Mixed C++ & Ada Program,A Simple Example,Interfacing to C++,Building Mixed Ada and C++ Programs
5705 @anchor{gnat_ugn/the_gnat_compilation_model linking-a-mixed-c-ada-program}@anchor{a8}@anchor{gnat_ugn/the_gnat_compilation_model linking-a-mixed-c-and-ada-program}@anchor{a9}
5706 @subsubsection Linking a Mixed C++ & Ada Program
5709 Usually the linker of the C++ development system must be used to link
5710 mixed applications because most C++ systems will resolve elaboration
5711 issues (such as calling constructors on global class instances)
5712 transparently during the link phase. GNAT has been adapted to ease the
5713 use of a foreign linker for the last phase. Three cases can be
5720 Using GNAT and G++ (GNU C++ compiler) from the same GCC installation:
5721 The C++ linker can simply be called by using the C++ specific driver
5724 Note that if the C++ code uses inline functions, you will need to
5725 compile your C++ code with the @code{-fkeep-inline-functions} switch in
5726 order to provide an existing function implementation that the Ada code can
5730 $ g++ -c -fkeep-inline-functions file1.C
5731 $ g++ -c -fkeep-inline-functions file2.C
5732 $ gnatmake ada_unit -largs file1.o file2.o --LINK=g++
5736 Using GNAT and G++ from two different GCC installations: If both
5737 compilers are on the :envvar`PATH`, the previous method may be used. It is
5738 important to note that environment variables such as
5739 @geindex C_INCLUDE_PATH
5740 @geindex environment variable; C_INCLUDE_PATH
5741 @code{C_INCLUDE_PATH},
5742 @geindex GCC_EXEC_PREFIX
5743 @geindex environment variable; GCC_EXEC_PREFIX
5744 @code{GCC_EXEC_PREFIX},
5745 @geindex BINUTILS_ROOT
5746 @geindex environment variable; BINUTILS_ROOT
5747 @code{BINUTILS_ROOT}, and
5749 @geindex environment variable; GCC_ROOT
5750 @code{GCC_ROOT} will affect both compilers
5751 at the same time and may make one of the two compilers operate
5752 improperly if set during invocation of the wrong compiler. It is also
5753 very important that the linker uses the proper @code{libgcc.a} GCC
5754 library – that is, the one from the C++ compiler installation. The
5755 implicit link command as suggested in the @code{gnatmake} command
5756 from the former example can be replaced by an explicit link command with
5757 the full-verbosity option in order to verify which library is used:
5761 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=c++
5764 If there is a problem due to interfering environment variables, it can
5765 be worked around by using an intermediate script. The following example
5766 shows the proper script to use when GNAT has not been installed at its
5767 default location and g++ has been installed at its default location:
5775 $ gnatlink -v -v ada_unit file1.o file2.o --LINK=./my_script
5779 Using a non-GNU C++ compiler: The commands previously described can be
5780 used to insure that the C++ linker is used. Nonetheless, you need to add
5781 a few more parameters to the link command line, depending on the exception
5784 If the @code{setjmp} / @code{longjmp} exception mechanism is used, only the paths
5785 to the @code{libgcc} libraries are required:
5790 CC $* gcc -print-file-name=libgcc.a gcc -print-file-name=libgcc_eh.a
5791 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
5794 where CC is the name of the non-GNU C++ compiler.
5796 If the “zero cost” exception mechanism is used, and the platform
5797 supports automatic registration of exception tables (e.g., Solaris),
5798 paths to more objects are required:
5803 CC gcc -print-file-name=crtbegin.o $* \\
5804 gcc -print-file-name=libgcc.a gcc -print-file-name=libgcc_eh.a \\
5805 gcc -print-file-name=crtend.o
5806 $ gnatlink ada_unit file1.o file2.o --LINK=./my_script
5809 If the “zero cost exception” mechanism is used, and the platform
5810 doesn’t support automatic registration of exception tables (e.g., HP-UX
5811 or AIX), the simple approach described above will not work and
5812 a pre-linking phase using GNAT will be necessary.
5815 Another alternative is to use the @code{gprbuild} multi-language builder
5816 which has a large knowledge base and knows how to link Ada and C++ code
5817 together automatically in most cases.
5819 @node A Simple Example,Interfacing with C++ constructors,Linking a Mixed C++ & Ada Program,Building Mixed Ada and C++ Programs
5820 @anchor{gnat_ugn/the_gnat_compilation_model a-simple-example}@anchor{aa}@anchor{gnat_ugn/the_gnat_compilation_model id67}@anchor{ab}
5821 @subsubsection A Simple Example
5824 The following example, provided as part of the GNAT examples, shows how
5825 to achieve procedural interfacing between Ada and C++ in both
5826 directions. The C++ class A has two methods. The first method is exported
5827 to Ada by the means of an extern C wrapper function. The second method
5828 calls an Ada subprogram. On the Ada side, the C++ calls are modelled by
5829 a limited record with a layout comparable to the C++ class. The Ada
5830 subprogram, in turn, calls the C++ method. So, starting from the C++
5831 main program, the process passes back and forth between the two
5834 Here are the compilation commands:
5837 $ gnatmake -c simple_cpp_interface
5840 $ gnatbind -n simple_cpp_interface
5841 $ gnatlink simple_cpp_interface -o cpp_main --LINK=g++ -lstdc++ ex7.o cpp_main.o
5844 Here are the corresponding sources:
5852 void adainit (void);
5853 void adafinal (void);
5854 void method1 (A *t);
5878 class A : public Origin @{
5880 void method1 (void);
5881 void method2 (int v);
5893 extern "C" @{ void ada_method2 (A *t, int v);@}
5895 void A::method1 (void)
5898 printf ("in A::method1, a_value = %d \\n",a_value);
5901 void A::method2 (int v)
5903 ada_method2 (this, v);
5904 printf ("in A::method2, a_value = %d \\n",a_value);
5910 printf ("in A::A, a_value = %d \\n",a_value);
5915 -- simple_cpp_interface.ads
5917 package Simple_Cpp_Interface is
5920 Vptr : System.Address;
5924 pragma Convention (C, A);
5926 procedure Method1 (This : in out A);
5927 pragma Import (C, Method1);
5929 procedure Ada_Method2 (This : in out A; V : Integer);
5930 pragma Export (C, Ada_Method2);
5932 end Simple_Cpp_Interface;
5936 -- simple_cpp_interface.adb
5937 package body Simple_Cpp_Interface is
5939 procedure Ada_Method2 (This : in out A; V : Integer) is
5945 end Simple_Cpp_Interface;
5948 @node Interfacing with C++ constructors,Interfacing with C++ at the Class Level,A Simple Example,Building Mixed Ada and C++ Programs
5949 @anchor{gnat_ugn/the_gnat_compilation_model id68}@anchor{ac}@anchor{gnat_ugn/the_gnat_compilation_model interfacing-with-c-constructors}@anchor{ad}
5950 @subsubsection Interfacing with C++ constructors
5953 In order to interface with C++ constructors GNAT provides the
5954 @code{pragma CPP_Constructor} (see the @cite{GNAT_Reference_Manual}
5955 for additional information).
5956 In this section we present some common uses of C++ constructors
5957 in mixed-languages programs in GNAT.
5959 Let us assume that we need to interface with the following
5967 virtual int Get_Value ();
5968 Root(); // Default constructor
5969 Root(int v); // 1st non-default constructor
5970 Root(int v, int w); // 2nd non-default constructor
5974 For this purpose we can write the following package spec (further
5975 information on how to build this spec is available in
5976 @ref{ae,,Interfacing with C++ at the Class Level} and
5977 @ref{a7,,Generating Ada Bindings for C and C++ headers}).
5980 with Interfaces.C; use Interfaces.C;
5982 type Root is tagged limited record
5986 pragma Import (CPP, Root);
5988 function Get_Value (Obj : Root) return int;
5989 pragma Import (CPP, Get_Value);
5991 function Constructor return Root;
5992 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ev");
5994 function Constructor (v : Integer) return Root;
5995 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Ei");
5997 function Constructor (v, w : Integer) return Root;
5998 pragma Cpp_Constructor (Constructor, "_ZN4RootC1Eii");
6002 On the Ada side the constructor is represented by a function (whose
6003 name is arbitrary) that returns the classwide type corresponding to
6004 the imported C++ class. Although the constructor is described as a
6005 function, it is typically a procedure with an extra implicit argument
6006 (the object being initialized) at the implementation level. GNAT
6007 issues the appropriate call, whatever it is, to get the object
6008 properly initialized.
6010 Constructors can only appear in the following contexts:
6016 On the right side of an initialization of an object of type @code{T}.
6019 On the right side of an initialization of a record component of type @code{T}.
6022 In an Ada 2005 limited aggregate.
6025 In an Ada 2005 nested limited aggregate.
6028 In an Ada 2005 limited aggregate that initializes an object built in
6029 place by an extended return statement.
6032 In a declaration of an object whose type is a class imported from C++,
6033 either the default C++ constructor is implicitly called by GNAT, or
6034 else the required C++ constructor must be explicitly called in the
6035 expression that initializes the object. For example:
6039 Obj2 : Root := Constructor;
6040 Obj3 : Root := Constructor (v => 10);
6041 Obj4 : Root := Constructor (30, 40);
6044 The first two declarations are equivalent: in both cases the default C++
6045 constructor is invoked (in the former case the call to the constructor is
6046 implicit, and in the latter case the call is explicit in the object
6047 declaration). @code{Obj3} is initialized by the C++ non-default constructor
6048 that takes an integer argument, and @code{Obj4} is initialized by the
6049 non-default C++ constructor that takes two integers.
6051 Let us derive the imported C++ class in the Ada side. For example:
6054 type DT is new Root with record
6055 C_Value : Natural := 2009;
6059 In this case the components DT inherited from the C++ side must be
6060 initialized by a C++ constructor, and the additional Ada components
6061 of type DT are initialized by GNAT. The initialization of such an
6062 object is done either by default, or by means of a function returning
6063 an aggregate of type DT, or by means of an extension aggregate.
6067 Obj6 : DT := Function_Returning_DT (50);
6068 Obj7 : DT := (Constructor (30,40) with C_Value => 50);
6071 The declaration of @code{Obj5} invokes the default constructors: the
6072 C++ default constructor of the parent type takes care of the initialization
6073 of the components inherited from Root, and GNAT takes care of the default
6074 initialization of the additional Ada components of type DT (that is,
6075 @code{C_Value} is initialized to value 2009). The order of invocation of
6076 the constructors is consistent with the order of elaboration required by
6077 Ada and C++. That is, the constructor of the parent type is always called
6078 before the constructor of the derived type.
6080 Let us now consider a record that has components whose type is imported
6081 from C++. For example:
6084 type Rec1 is limited record
6085 Data1 : Root := Constructor (10);
6086 Value : Natural := 1000;
6089 type Rec2 (D : Integer := 20) is limited record
6091 Data2 : Root := Constructor (D, 30);
6095 The initialization of an object of type @code{Rec2} will call the
6096 non-default C++ constructors specified for the imported components.
6103 Using Ada 2005 we can use limited aggregates to initialize an object
6104 invoking C++ constructors that differ from those specified in the type
6105 declarations. For example:
6108 Obj9 : Rec2 := (Rec => (Data1 => Constructor (15, 16),
6113 The above declaration uses an Ada 2005 limited aggregate to
6114 initialize @code{Obj9}, and the C++ constructor that has two integer
6115 arguments is invoked to initialize the @code{Data1} component instead
6116 of the constructor specified in the declaration of type @code{Rec1}. In
6117 Ada 2005 the box in the aggregate indicates that unspecified components
6118 are initialized using the expression (if any) available in the component
6119 declaration. That is, in this case discriminant @code{D} is initialized
6120 to value @code{20}, @code{Value} is initialized to value 1000, and the
6121 non-default C++ constructor that handles two integers takes care of
6122 initializing component @code{Data2} with values @code{20,30}.
6124 In Ada 2005 we can use the extended return statement to build the Ada
6125 equivalent to C++ non-default constructors. For example:
6128 function Constructor (V : Integer) return Rec2 is
6130 return Obj : Rec2 := (Rec => (Data1 => Constructor (V, 20),
6133 -- Further actions required for construction of
6134 -- objects of type Rec2
6140 In this example the extended return statement construct is used to
6141 build in place the returned object whose components are initialized
6142 by means of a limited aggregate. Any further action associated with
6143 the constructor can be placed inside the construct.
6145 @node Interfacing with C++ at the Class Level,,Interfacing with C++ constructors,Building Mixed Ada and C++ Programs
6146 @anchor{gnat_ugn/the_gnat_compilation_model id69}@anchor{af}@anchor{gnat_ugn/the_gnat_compilation_model interfacing-with-c-at-the-class-level}@anchor{ae}
6147 @subsubsection Interfacing with C++ at the Class Level
6150 In this section we demonstrate the GNAT features for interfacing with
6151 C++ by means of an example making use of Ada 2005 abstract interface
6152 types. This example consists of a classification of animals; classes
6153 have been used to model our main classification of animals, and
6154 interfaces provide support for the management of secondary
6155 classifications. We first demonstrate a case in which the types and
6156 constructors are defined on the C++ side and imported from the Ada
6157 side, and latter the reverse case.
6159 The root of our derivation will be the @code{Animal} class, with a
6160 single private attribute (the @code{Age} of the animal), a constructor,
6161 and two public primitives to set and get the value of this attribute.
6166 virtual void Set_Age (int New_Age);
6168 Animal() @{Age_Count = 0;@};
6174 Abstract interface types are defined in C++ by means of classes with pure
6175 virtual functions and no data members. In our example we will use two
6176 interfaces that provide support for the common management of @code{Carnivore}
6177 and @code{Domestic} animals:
6182 virtual int Number_Of_Teeth () = 0;
6187 virtual void Set_Owner (char* Name) = 0;
6191 Using these declarations, we can now say that a @code{Dog} is an animal that is
6192 both Carnivore and Domestic, that is:
6195 class Dog : Animal, Carnivore, Domestic @{
6197 virtual int Number_Of_Teeth ();
6198 virtual void Set_Owner (char* Name);
6200 Dog(); // Constructor
6207 In the following examples we will assume that the previous declarations are
6208 located in a file named @code{animals.h}. The following package demonstrates
6209 how to import these C++ declarations from the Ada side:
6212 with Interfaces.C.Strings; use Interfaces.C.Strings;
6214 type Carnivore is limited interface;
6215 pragma Convention (C_Plus_Plus, Carnivore);
6216 function Number_Of_Teeth (X : Carnivore)
6217 return Natural is abstract;
6219 type Domestic is limited interface;
6220 pragma Convention (C_Plus_Plus, Domestic);
6222 (X : in out Domestic;
6223 Name : Chars_Ptr) is abstract;
6225 type Animal is tagged limited record
6228 pragma Import (C_Plus_Plus, Animal);
6230 procedure Set_Age (X : in out Animal; Age : Integer);
6231 pragma Import (C_Plus_Plus, Set_Age);
6233 function Age (X : Animal) return Integer;
6234 pragma Import (C_Plus_Plus, Age);
6236 function New_Animal return Animal;
6237 pragma CPP_Constructor (New_Animal);
6238 pragma Import (CPP, New_Animal, "_ZN6AnimalC1Ev");
6240 type Dog is new Animal and Carnivore and Domestic with record
6241 Tooth_Count : Natural;
6244 pragma Import (C_Plus_Plus, Dog);
6246 function Number_Of_Teeth (A : Dog) return Natural;
6247 pragma Import (C_Plus_Plus, Number_Of_Teeth);
6249 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
6250 pragma Import (C_Plus_Plus, Set_Owner);
6252 function New_Dog return Dog;
6253 pragma CPP_Constructor (New_Dog);
6254 pragma Import (CPP, New_Dog, "_ZN3DogC2Ev");
6258 Thanks to the compatibility between GNAT run-time structures and the C++ ABI,
6259 interfacing with these C++ classes is easy. The only requirement is that all
6260 the primitives and components must be declared exactly in the same order in
6263 Regarding the abstract interfaces, we must indicate to the GNAT compiler by
6264 means of a @code{pragma Convention (C_Plus_Plus)}, the convention used to pass
6265 the arguments to the called primitives will be the same as for C++. For the
6266 imported classes we use @code{pragma Import} with convention @code{C_Plus_Plus}
6267 to indicate that they have been defined on the C++ side; this is required
6268 because the dispatch table associated with these tagged types will be built
6269 in the C++ side and therefore will not contain the predefined Ada primitives
6270 which Ada would otherwise expect.
6272 As the reader can see there is no need to indicate the C++ mangled names
6273 associated with each subprogram because it is assumed that all the calls to
6274 these primitives will be dispatching calls. The only exception is the
6275 constructor, which must be registered with the compiler by means of
6276 @code{pragma CPP_Constructor} and needs to provide its associated C++
6277 mangled name because the Ada compiler generates direct calls to it.
6279 With the above packages we can now declare objects of type Dog on the Ada side
6280 and dispatch calls to the corresponding subprograms on the C++ side. We can
6281 also extend the tagged type Dog with further fields and primitives, and
6282 override some of its C++ primitives on the Ada side. For example, here we have
6283 a type derivation defined on the Ada side that inherits all the dispatching
6284 primitives of the ancestor from the C++ side.
6287 with Animals; use Animals;
6288 package Vaccinated_Animals is
6289 type Vaccinated_Dog is new Dog with null record;
6290 function Vaccination_Expired (A : Vaccinated_Dog) return Boolean;
6291 end Vaccinated_Animals;
6294 It is important to note that, because of the ABI compatibility, the programmer
6295 does not need to add any further information to indicate either the object
6296 layout or the dispatch table entry associated with each dispatching operation.
6298 Now let us define all the types and constructors on the Ada side and export
6299 them to C++, using the same hierarchy of our previous example:
6302 with Interfaces.C.Strings;
6303 use Interfaces.C.Strings;
6305 type Carnivore is limited interface;
6306 pragma Convention (C_Plus_Plus, Carnivore);
6307 function Number_Of_Teeth (X : Carnivore)
6308 return Natural is abstract;
6310 type Domestic is limited interface;
6311 pragma Convention (C_Plus_Plus, Domestic);
6313 (X : in out Domestic;
6314 Name : Chars_Ptr) is abstract;
6316 type Animal is tagged record
6319 pragma Convention (C_Plus_Plus, Animal);
6321 procedure Set_Age (X : in out Animal; Age : Integer);
6322 pragma Export (C_Plus_Plus, Set_Age);
6324 function Age (X : Animal) return Integer;
6325 pragma Export (C_Plus_Plus, Age);
6327 function New_Animal return Animal'Class;
6328 pragma Export (C_Plus_Plus, New_Animal);
6330 type Dog is new Animal and Carnivore and Domestic with record
6331 Tooth_Count : Natural;
6332 Owner : String (1 .. 30);
6334 pragma Convention (C_Plus_Plus, Dog);
6336 function Number_Of_Teeth (A : Dog) return Natural;
6337 pragma Export (C_Plus_Plus, Number_Of_Teeth);
6339 procedure Set_Owner (A : in out Dog; Name : Chars_Ptr);
6340 pragma Export (C_Plus_Plus, Set_Owner);
6342 function New_Dog return Dog'Class;
6343 pragma Export (C_Plus_Plus, New_Dog);
6347 Compared with our previous example the only differences are the use of
6348 @code{pragma Convention} (instead of @code{pragma Import}), and the use of
6349 @code{pragma Export} to indicate to the GNAT compiler that the primitives will
6350 be available to C++. Thanks to the ABI compatibility, on the C++ side there is
6351 nothing else to be done; as explained above, the only requirement is that all
6352 the primitives and components are declared in exactly the same order.
6354 For completeness, let us see a brief C++ main program that uses the
6355 declarations available in @code{animals.h} (presented in our first example) to
6356 import and use the declarations from the Ada side, properly initializing and
6357 finalizing the Ada run-time system along the way:
6360 #include "animals.h"
6362 using namespace std;
6364 void Check_Carnivore (Carnivore *obj) @{...@}
6365 void Check_Domestic (Domestic *obj) @{...@}
6366 void Check_Animal (Animal *obj) @{...@}
6367 void Check_Dog (Dog *obj) @{...@}
6370 void adainit (void);
6371 void adafinal (void);
6377 Dog *obj = new_dog(); // Ada constructor
6378 Check_Carnivore (obj); // Check secondary DT
6379 Check_Domestic (obj); // Check secondary DT
6380 Check_Animal (obj); // Check primary DT
6381 Check_Dog (obj); // Check primary DT
6386 adainit (); test(); adafinal ();
6391 @node Generating Ada Bindings for C and C++ headers,Generating C Headers for Ada Specifications,Building Mixed Ada and C++ Programs,Mixed Language Programming
6392 @anchor{gnat_ugn/the_gnat_compilation_model generating-ada-bindings-for-c-and-c-headers}@anchor{a7}@anchor{gnat_ugn/the_gnat_compilation_model id70}@anchor{b0}
6393 @subsection Generating Ada Bindings for C and C++ headers
6396 @geindex Binding generation (for C and C++ headers)
6398 @geindex C headers (binding generation)
6400 @geindex C++ headers (binding generation)
6402 GNAT includes a binding generator for C and C++ headers which is
6403 intended to do 95% of the tedious work of generating Ada specs from C
6404 or C++ header files.
6406 Note that this capability is not intended to generate 100% correct Ada specs,
6407 and will is some cases require manual adjustments, although it can often
6408 be used out of the box in practice.
6410 Some of the known limitations include:
6416 only very simple character constant macros are translated into Ada
6417 constants. Function macros (macros with arguments) are partially translated
6418 as comments, to be completed manually if needed.
6421 some extensions (e.g. vector types) are not supported
6424 pointers to pointers or complex structures are mapped to System.Address
6427 identifiers with identical name (except casing) will generate compilation
6428 errors (e.g. @code{shm_get} vs @code{SHM_GET}).
6431 The code is generated using Ada 2012 syntax, which makes it easier to interface
6432 with other languages. In most cases you can still use the generated binding
6433 even if your code is compiled using earlier versions of Ada (e.g. @code{-gnat95}).
6436 * Running the Binding Generator::
6437 * Generating Bindings for C++ Headers::
6442 @node Running the Binding Generator,Generating Bindings for C++ Headers,,Generating Ada Bindings for C and C++ headers
6443 @anchor{gnat_ugn/the_gnat_compilation_model id71}@anchor{b1}@anchor{gnat_ugn/the_gnat_compilation_model running-the-binding-generator}@anchor{b2}
6444 @subsubsection Running the Binding Generator
6447 The binding generator is part of the @code{gcc} compiler and can be
6448 invoked via the @code{-fdump-ada-spec} switch, which will generate Ada
6449 spec files for the header files specified on the command line, and all
6450 header files needed by these files transitively. For example:
6453 $ g++ -c -fdump-ada-spec -C /usr/include/time.h
6457 will generate, under GNU/Linux, the following files: @code{time_h.ads},
6458 @code{bits_time_h.ads}, @code{stddef_h.ads}, @code{bits_types_h.ads} which
6459 correspond to the files @code{/usr/include/time.h},
6460 @code{/usr/include/bits/time.h}, etc…, and will then compile these Ada specs
6463 The @code{-C} switch tells @code{gcc} to extract comments from headers,
6464 and will attempt to generate corresponding Ada comments.
6466 If you want to generate a single Ada file and not the transitive closure, you
6467 can use instead the @code{-fdump-ada-spec-slim} switch.
6469 You can optionally specify a parent unit, of which all generated units will
6470 be children, using @code{-fada-spec-parent=@emph{unit}}.
6472 Note that we recommend when possible to use the @emph{g++} driver to
6473 generate bindings, even for most C headers, since this will in general
6474 generate better Ada specs. For generating bindings for C++ headers, it is
6475 mandatory to use the @emph{g++} command, or @emph{gcc -x c++} which
6476 is equivalent in this case. If @emph{g++} cannot work on your C headers
6477 because of incompatibilities between C and C++, then you can fallback to
6480 For an example of better bindings generated from the C++ front-end,
6481 the name of the parameters (when available) are actually ignored by the C
6482 front-end. Consider the following C header:
6485 extern void foo (int variable);
6488 with the C front-end, @code{variable} is ignored, and the above is handled as:
6491 extern void foo (int);
6494 generating a generic:
6497 procedure foo (param1 : int);
6500 with the C++ front-end, the name is available, and we generate:
6503 procedure foo (variable : int);
6506 In some cases, the generated bindings will be more complete or more meaningful
6507 when defining some macros, which you can do via the @code{-D} switch. This
6508 is for example the case with @code{Xlib.h} under GNU/Linux:
6511 $ g++ -c -fdump-ada-spec -DXLIB_ILLEGAL_ACCESS -C /usr/include/X11/Xlib.h
6514 The above will generate more complete bindings than a straight call without
6515 the @code{-DXLIB_ILLEGAL_ACCESS} switch.
6517 In other cases, it is not possible to parse a header file in a stand-alone
6518 manner, because other include files need to be included first. In this
6519 case, the solution is to create a small header file including the needed
6520 @code{#include} and possible @code{#define} directives. For example, to
6521 generate Ada bindings for @code{readline/readline.h}, you need to first
6522 include @code{stdio.h}, so you can create a file with the following two
6523 lines in e.g. @code{readline1.h}:
6527 #include <readline/readline.h>
6530 and then generate Ada bindings from this file:
6533 $ g++ -c -fdump-ada-spec readline1.h
6536 @node Generating Bindings for C++ Headers,Switches,Running the Binding Generator,Generating Ada Bindings for C and C++ headers
6537 @anchor{gnat_ugn/the_gnat_compilation_model generating-bindings-for-c-headers}@anchor{b3}@anchor{gnat_ugn/the_gnat_compilation_model id72}@anchor{b4}
6538 @subsubsection Generating Bindings for C++ Headers
6541 Generating bindings for C++ headers is done using the same options, always
6542 with the @emph{g++} compiler. Note that generating Ada spec from C++ headers is a
6543 much more complex job and support for C++ headers is much more limited that
6544 support for C headers. As a result, you will need to modify the resulting
6545 bindings by hand more extensively when using C++ headers.
6547 In this mode, C++ classes will be mapped to Ada tagged types, constructors
6548 will be mapped using the @code{CPP_Constructor} pragma, and when possible,
6549 multiple inheritance of abstract classes will be mapped to Ada interfaces
6550 (see the @emph{Interfacing to C++} section in the @cite{GNAT Reference Manual}
6551 for additional information on interfacing to C++).
6553 For example, given the following C++ header file:
6558 virtual int Number_Of_Teeth () = 0;
6563 virtual void Set_Owner (char* Name) = 0;
6569 virtual void Set_Age (int New_Age);
6572 class Dog : Animal, Carnivore, Domestic @{
6577 virtual int Number_Of_Teeth ();
6578 virtual void Set_Owner (char* Name);
6584 The corresponding Ada code is generated:
6587 package Class_Carnivore is
6588 type Carnivore is limited interface;
6589 pragma Import (CPP, Carnivore);
6591 function Number_Of_Teeth (this : access Carnivore) return int is abstract;
6593 use Class_Carnivore;
6595 package Class_Domestic is
6596 type Domestic is limited interface;
6597 pragma Import (CPP, Domestic);
6600 (this : access Domestic;
6601 Name : Interfaces.C.Strings.chars_ptr) is abstract;
6605 package Class_Animal is
6606 type Animal is tagged limited record
6607 Age_Count : aliased int;
6609 pragma Import (CPP, Animal);
6611 procedure Set_Age (this : access Animal; New_Age : int);
6612 pragma Import (CPP, Set_Age, "_ZN6Animal7Set_AgeEi");
6616 package Class_Dog is
6617 type Dog is new Animal and Carnivore and Domestic with record
6618 Tooth_Count : aliased int;
6619 Owner : Interfaces.C.Strings.chars_ptr;
6621 pragma Import (CPP, Dog);
6623 function Number_Of_Teeth (this : access Dog) return int;
6624 pragma Import (CPP, Number_Of_Teeth, "_ZN3Dog15Number_Of_TeethEv");
6627 (this : access Dog; Name : Interfaces.C.Strings.chars_ptr);
6628 pragma Import (CPP, Set_Owner, "_ZN3Dog9Set_OwnerEPc");
6630 function New_Dog return Dog;
6631 pragma CPP_Constructor (New_Dog);
6632 pragma Import (CPP, New_Dog, "_ZN3DogC1Ev");
6637 @node Switches,,Generating Bindings for C++ Headers,Generating Ada Bindings for C and C++ headers
6638 @anchor{gnat_ugn/the_gnat_compilation_model switches}@anchor{b5}@anchor{gnat_ugn/the_gnat_compilation_model switches-for-ada-binding-generation}@anchor{b6}
6639 @subsubsection Switches
6642 @geindex -fdump-ada-spec (gcc)
6647 @item @code{-fdump-ada-spec}
6649 Generate Ada spec files for the given header files transitively (including
6650 all header files that these headers depend upon).
6653 @geindex -fdump-ada-spec-slim (gcc)
6658 @item @code{-fdump-ada-spec-slim}
6660 Generate Ada spec files for the header files specified on the command line
6664 @geindex -fada-spec-parent (gcc)
6669 @item @code{-fada-spec-parent=@emph{unit}}
6671 Specifies that all files generated by @code{-fdump-ada-spec} are
6672 to be child units of the specified parent unit.
6682 Extract comments from headers and generate Ada comments in the Ada spec files.
6685 @node Generating C Headers for Ada Specifications,,Generating Ada Bindings for C and C++ headers,Mixed Language Programming
6686 @anchor{gnat_ugn/the_gnat_compilation_model generating-c-headers-for-ada-specifications}@anchor{b7}@anchor{gnat_ugn/the_gnat_compilation_model id73}@anchor{b8}
6687 @subsection Generating C Headers for Ada Specifications
6690 @geindex Binding generation (for Ada specs)
6692 @geindex C headers (binding generation)
6694 GNAT includes a C header generator for Ada specifications which supports
6695 Ada types that have a direct mapping to C types. This includes in particular
6711 Composition of the above types
6714 Constant declarations
6720 Subprogram declarations
6724 * Running the C Header Generator::
6728 @node Running the C Header Generator,,,Generating C Headers for Ada Specifications
6729 @anchor{gnat_ugn/the_gnat_compilation_model running-the-c-header-generator}@anchor{b9}
6730 @subsubsection Running the C Header Generator
6733 The C header generator is part of the GNAT compiler and can be invoked via
6734 the @code{-gnatceg} combination of switches, which will generate a @code{.h}
6735 file corresponding to the given input file (Ada spec or body). Note that
6736 only spec files are processed in any case, so giving a spec or a body file
6737 as input is equivalent. For example:
6740 $ gcc -c -gnatceg pack1.ads
6743 will generate a self-contained file called @code{pack1.h} including
6744 common definitions from the Ada Standard package, followed by the
6745 definitions included in @code{pack1.ads}, as well as all the other units
6746 withed by this file.
6748 For instance, given the following Ada files:
6752 type Int is range 1 .. 10;
6761 Field1, Field2 : Pack2.Int;
6764 Global : Rec := (1, 2);
6766 procedure Proc1 (R : Rec);
6767 procedure Proc2 (R : in out Rec);
6771 The above @code{gcc} command will generate the following @code{pack1.h} file:
6774 /* Standard definitions skipped */
6777 typedef short_short_integer pack2__TintB;
6778 typedef pack2__TintB pack2__int;
6779 #endif /* PACK2_ADS */
6783 typedef struct _pack1__rec @{
6787 extern pack1__rec pack1__global;
6788 extern void pack1__proc1(const pack1__rec r);
6789 extern void pack1__proc2(pack1__rec *r);
6790 #endif /* PACK1_ADS */
6793 You can then @code{include} @code{pack1.h} from a C source file and use the types,
6794 call subprograms, reference objects, and constants.
6796 @node GNAT and Other Compilation Models,Using GNAT Files with External Tools,Mixed Language Programming,The GNAT Compilation Model
6797 @anchor{gnat_ugn/the_gnat_compilation_model gnat-and-other-compilation-models}@anchor{2d}@anchor{gnat_ugn/the_gnat_compilation_model id74}@anchor{ba}
6798 @section GNAT and Other Compilation Models
6801 This section compares the GNAT model with the approaches taken in
6802 other environents, first the C/C++ model and then the mechanism that
6803 has been used in other Ada systems, in particular those traditionally
6807 * Comparison between GNAT and C/C++ Compilation Models::
6808 * Comparison between GNAT and Conventional Ada Library Models::
6812 @node Comparison between GNAT and C/C++ Compilation Models,Comparison between GNAT and Conventional Ada Library Models,,GNAT and Other Compilation Models
6813 @anchor{gnat_ugn/the_gnat_compilation_model comparison-between-gnat-and-c-c-compilation-models}@anchor{bb}@anchor{gnat_ugn/the_gnat_compilation_model id75}@anchor{bc}
6814 @subsection Comparison between GNAT and C/C++ Compilation Models
6817 The GNAT model of compilation is close to the C and C++ models. You can
6818 think of Ada specs as corresponding to header files in C. As in C, you
6819 don’t need to compile specs; they are compiled when they are used. The
6820 Ada @emph{with} is similar in effect to the @code{#include} of a C
6823 One notable difference is that, in Ada, you may compile specs separately
6824 to check them for semantic and syntactic accuracy. This is not always
6825 possible with C headers because they are fragments of programs that have
6826 less specific syntactic or semantic rules.
6828 The other major difference is the requirement for running the binder,
6829 which performs two important functions. First, it checks for
6830 consistency. In C or C++, the only defense against assembling
6831 inconsistent programs lies outside the compiler, in a makefile, for
6832 example. The binder satisfies the Ada requirement that it be impossible
6833 to construct an inconsistent program when the compiler is used in normal
6836 @geindex Elaboration order control
6838 The other important function of the binder is to deal with elaboration
6839 issues. There are also elaboration issues in C++ that are handled
6840 automatically. This automatic handling has the advantage of being
6841 simpler to use, but the C++ programmer has no control over elaboration.
6842 Where @code{gnatbind} might complain there was no valid order of
6843 elaboration, a C++ compiler would simply construct a program that
6844 malfunctioned at run time.
6846 @node Comparison between GNAT and Conventional Ada Library Models,,Comparison between GNAT and C/C++ Compilation Models,GNAT and Other Compilation Models
6847 @anchor{gnat_ugn/the_gnat_compilation_model comparison-between-gnat-and-conventional-ada-library-models}@anchor{bd}@anchor{gnat_ugn/the_gnat_compilation_model id76}@anchor{be}
6848 @subsection Comparison between GNAT and Conventional Ada Library Models
6851 This section is intended for Ada programmers who have
6852 used an Ada compiler implementing the traditional Ada library
6853 model, as described in the Ada Reference Manual.
6855 @geindex GNAT library
6857 In GNAT, there is no ‘library’ in the normal sense. Instead, the set of
6858 source files themselves acts as the library. Compiling Ada programs does
6859 not generate any centralized information, but rather an object file and
6860 a ALI file, which are of interest only to the binder and linker.
6861 In a traditional system, the compiler reads information not only from
6862 the source file being compiled, but also from the centralized library.
6863 This means that the effect of a compilation depends on what has been
6864 previously compiled. In particular:
6870 When a unit is @emph{with}ed, the unit seen by the compiler corresponds
6871 to the version of the unit most recently compiled into the library.
6874 Inlining is effective only if the necessary body has already been
6875 compiled into the library.
6878 Compiling a unit may obsolete other units in the library.
6881 In GNAT, compiling one unit never affects the compilation of any other
6882 units because the compiler reads only source files. Only changes to source
6883 files can affect the results of a compilation. In particular:
6889 When a unit is @emph{with}ed, the unit seen by the compiler corresponds
6890 to the source version of the unit that is currently accessible to the
6896 Inlining requires the appropriate source files for the package or
6897 subprogram bodies to be available to the compiler. Inlining is always
6898 effective, independent of the order in which units are compiled.
6901 Compiling a unit never affects any other compilations. The editing of
6902 sources may cause previous compilations to be out of date if they
6903 depended on the source file being modified.
6906 The most important result of these differences is that order of compilation
6907 is never significant in GNAT. There is no situation in which one is
6908 required to do one compilation before another. What shows up as order of
6909 compilation requirements in the traditional Ada library becomes, in
6910 GNAT, simple source dependencies; in other words, there is only a set
6911 of rules saying what source files must be present when a file is
6914 @node Using GNAT Files with External Tools,,GNAT and Other Compilation Models,The GNAT Compilation Model
6915 @anchor{gnat_ugn/the_gnat_compilation_model id77}@anchor{bf}@anchor{gnat_ugn/the_gnat_compilation_model using-gnat-files-with-external-tools}@anchor{2e}
6916 @section Using GNAT Files with External Tools
6919 This section explains how files that are produced by GNAT may be
6920 used with tools designed for other languages.
6923 * Using Other Utility Programs with GNAT::
6924 * The External Symbol Naming Scheme of GNAT::
6928 @node Using Other Utility Programs with GNAT,The External Symbol Naming Scheme of GNAT,,Using GNAT Files with External Tools
6929 @anchor{gnat_ugn/the_gnat_compilation_model id78}@anchor{c0}@anchor{gnat_ugn/the_gnat_compilation_model using-other-utility-programs-with-gnat}@anchor{c1}
6930 @subsection Using Other Utility Programs with GNAT
6933 The object files generated by GNAT are in standard system format and in
6934 particular the debugging information uses this format. This means
6935 programs generated by GNAT can be used with existing utilities that
6936 depend on these formats.
6938 In general, any utility program that works with C will also often work with
6939 Ada programs generated by GNAT. This includes software utilities such as
6940 gprof (a profiling program), gdb (the FSF debugger), and utilities such
6943 @node The External Symbol Naming Scheme of GNAT,,Using Other Utility Programs with GNAT,Using GNAT Files with External Tools
6944 @anchor{gnat_ugn/the_gnat_compilation_model id79}@anchor{c2}@anchor{gnat_ugn/the_gnat_compilation_model the-external-symbol-naming-scheme-of-gnat}@anchor{c3}
6945 @subsection The External Symbol Naming Scheme of GNAT
6948 In order to interpret the output from GNAT, when using tools that are
6949 originally intended for use with other languages, it is useful to
6950 understand the conventions used to generate link names from the Ada
6953 All link names are in all lowercase letters. With the exception of library
6954 procedure names, the mechanism used is simply to use the full expanded
6955 Ada name with dots replaced by double underscores. For example, suppose
6956 we have the following package spec:
6964 @geindex pragma Export
6966 The variable @code{MN} has a full expanded Ada name of @code{QRS.MN}, so
6967 the corresponding link name is @code{qrs__mn}.
6968 Of course if a @code{pragma Export} is used this may be overridden:
6973 pragma Export (Var1, C, External_Name => "var1_name");
6975 pragma Export (Var2, C, Link_Name => "var2_link_name");
6979 In this case, the link name for @code{Var1} is whatever link name the
6980 C compiler would assign for the C function @code{var1_name}. This typically
6981 would be either @code{var1_name} or @code{_var1_name}, depending on operating
6982 system conventions, but other possibilities exist. The link name for
6983 @code{Var2} is @code{var2_link_name}, and this is not operating system
6986 One exception occurs for library level procedures. A potential ambiguity
6987 arises between the required name @code{_main} for the C main program,
6988 and the name we would otherwise assign to an Ada library level procedure
6989 called @code{Main} (which might well not be the main program).
6991 To avoid this ambiguity, we attach the prefix @code{_ada_} to such
6992 names. So if we have a library level procedure such as:
6995 procedure Hello (S : String);
6998 the external name of this procedure will be @code{_ada_hello}.
7000 @c -- Example: A |withing| unit has a |with| clause, it |withs| a |withed| unit
7002 @node Building Executable Programs with GNAT,GNAT Utility Programs,The GNAT Compilation Model,Top
7003 @anchor{gnat_ugn/building_executable_programs_with_gnat doc}@anchor{c4}@anchor{gnat_ugn/building_executable_programs_with_gnat building-executable-programs-with-gnat}@anchor{a}@anchor{gnat_ugn/building_executable_programs_with_gnat id1}@anchor{c5}
7004 @chapter Building Executable Programs with GNAT
7007 This chapter describes first the gnatmake tool
7008 (@ref{c6,,Building with gnatmake}),
7009 which automatically determines the set of sources
7010 needed by an Ada compilation unit and executes the necessary
7011 (re)compilations, binding and linking.
7012 It also explains how to use each tool individually: the
7013 compiler (gcc, see @ref{c7,,Compiling with gcc}),
7014 binder (gnatbind, see @ref{c8,,Binding with gnatbind}),
7015 and linker (gnatlink, see @ref{c9,,Linking with gnatlink})
7016 to build executable programs.
7017 Finally, this chapter provides examples of
7018 how to make use of the general GNU make mechanism
7019 in a GNAT context (see @ref{70,,Using the GNU make Utility}).
7023 * Building with gnatmake::
7024 * Compiling with gcc::
7025 * Compiler Switches::
7027 * Binding with gnatbind::
7028 * Linking with gnatlink::
7029 * Using the GNU make Utility::
7033 @node Building with gnatmake,Compiling with gcc,,Building Executable Programs with GNAT
7034 @anchor{gnat_ugn/building_executable_programs_with_gnat building-with-gnatmake}@anchor{ca}@anchor{gnat_ugn/building_executable_programs_with_gnat the-gnat-make-program-gnatmake}@anchor{c6}
7035 @section Building with @code{gnatmake}
7040 A typical development cycle when working on an Ada program consists of
7041 the following steps:
7047 Edit some sources to fix bugs;
7053 Compile all sources affected;
7056 Rebind and relink; and
7062 @geindex Dependency rules (compilation)
7064 The third step in particular can be tricky, because not only do the modified
7065 files have to be compiled, but any files depending on these files must also be
7066 recompiled. The dependency rules in Ada can be quite complex, especially
7067 in the presence of overloading, @code{use} clauses, generics and inlined
7070 @code{gnatmake} automatically takes care of the third and fourth steps
7071 of this process. It determines which sources need to be compiled,
7072 compiles them, and binds and links the resulting object files.
7074 Unlike some other Ada make programs, the dependencies are always
7075 accurately recomputed from the new sources. The source based approach of
7076 the GNAT compilation model makes this possible. This means that if
7077 changes to the source program cause corresponding changes in
7078 dependencies, they will always be tracked exactly correctly by
7081 Note that for advanced forms of project structure, we recommend creating
7082 a project file as explained in the @emph{GNAT_Project_Manager} chapter in the
7083 @emph{GPRbuild User’s Guide}, and using the
7084 @code{gprbuild} tool which supports building with project files and works similarly
7088 * Running gnatmake::
7089 * Switches for gnatmake::
7090 * Mode Switches for gnatmake::
7091 * Notes on the Command Line::
7092 * How gnatmake Works::
7093 * Examples of gnatmake Usage::
7097 @node Running gnatmake,Switches for gnatmake,,Building with gnatmake
7098 @anchor{gnat_ugn/building_executable_programs_with_gnat id2}@anchor{cb}@anchor{gnat_ugn/building_executable_programs_with_gnat running-gnatmake}@anchor{cc}
7099 @subsection Running @code{gnatmake}
7102 The usual form of the @code{gnatmake} command is
7105 $ gnatmake [<switches>] <file_name> [<file_names>] [<mode_switches>]
7108 The only required argument is one @code{file_name}, which specifies
7109 a compilation unit that is a main program. Several @code{file_names} can be
7110 specified: this will result in several executables being built.
7111 If @code{switches} are present, they can be placed before the first
7112 @code{file_name}, between @code{file_names} or after the last @code{file_name}.
7113 If @code{mode_switches} are present, they must always be placed after
7114 the last @code{file_name} and all @code{switches}.
7116 If you are using standard file extensions (@code{.adb} and
7117 @code{.ads}), then the
7118 extension may be omitted from the @code{file_name} arguments. However, if
7119 you are using non-standard extensions, then it is required that the
7120 extension be given. A relative or absolute directory path can be
7121 specified in a @code{file_name}, in which case, the input source file will
7122 be searched for in the specified directory only. Otherwise, the input
7123 source file will first be searched in the directory where
7124 @code{gnatmake} was invoked and if it is not found, it will be search on
7125 the source path of the compiler as described in
7126 @ref{73,,Search Paths and the Run-Time Library (RTL)}.
7128 All @code{gnatmake} output (except when you specify @code{-M}) is sent to
7129 @code{stderr}. The output produced by the
7130 @code{-M} switch is sent to @code{stdout}.
7132 @node Switches for gnatmake,Mode Switches for gnatmake,Running gnatmake,Building with gnatmake
7133 @anchor{gnat_ugn/building_executable_programs_with_gnat id3}@anchor{cd}@anchor{gnat_ugn/building_executable_programs_with_gnat switches-for-gnatmake}@anchor{ce}
7134 @subsection Switches for @code{gnatmake}
7137 You may specify any of the following switches to @code{gnatmake}:
7139 @geindex --version (gnatmake)
7144 @item @code{--version}
7146 Display Copyright and version, then exit disregarding all other options.
7149 @geindex --help (gnatmake)
7156 If @code{--version} was not used, display usage, then exit disregarding
7160 @geindex --GCC=compiler_name (gnatmake)
7165 @item @code{--GCC=@emph{compiler_name}}
7167 Program used for compiling. The default is @code{gcc}. You need to use
7168 quotes around @code{compiler_name} if @code{compiler_name} contains
7169 spaces or other separator characters.
7170 As an example @code{--GCC="foo -x -y"}
7171 will instruct @code{gnatmake} to use @code{foo -x -y} as your
7172 compiler. A limitation of this syntax is that the name and path name of
7173 the executable itself must not include any embedded spaces. Note that
7174 switch @code{-c} is always inserted after your command name. Thus in the
7175 above example the compiler command that will be used by @code{gnatmake}
7176 will be @code{foo -c -x -y}. If several @code{--GCC=compiler_name} are
7177 used, only the last @code{compiler_name} is taken into account. However,
7178 all the additional switches are also taken into account. Thus,
7179 @code{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
7180 @code{--GCC="bar -x -y -z -t"}.
7183 @geindex --GNATBIND=binder_name (gnatmake)
7188 @item @code{--GNATBIND=@emph{binder_name}}
7190 Program used for binding. The default is @code{gnatbind}. You need to
7191 use quotes around @code{binder_name} if @code{binder_name} contains spaces
7192 or other separator characters.
7193 As an example @code{--GNATBIND="bar -x -y"}
7194 will instruct @code{gnatmake} to use @code{bar -x -y} as your
7195 binder. Binder switches that are normally appended by @code{gnatmake}
7196 to @code{gnatbind} are now appended to the end of @code{bar -x -y}.
7197 A limitation of this syntax is that the name and path name of the executable
7198 itself must not include any embedded spaces.
7201 @geindex --GNATLINK=linker_name (gnatmake)
7206 @item @code{--GNATLINK=@emph{linker_name}}
7208 Program used for linking. The default is @code{gnatlink}. You need to
7209 use quotes around @code{linker_name} if @code{linker_name} contains spaces
7210 or other separator characters.
7211 As an example @code{--GNATLINK="lan -x -y"}
7212 will instruct @code{gnatmake} to use @code{lan -x -y} as your
7213 linker. Linker switches that are normally appended by @code{gnatmake} to
7214 @code{gnatlink} are now appended to the end of @code{lan -x -y}.
7215 A limitation of this syntax is that the name and path name of the executable
7216 itself must not include any embedded spaces.
7218 @item @code{--create-map-file}
7220 When linking an executable, create a map file. The name of the map file
7221 has the same name as the executable with extension “.map”.
7223 @item @code{--create-map-file=@emph{mapfile}}
7225 When linking an executable, create a map file with the specified name.
7228 @geindex --create-missing-dirs (gnatmake)
7233 @item @code{--create-missing-dirs}
7235 When using project files (@code{-P@emph{project}}), automatically create
7236 missing object directories, library directories and exec
7239 @item @code{--single-compile-per-obj-dir}
7241 Disallow simultaneous compilations in the same object directory when
7242 project files are used.
7244 @item @code{--subdirs=@emph{subdir}}
7246 Actual object directory of each project file is the subdirectory subdir of the
7247 object directory specified or defaulted in the project file.
7249 @item @code{--unchecked-shared-lib-imports}
7251 By default, shared library projects are not allowed to import static library
7252 projects. When this switch is used on the command line, this restriction is
7255 @item @code{--source-info=@emph{source info file}}
7257 Specify a source info file. This switch is active only when project files
7258 are used. If the source info file is specified as a relative path, then it is
7259 relative to the object directory of the main project. If the source info file
7260 does not exist, then after the Project Manager has successfully parsed and
7261 processed the project files and found the sources, it creates the source info
7262 file. If the source info file already exists and can be read successfully,
7263 then the Project Manager will get all the needed information about the sources
7264 from the source info file and will not look for them. This reduces the time
7265 to process the project files, especially when looking for sources that take a
7266 long time. If the source info file exists but cannot be parsed successfully,
7267 the Project Manager will attempt to recreate it. If the Project Manager fails
7268 to create the source info file, a message is issued, but gnatmake does not
7269 fail. @code{gnatmake} “trusts” the source info file. This means that
7270 if the source files have changed (addition, deletion, moving to a different
7271 source directory), then the source info file need to be deleted and recreated.
7274 @geindex -a (gnatmake)
7281 Consider all files in the make process, even the GNAT internal system
7282 files (for example, the predefined Ada library files), as well as any
7283 locked files. Locked files are files whose ALI file is write-protected.
7285 @code{gnatmake} does not check these files,
7286 because the assumption is that the GNAT internal files are properly up
7287 to date, and also that any write protected ALI files have been properly
7288 installed. Note that if there is an installation problem, such that one
7289 of these files is not up to date, it will be properly caught by the
7291 You may have to specify this switch if you are working on GNAT
7292 itself. The switch @code{-a} is also useful
7293 in conjunction with @code{-f}
7294 if you need to recompile an entire application,
7295 including run-time files, using special configuration pragmas,
7296 such as a @code{Normalize_Scalars} pragma.
7299 @code{gnatmake -a} compiles all GNAT
7301 @code{gcc -c -gnatpg} rather than @code{gcc -c}.
7304 @geindex -b (gnatmake)
7311 Bind only. Can be combined with @code{-c} to do
7312 compilation and binding, but no link.
7313 Can be combined with @code{-l}
7314 to do binding and linking. When not combined with
7316 all the units in the closure of the main program must have been previously
7317 compiled and must be up to date. The root unit specified by @code{file_name}
7318 may be given without extension, with the source extension or, if no GNAT
7319 Project File is specified, with the ALI file extension.
7322 @geindex -c (gnatmake)
7329 Compile only. Do not perform binding, except when @code{-b}
7330 is also specified. Do not perform linking, except if both
7332 @code{-l} are also specified.
7333 If the root unit specified by @code{file_name} is not a main unit, this is the
7334 default. Otherwise @code{gnatmake} will attempt binding and linking
7335 unless all objects are up to date and the executable is more recent than
7339 @geindex -C (gnatmake)
7346 Use a temporary mapping file. A mapping file is a way to communicate
7347 to the compiler two mappings: from unit names to file names (without
7348 any directory information) and from file names to path names (with
7349 full directory information). A mapping file can make the compiler’s
7350 file searches faster, especially if there are many source directories,
7351 or the sources are read over a slow network connection. If
7352 @code{-P} is used, a mapping file is always used, so
7353 @code{-C} is unnecessary; in this case the mapping file
7354 is initially populated based on the project file. If
7355 @code{-C} is used without
7357 the mapping file is initially empty. Each invocation of the compiler
7358 will add any newly accessed sources to the mapping file.
7361 @geindex -C= (gnatmake)
7366 @item @code{-C=@emph{file}}
7368 Use a specific mapping file. The file, specified as a path name (absolute or
7369 relative) by this switch, should already exist, otherwise the switch is
7370 ineffective. The specified mapping file will be communicated to the compiler.
7371 This switch is not compatible with a project file
7372 (-P`file`) or with multiple compiling processes
7373 (-jnnn, when nnn is greater than 1).
7376 @geindex -d (gnatmake)
7383 Display progress for each source, up to date or not, as a single line:
7386 completed x out of y (zz%)
7389 If the file needs to be compiled this is displayed after the invocation of
7390 the compiler. These lines are displayed even in quiet output mode.
7393 @geindex -D (gnatmake)
7398 @item @code{-D @emph{dir}}
7400 Put all object files and ALI file in directory @code{dir}.
7401 If the @code{-D} switch is not used, all object files
7402 and ALI files go in the current working directory.
7404 This switch cannot be used when using a project file.
7407 @geindex -eI (gnatmake)
7412 @item @code{-eI@emph{nnn}}
7414 Indicates that the main source is a multi-unit source and the rank of the unit
7415 in the source file is nnn. nnn needs to be a positive number and a valid
7416 index in the source. This switch cannot be used when @code{gnatmake} is
7417 invoked for several mains.
7420 @geindex -eL (gnatmake)
7422 @geindex symbolic links
7429 Follow all symbolic links when processing project files.
7430 This should be used if your project uses symbolic links for files or
7431 directories, but is not needed in other cases.
7433 @geindex naming scheme
7435 This also assumes that no directory matches the naming scheme for files (for
7436 instance that you do not have a directory called “sources.ads” when using the
7437 default GNAT naming scheme).
7439 When you do not have to use this switch (i.e., by default), gnatmake is able to
7440 save a lot of system calls (several per source file and object file), which
7441 can result in a significant speed up to load and manipulate a project file,
7442 especially when using source files from a remote system.
7445 @geindex -eS (gnatmake)
7452 Output the commands for the compiler, the binder and the linker
7454 instead of standard error.
7457 @geindex -f (gnatmake)
7464 Force recompilations. Recompile all sources, even though some object
7465 files may be up to date, but don’t recompile predefined or GNAT internal
7466 files or locked files (files with a write-protected ALI file),
7467 unless the @code{-a} switch is also specified.
7470 @geindex -F (gnatmake)
7477 When using project files, if some errors or warnings are detected during
7478 parsing and verbose mode is not in effect (no use of switch
7479 -v), then error lines start with the full path name of the project
7480 file, rather than its simple file name.
7483 @geindex -g (gnatmake)
7490 Enable debugging. This switch is simply passed to the compiler and to the
7494 @geindex -i (gnatmake)
7501 In normal mode, @code{gnatmake} compiles all object files and ALI files
7502 into the current directory. If the @code{-i} switch is used,
7503 then instead object files and ALI files that already exist are overwritten
7504 in place. This means that once a large project is organized into separate
7505 directories in the desired manner, then @code{gnatmake} will automatically
7506 maintain and update this organization. If no ALI files are found on the
7507 Ada object path (see @ref{73,,Search Paths and the Run-Time Library (RTL)}),
7508 the new object and ALI files are created in the
7509 directory containing the source being compiled. If another organization
7510 is desired, where objects and sources are kept in different directories,
7511 a useful technique is to create dummy ALI files in the desired directories.
7512 When detecting such a dummy file, @code{gnatmake} will be forced to
7513 recompile the corresponding source file, and it will be put the resulting
7514 object and ALI files in the directory where it found the dummy file.
7517 @geindex -j (gnatmake)
7519 @geindex Parallel make
7524 @item @code{-j@emph{n}}
7526 Use @code{n} processes to carry out the (re)compilations. On a multiprocessor
7527 machine compilations will occur in parallel. If @code{n} is 0, then the
7528 maximum number of parallel compilations is the number of core processors
7529 on the platform. In the event of compilation errors, messages from various
7530 compilations might get interspersed (but @code{gnatmake} will give you the
7531 full ordered list of failing compiles at the end). If this is problematic,
7532 rerun the make process with n set to 1 to get a clean list of messages.
7535 @geindex -k (gnatmake)
7542 Keep going. Continue as much as possible after a compilation error. To
7543 ease the programmer’s task in case of compilation errors, the list of
7544 sources for which the compile fails is given when @code{gnatmake}
7547 If @code{gnatmake} is invoked with several @code{file_names} and with this
7548 switch, if there are compilation errors when building an executable,
7549 @code{gnatmake} will not attempt to build the following executables.
7552 @geindex -l (gnatmake)
7559 Link only. Can be combined with @code{-b} to binding
7560 and linking. Linking will not be performed if combined with
7562 but not with @code{-b}.
7563 When not combined with @code{-b}
7564 all the units in the closure of the main program must have been previously
7565 compiled and must be up to date, and the main program needs to have been bound.
7566 The root unit specified by @code{file_name}
7567 may be given without extension, with the source extension or, if no GNAT
7568 Project File is specified, with the ALI file extension.
7571 @geindex -m (gnatmake)
7578 Specify that the minimum necessary amount of recompilations
7579 be performed. In this mode @code{gnatmake} ignores time
7580 stamp differences when the only
7581 modifications to a source file consist in adding/removing comments,
7582 empty lines, spaces or tabs. This means that if you have changed the
7583 comments in a source file or have simply reformatted it, using this
7584 switch will tell @code{gnatmake} not to recompile files that depend on it
7585 (provided other sources on which these files depend have undergone no
7586 semantic modifications). Note that the debugging information may be
7587 out of date with respect to the sources if the @code{-m} switch causes
7588 a compilation to be switched, so the use of this switch represents a
7589 trade-off between compilation time and accurate debugging information.
7592 @geindex Dependencies
7593 @geindex producing list
7595 @geindex -M (gnatmake)
7602 Check if all objects are up to date. If they are, output the object
7603 dependences to @code{stdout} in a form that can be directly exploited in
7604 a @code{Makefile}. By default, each source file is prefixed with its
7605 (relative or absolute) directory name. This name is whatever you
7606 specified in the various @code{-aI}
7607 and @code{-I} switches. If you use
7608 @code{gnatmake -M} @code{-q}
7609 (see below), only the source file names,
7610 without relative paths, are output. If you just specify the @code{-M}
7611 switch, dependencies of the GNAT internal system files are omitted. This
7612 is typically what you want. If you also specify
7613 the @code{-a} switch,
7614 dependencies of the GNAT internal files are also listed. Note that
7615 dependencies of the objects in external Ada libraries (see
7616 switch @code{-aL@emph{dir}} in the following list)
7620 @geindex -n (gnatmake)
7627 Don’t compile, bind, or link. Checks if all objects are up to date.
7628 If they are not, the full name of the first file that needs to be
7629 recompiled is printed.
7630 Repeated use of this option, followed by compiling the indicated source
7631 file, will eventually result in recompiling all required units.
7634 @geindex -o (gnatmake)
7639 @item @code{-o @emph{exec_name}}
7641 Output executable name. The name of the final executable program will be
7642 @code{exec_name}. If the @code{-o} switch is omitted the default
7643 name for the executable will be the name of the input file in appropriate form
7644 for an executable file on the host system.
7646 This switch cannot be used when invoking @code{gnatmake} with several
7650 @geindex -p (gnatmake)
7657 Same as @code{--create-missing-dirs}
7660 @geindex -P (gnatmake)
7665 @item @code{-P@emph{project}}
7667 Use project file @code{project}. Only one such switch can be used.
7671 @c :ref:`gnatmake_and_Project_Files`.
7673 @geindex -q (gnatmake)
7680 Quiet. When this flag is not set, the commands carried out by
7681 @code{gnatmake} are displayed.
7684 @geindex -s (gnatmake)
7691 Recompile if compiler switches have changed since last compilation.
7692 All compiler switches but -I and -o are taken into account in the
7694 orders between different ‘first letter’ switches are ignored, but
7695 orders between same switches are taken into account. For example,
7696 @code{-O -O2} is different than @code{-O2 -O}, but @code{-g -O}
7697 is equivalent to @code{-O -g}.
7699 This switch is recommended when Integrated Preprocessing is used.
7702 @geindex -u (gnatmake)
7709 Unique. Recompile at most the main files. It implies -c. Combined with
7710 -f, it is equivalent to calling the compiler directly. Note that using
7711 -u with a project file and no main has a special meaning.
7715 @c (See :ref:`Project_Files_and_Main_Subprograms`.)
7717 @geindex -U (gnatmake)
7724 When used without a project file or with one or several mains on the command
7725 line, is equivalent to -u. When used with a project file and no main
7726 on the command line, all sources of all project files are checked and compiled
7727 if not up to date, and libraries are rebuilt, if necessary.
7730 @geindex -v (gnatmake)
7737 Verbose. Display the reason for all recompilations @code{gnatmake}
7738 decides are necessary, with the highest verbosity level.
7741 @geindex -vl (gnatmake)
7748 Verbosity level Low. Display fewer lines than in verbosity Medium.
7751 @geindex -vm (gnatmake)
7758 Verbosity level Medium. Potentially display fewer lines than in verbosity High.
7761 @geindex -vm (gnatmake)
7768 Verbosity level High. Equivalent to -v.
7770 @item @code{-vP@emph{x}}
7772 Indicate the verbosity of the parsing of GNAT project files.
7773 See @ref{cf,,Switches Related to Project Files}.
7776 @geindex -x (gnatmake)
7783 Indicate that sources that are not part of any Project File may be compiled.
7784 Normally, when using Project Files, only sources that are part of a Project
7785 File may be compile. When this switch is used, a source outside of all Project
7786 Files may be compiled. The ALI file and the object file will be put in the
7787 object directory of the main Project. The compilation switches used will only
7788 be those specified on the command line. Even when
7789 @code{-x} is used, mains specified on the
7790 command line need to be sources of a project file.
7792 @item @code{-X@emph{name}=@emph{value}}
7794 Indicate that external variable @code{name} has the value @code{value}.
7795 The Project Manager will use this value for occurrences of
7796 @code{external(name)} when parsing the project file.
7797 @ref{cf,,Switches Related to Project Files}.
7800 @geindex -z (gnatmake)
7807 No main subprogram. Bind and link the program even if the unit name
7808 given on the command line is a package name. The resulting executable
7809 will execute the elaboration routines of the package and its closure,
7810 then the finalization routines.
7813 @subsubheading GCC switches
7816 Any uppercase or multi-character switch that is not a @code{gnatmake} switch
7817 is passed to @code{gcc} (e.g., @code{-O}, @code{-gnato,} etc.)
7819 @subsubheading Source and library search path switches
7822 @geindex -aI (gnatmake)
7827 @item @code{-aI@emph{dir}}
7829 When looking for source files also look in directory @code{dir}.
7830 The order in which source files search is undertaken is
7831 described in @ref{73,,Search Paths and the Run-Time Library (RTL)}.
7834 @geindex -aL (gnatmake)
7839 @item @code{-aL@emph{dir}}
7841 Consider @code{dir} as being an externally provided Ada library.
7842 Instructs @code{gnatmake} to skip compilation units whose @code{.ALI}
7843 files have been located in directory @code{dir}. This allows you to have
7844 missing bodies for the units in @code{dir} and to ignore out of date bodies
7845 for the same units. You still need to specify
7846 the location of the specs for these units by using the switches
7847 @code{-aI@emph{dir}} or @code{-I@emph{dir}}.
7848 Note: this switch is provided for compatibility with previous versions
7849 of @code{gnatmake}. The easier method of causing standard libraries
7850 to be excluded from consideration is to write-protect the corresponding
7854 @geindex -aO (gnatmake)
7859 @item @code{-aO@emph{dir}}
7861 When searching for library and object files, look in directory
7862 @code{dir}. The order in which library files are searched is described in
7863 @ref{76,,Search Paths for gnatbind}.
7866 @geindex Search paths
7867 @geindex for gnatmake
7869 @geindex -A (gnatmake)
7874 @item @code{-A@emph{dir}}
7876 Equivalent to @code{-aL@emph{dir}} @code{-aI@emph{dir}}.
7878 @geindex -I (gnatmake)
7880 @item @code{-I@emph{dir}}
7882 Equivalent to @code{-aO@emph{dir} -aI@emph{dir}}.
7885 @geindex -I- (gnatmake)
7887 @geindex Source files
7888 @geindex suppressing search
7895 Do not look for source files in the directory containing the source
7896 file named in the command line.
7897 Do not look for ALI or object files in the directory
7898 where @code{gnatmake} was invoked.
7901 @geindex -L (gnatmake)
7903 @geindex Linker libraries
7908 @item @code{-L@emph{dir}}
7910 Add directory @code{dir} to the list of directories in which the linker
7911 will search for libraries. This is equivalent to
7912 @code{-largs} @code{-L@emph{dir}}.
7913 Furthermore, under Windows, the sources pointed to by the libraries path
7914 set in the registry are not searched for.
7917 @geindex -nostdinc (gnatmake)
7922 @item @code{-nostdinc}
7924 Do not look for source files in the system default directory.
7927 @geindex -nostdlib (gnatmake)
7932 @item @code{-nostdlib}
7934 Do not look for library files in the system default directory.
7937 @geindex --RTS (gnatmake)
7942 @item @code{--RTS=@emph{rts-path}}
7944 Specifies the default location of the run-time library. GNAT looks for the
7946 in the following directories, and stops as soon as a valid run-time is found
7947 (@code{adainclude} or @code{ada_source_path}, and @code{adalib} or
7948 @code{ada_object_path} present):
7954 @emph{<current directory>/$rts_path}
7957 @emph{<default-search-dir>/$rts_path}
7960 @emph{<default-search-dir>/rts-$rts_path}
7963 The selected path is handled like a normal RTS path.
7967 @node Mode Switches for gnatmake,Notes on the Command Line,Switches for gnatmake,Building with gnatmake
7968 @anchor{gnat_ugn/building_executable_programs_with_gnat id4}@anchor{d0}@anchor{gnat_ugn/building_executable_programs_with_gnat mode-switches-for-gnatmake}@anchor{d1}
7969 @subsection Mode Switches for @code{gnatmake}
7972 The mode switches (referred to as @code{mode_switches}) allow the
7973 inclusion of switches that are to be passed to the compiler itself, the
7974 binder or the linker. The effect of a mode switch is to cause all
7975 subsequent switches up to the end of the switch list, or up to the next
7976 mode switch, to be interpreted as switches to be passed on to the
7977 designated component of GNAT.
7979 @geindex -cargs (gnatmake)
7984 @item @code{-cargs @emph{switches}}
7986 Compiler switches. Here @code{switches} is a list of switches
7987 that are valid switches for @code{gcc}. They will be passed on to
7988 all compile steps performed by @code{gnatmake}.
7991 @geindex -bargs (gnatmake)
7996 @item @code{-bargs @emph{switches}}
7998 Binder switches. Here @code{switches} is a list of switches
7999 that are valid switches for @code{gnatbind}. They will be passed on to
8000 all bind steps performed by @code{gnatmake}.
8003 @geindex -largs (gnatmake)
8008 @item @code{-largs @emph{switches}}
8010 Linker switches. Here @code{switches} is a list of switches
8011 that are valid switches for @code{gnatlink}. They will be passed on to
8012 all link steps performed by @code{gnatmake}.
8015 @geindex -margs (gnatmake)
8020 @item @code{-margs @emph{switches}}
8022 Make switches. The switches are directly interpreted by @code{gnatmake},
8023 regardless of any previous occurrence of @code{-cargs}, @code{-bargs}
8027 @node Notes on the Command Line,How gnatmake Works,Mode Switches for gnatmake,Building with gnatmake
8028 @anchor{gnat_ugn/building_executable_programs_with_gnat id5}@anchor{d2}@anchor{gnat_ugn/building_executable_programs_with_gnat notes-on-the-command-line}@anchor{d3}
8029 @subsection Notes on the Command Line
8032 This section contains some additional useful notes on the operation
8033 of the @code{gnatmake} command.
8035 @geindex Recompilation (by gnatmake)
8041 If @code{gnatmake} finds no ALI files, it recompiles the main program
8042 and all other units required by the main program.
8043 This means that @code{gnatmake}
8044 can be used for the initial compile, as well as during subsequent steps of
8045 the development cycle.
8048 If you enter @code{gnatmake foo.adb}, where @code{foo}
8049 is a subunit or body of a generic unit, @code{gnatmake} recompiles
8050 @code{foo.adb} (because it finds no ALI) and stops, issuing a
8054 In @code{gnatmake} the switch @code{-I}
8055 is used to specify both source and
8056 library file paths. Use @code{-aI}
8057 instead if you just want to specify
8058 source paths only and @code{-aO}
8059 if you want to specify library paths
8063 @code{gnatmake} will ignore any files whose ALI file is write-protected.
8064 This may conveniently be used to exclude standard libraries from
8065 consideration and in particular it means that the use of the
8066 @code{-f} switch will not recompile these files
8067 unless @code{-a} is also specified.
8070 @code{gnatmake} has been designed to make the use of Ada libraries
8071 particularly convenient. Assume you have an Ada library organized
8072 as follows: @emph{obj-dir} contains the objects and ALI files for
8073 of your Ada compilation units,
8074 whereas @emph{include-dir} contains the
8075 specs of these units, but no bodies. Then to compile a unit
8076 stored in @code{main.adb}, which uses this Ada library you would just type:
8079 $ gnatmake -aI`include-dir` -aL`obj-dir` main
8083 Using @code{gnatmake} along with the @code{-m (minimal recompilation)}
8084 switch provides a mechanism for avoiding unnecessary recompilations. Using
8086 you can update the comments/format of your
8087 source files without having to recompile everything. Note, however, that
8088 adding or deleting lines in a source files may render its debugging
8089 info obsolete. If the file in question is a spec, the impact is rather
8090 limited, as that debugging info will only be useful during the
8091 elaboration phase of your program. For bodies the impact can be more
8092 significant. In all events, your debugger will warn you if a source file
8093 is more recent than the corresponding object, and alert you to the fact
8094 that the debugging information may be out of date.
8097 @node How gnatmake Works,Examples of gnatmake Usage,Notes on the Command Line,Building with gnatmake
8098 @anchor{gnat_ugn/building_executable_programs_with_gnat how-gnatmake-works}@anchor{d4}@anchor{gnat_ugn/building_executable_programs_with_gnat id6}@anchor{d5}
8099 @subsection How @code{gnatmake} Works
8102 Generally @code{gnatmake} automatically performs all necessary
8103 recompilations and you don’t need to worry about how it works. However,
8104 it may be useful to have some basic understanding of the @code{gnatmake}
8105 approach and in particular to understand how it uses the results of
8106 previous compilations without incorrectly depending on them.
8108 First a definition: an object file is considered @emph{up to date} if the
8109 corresponding ALI file exists and if all the source files listed in the
8110 dependency section of this ALI file have time stamps matching those in
8111 the ALI file. This means that neither the source file itself nor any
8112 files that it depends on have been modified, and hence there is no need
8113 to recompile this file.
8115 @code{gnatmake} works by first checking if the specified main unit is up
8116 to date. If so, no compilations are required for the main unit. If not,
8117 @code{gnatmake} compiles the main program to build a new ALI file that
8118 reflects the latest sources. Then the ALI file of the main unit is
8119 examined to find all the source files on which the main program depends,
8120 and @code{gnatmake} recursively applies the above procedure on all these
8123 This process ensures that @code{gnatmake} only trusts the dependencies
8124 in an existing ALI file if they are known to be correct. Otherwise it
8125 always recompiles to determine a new, guaranteed accurate set of
8126 dependencies. As a result the program is compiled ‘upside down’ from what may
8127 be more familiar as the required order of compilation in some other Ada
8128 systems. In particular, clients are compiled before the units on which
8129 they depend. The ability of GNAT to compile in any order is critical in
8130 allowing an order of compilation to be chosen that guarantees that
8131 @code{gnatmake} will recompute a correct set of new dependencies if
8134 When invoking @code{gnatmake} with several @code{file_names}, if a unit is
8135 imported by several of the executables, it will be recompiled at most once.
8137 Note: when using non-standard naming conventions
8138 (@ref{1c,,Using Other File Names}), changing through a configuration pragmas
8139 file the version of a source and invoking @code{gnatmake} to recompile may
8140 have no effect, if the previous version of the source is still accessible
8141 by @code{gnatmake}. It may be necessary to use the switch
8144 @node Examples of gnatmake Usage,,How gnatmake Works,Building with gnatmake
8145 @anchor{gnat_ugn/building_executable_programs_with_gnat examples-of-gnatmake-usage}@anchor{d6}@anchor{gnat_ugn/building_executable_programs_with_gnat id7}@anchor{d7}
8146 @subsection Examples of @code{gnatmake} Usage
8152 @item @emph{gnatmake hello.adb}
8154 Compile all files necessary to bind and link the main program
8155 @code{hello.adb} (containing unit @code{Hello}) and bind and link the
8156 resulting object files to generate an executable file @code{hello}.
8158 @item @emph{gnatmake main1 main2 main3}
8160 Compile all files necessary to bind and link the main programs
8161 @code{main1.adb} (containing unit @code{Main1}), @code{main2.adb}
8162 (containing unit @code{Main2}) and @code{main3.adb}
8163 (containing unit @code{Main3}) and bind and link the resulting object files
8164 to generate three executable files @code{main1},
8165 @code{main2} and @code{main3}.
8167 @item @emph{gnatmake -q Main_Unit -cargs -O2 -bargs -l}
8169 Compile all files necessary to bind and link the main program unit
8170 @code{Main_Unit} (from file @code{main_unit.adb}). All compilations will
8171 be done with optimization level 2 and the order of elaboration will be
8172 listed by the binder. @code{gnatmake} will operate in quiet mode, not
8173 displaying commands it is executing.
8176 @node Compiling with gcc,Compiler Switches,Building with gnatmake,Building Executable Programs with GNAT
8177 @anchor{gnat_ugn/building_executable_programs_with_gnat compiling-with-gcc}@anchor{c7}@anchor{gnat_ugn/building_executable_programs_with_gnat id8}@anchor{d8}
8178 @section Compiling with @code{gcc}
8181 This section discusses how to compile Ada programs using the @code{gcc}
8182 command. It also describes the set of switches
8183 that can be used to control the behavior of the compiler.
8186 * Compiling Programs::
8187 * Search Paths and the Run-Time Library (RTL): Search Paths and the Run-Time Library RTL.
8188 * Order of Compilation Issues::
8193 @node Compiling Programs,Search Paths and the Run-Time Library RTL,,Compiling with gcc
8194 @anchor{gnat_ugn/building_executable_programs_with_gnat compiling-programs}@anchor{d9}@anchor{gnat_ugn/building_executable_programs_with_gnat id9}@anchor{da}
8195 @subsection Compiling Programs
8198 The first step in creating an executable program is to compile the units
8199 of the program using the @code{gcc} command. You must compile the
8206 the body file (@code{.adb}) for a library level subprogram or generic
8210 the spec file (@code{.ads}) for a library level package or generic
8211 package that has no body
8214 the body file (@code{.adb}) for a library level package
8215 or generic package that has a body
8218 You need @emph{not} compile the following files
8224 the spec of a library unit which has a body
8230 because they are compiled as part of compiling related units. GNAT
8232 when the corresponding body is compiled, and subunits when the parent is
8235 @geindex cannot generate code
8237 If you attempt to compile any of these files, you will get one of the
8238 following error messages (where @code{fff} is the name of the file you
8244 cannot generate code for file `@w{`}fff`@w{`} (package spec)
8245 to check package spec, use -gnatc
8247 cannot generate code for file `@w{`}fff`@w{`} (missing subunits)
8248 to check parent unit, use -gnatc
8250 cannot generate code for file `@w{`}fff`@w{`} (subprogram spec)
8251 to check subprogram spec, use -gnatc
8253 cannot generate code for file `@w{`}fff`@w{`} (subunit)
8254 to check subunit, use -gnatc
8258 As indicated by the above error messages, if you want to submit
8259 one of these files to the compiler to check for correct semantics
8260 without generating code, then use the @code{-gnatc} switch.
8262 The basic command for compiling a file containing an Ada unit is:
8265 $ gcc -c [switches] <file name>
8268 where @code{file name} is the name of the Ada file (usually
8269 having an extension @code{.ads} for a spec or @code{.adb} for a body).
8271 @code{-c} switch to tell @code{gcc} to compile, but not link, the file.
8272 The result of a successful compilation is an object file, which has the
8273 same name as the source file but an extension of @code{.o} and an Ada
8274 Library Information (ALI) file, which also has the same name as the
8275 source file, but with @code{.ali} as the extension. GNAT creates these
8276 two output files in the current directory, but you may specify a source
8277 file in any directory using an absolute or relative path specification
8278 containing the directory information.
8280 TESTING: the @code{--foobar@emph{NN}} switch
8284 @code{gcc} is actually a driver program that looks at the extensions of
8285 the file arguments and loads the appropriate compiler. For example, the
8286 GNU C compiler is @code{cc1}, and the Ada compiler is @code{gnat1}.
8287 These programs are in directories known to the driver program (in some
8288 configurations via environment variables you set), but need not be in
8289 your path. The @code{gcc} driver also calls the assembler and any other
8290 utilities needed to complete the generation of the required object
8293 It is possible to supply several file names on the same @code{gcc}
8294 command. This causes @code{gcc} to call the appropriate compiler for
8295 each file. For example, the following command lists two separate
8296 files to be compiled:
8299 $ gcc -c x.adb y.adb
8302 calls @code{gnat1} (the Ada compiler) twice to compile @code{x.adb} and
8304 The compiler generates two object files @code{x.o} and @code{y.o}
8305 and the two ALI files @code{x.ali} and @code{y.ali}.
8307 Any switches apply to all the files listed, see @ref{db,,Compiler Switches} for a
8308 list of available @code{gcc} switches.
8310 @node Search Paths and the Run-Time Library RTL,Order of Compilation Issues,Compiling Programs,Compiling with gcc
8311 @anchor{gnat_ugn/building_executable_programs_with_gnat id10}@anchor{dc}@anchor{gnat_ugn/building_executable_programs_with_gnat search-paths-and-the-run-time-library-rtl}@anchor{73}
8312 @subsection Search Paths and the Run-Time Library (RTL)
8315 With the GNAT source-based library system, the compiler must be able to
8316 find source files for units that are needed by the unit being compiled.
8317 Search paths are used to guide this process.
8319 The compiler compiles one source file whose name must be given
8320 explicitly on the command line. In other words, no searching is done
8321 for this file. To find all other source files that are needed (the most
8322 common being the specs of units), the compiler examines the following
8323 directories, in the following order:
8329 The directory containing the source file of the main unit being compiled
8330 (the file name on the command line).
8333 Each directory named by an @code{-I} switch given on the @code{gcc}
8334 command line, in the order given.
8336 @geindex ADA_PRJ_INCLUDE_FILE
8339 Each of the directories listed in the text file whose name is given
8341 @geindex ADA_PRJ_INCLUDE_FILE
8342 @geindex environment variable; ADA_PRJ_INCLUDE_FILE
8343 @code{ADA_PRJ_INCLUDE_FILE} environment variable.
8344 @geindex ADA_PRJ_INCLUDE_FILE
8345 @geindex environment variable; ADA_PRJ_INCLUDE_FILE
8346 @code{ADA_PRJ_INCLUDE_FILE} is normally set by gnatmake or by the gnat
8347 driver when project files are used. It should not normally be set
8350 @geindex ADA_INCLUDE_PATH
8353 Each of the directories listed in the value of the
8354 @geindex ADA_INCLUDE_PATH
8355 @geindex environment variable; ADA_INCLUDE_PATH
8356 @code{ADA_INCLUDE_PATH} environment variable.
8357 Construct this value
8360 @geindex environment variable; PATH
8361 @code{PATH} environment variable: a list of directory
8362 names separated by colons (semicolons when working with the NT version).
8365 The content of the @code{ada_source_path} file which is part of the GNAT
8366 installation tree and is used to store standard libraries such as the
8367 GNAT Run Time Library (RTL) source files.
8368 @ref{72,,Installing a library}
8371 Specifying the switch @code{-I-}
8372 inhibits the use of the directory
8373 containing the source file named in the command line. You can still
8374 have this directory on your search path, but in this case it must be
8375 explicitly requested with a @code{-I} switch.
8377 Specifying the switch @code{-nostdinc}
8378 inhibits the search of the default location for the GNAT Run Time
8379 Library (RTL) source files.
8381 The compiler outputs its object files and ALI files in the current
8383 Caution: The object file can be redirected with the @code{-o} switch;
8384 however, @code{gcc} and @code{gnat1} have not been coordinated on this
8385 so the @code{ALI} file will not go to the right place. Therefore, you should
8386 avoid using the @code{-o} switch.
8390 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
8391 children make up the GNAT RTL, together with the simple @code{System.IO}
8392 package used in the @code{"Hello World"} example. The sources for these units
8393 are needed by the compiler and are kept together in one directory. Not
8394 all of the bodies are needed, but all of the sources are kept together
8395 anyway. In a normal installation, you need not specify these directory
8396 names when compiling or binding. Either the environment variables or
8397 the built-in defaults cause these files to be found.
8399 In addition to the language-defined hierarchies (@code{System}, @code{Ada} and
8400 @code{Interfaces}), the GNAT distribution provides a fourth hierarchy,
8401 consisting of child units of @code{GNAT}. This is a collection of generally
8402 useful types, subprograms, etc. See the @cite{GNAT_Reference_Manual}
8403 for further details.
8405 Besides simplifying access to the RTL, a major use of search paths is
8406 in compiling sources from multiple directories. This can make
8407 development environments much more flexible.
8409 @node Order of Compilation Issues,Examples,Search Paths and the Run-Time Library RTL,Compiling with gcc
8410 @anchor{gnat_ugn/building_executable_programs_with_gnat id11}@anchor{dd}@anchor{gnat_ugn/building_executable_programs_with_gnat order-of-compilation-issues}@anchor{de}
8411 @subsection Order of Compilation Issues
8414 If, in our earlier example, there was a spec for the @code{hello}
8415 procedure, it would be contained in the file @code{hello.ads}; yet this
8416 file would not have to be explicitly compiled. This is the result of the
8417 model we chose to implement library management. Some of the consequences
8418 of this model are as follows:
8424 There is no point in compiling specs (except for package
8425 specs with no bodies) because these are compiled as needed by clients. If
8426 you attempt a useless compilation, you will receive an error message.
8427 It is also useless to compile subunits because they are compiled as needed
8431 There are no order of compilation requirements: performing a
8432 compilation never obsoletes anything. The only way you can obsolete
8433 something and require recompilations is to modify one of the
8434 source files on which it depends.
8437 There is no library as such, apart from the ALI files
8438 (@ref{28,,The Ada Library Information Files}, for information on the format
8439 of these files). For now we find it convenient to create separate ALI files,
8440 but eventually the information therein may be incorporated into the object
8444 When you compile a unit, the source files for the specs of all units
8445 that it @emph{with}s, all its subunits, and the bodies of any generics it
8446 instantiates must be available (reachable by the search-paths mechanism
8447 described above), or you will receive a fatal error message.
8450 @node Examples,,Order of Compilation Issues,Compiling with gcc
8451 @anchor{gnat_ugn/building_executable_programs_with_gnat examples}@anchor{df}@anchor{gnat_ugn/building_executable_programs_with_gnat id12}@anchor{e0}
8452 @subsection Examples
8455 The following are some typical Ada compilation command line examples:
8461 Compile body in file @code{xyz.adb} with all default options.
8464 $ gcc -c -O2 -gnata xyz-def.adb
8467 Compile the child unit package in file @code{xyz-def.adb} with extensive
8468 optimizations, and pragma @code{Assert}/@cite{Debug} statements
8472 $ gcc -c -gnatc abc-def.adb
8475 Compile the subunit in file @code{abc-def.adb} in semantic-checking-only
8478 @node Compiler Switches,Linker Switches,Compiling with gcc,Building Executable Programs with GNAT
8479 @anchor{gnat_ugn/building_executable_programs_with_gnat compiler-switches}@anchor{e1}@anchor{gnat_ugn/building_executable_programs_with_gnat switches-for-gcc}@anchor{db}
8480 @section Compiler Switches
8483 The @code{gcc} command accepts switches that control the
8484 compilation process. These switches are fully described in this section:
8485 first an alphabetical listing of all switches with a brief description,
8486 and then functionally grouped sets of switches with more detailed
8489 More switches exist for GCC than those documented here, especially
8490 for specific targets. However, their use is not recommended as
8491 they may change code generation in ways that are incompatible with
8492 the Ada run-time library, or can cause inconsistencies between
8496 * Alphabetical List of All Switches::
8497 * Output and Error Message Control::
8498 * Warning Message Control::
8499 * Debugging and Assertion Control::
8500 * Validity Checking::
8503 * Using gcc for Syntax Checking::
8504 * Using gcc for Semantic Checking::
8505 * Compiling Different Versions of Ada::
8506 * Character Set Control::
8507 * File Naming Control::
8508 * Subprogram Inlining Control::
8509 * Auxiliary Output Control::
8510 * Debugging Control::
8511 * Exception Handling Control::
8512 * Units to Sources Mapping Files::
8513 * Code Generation Control::
8517 @node Alphabetical List of All Switches,Output and Error Message Control,,Compiler Switches
8518 @anchor{gnat_ugn/building_executable_programs_with_gnat alphabetical-list-of-all-switches}@anchor{e2}@anchor{gnat_ugn/building_executable_programs_with_gnat id13}@anchor{e3}
8519 @subsection Alphabetical List of All Switches
8527 @item @code{-b @emph{target}}
8529 Compile your program to run on @code{target}, which is the name of a
8530 system configuration. You must have a GNAT cross-compiler built if
8531 @code{target} is not the same as your host system.
8539 @item @code{-B@emph{dir}}
8541 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
8542 from @code{dir} instead of the default location. Only use this switch
8543 when multiple versions of the GNAT compiler are available.
8544 See the “Options for Directory Search” section in the
8545 @cite{Using the GNU Compiler Collection (GCC)} manual for further details.
8546 You would normally use the @code{-b} or @code{-V} switch instead.
8556 Compile. Always use this switch when compiling Ada programs.
8558 Note: for some other languages when using @code{gcc}, notably in
8559 the case of C and C++, it is possible to use
8560 use @code{gcc} without a @code{-c} switch to
8561 compile and link in one step. In the case of GNAT, you
8562 cannot use this approach, because the binder must be run
8563 and @code{gcc} cannot be used to run the GNAT binder.
8566 @geindex -fcallgraph-info (gcc)
8571 @item @code{-fcallgraph-info[=su,da]}
8573 Makes the compiler output callgraph information for the program, on a
8574 per-file basis. The information is generated in the VCG format. It can
8575 be decorated with additional, per-node and/or per-edge information, if a
8576 list of comma-separated markers is additionally specified. When the
8577 @code{su} marker is specified, the callgraph is decorated with stack usage
8578 information; it is equivalent to @code{-fstack-usage}. When the @code{da}
8579 marker is specified, the callgraph is decorated with information about
8580 dynamically allocated objects.
8583 @geindex -fdiagnostics-format (gcc)
8588 @item @code{-fdiagnostics-format=json}
8590 Makes GNAT emit warning and error messages as JSON. Inhibits printing of
8591 text warning and errors messages except if @code{-gnatv} or
8592 @code{-gnatl} are present.
8595 @geindex -fdump-scos (gcc)
8600 @item @code{-fdump-scos}
8602 Generates SCO (Source Coverage Obligation) information in the ALI file.
8603 This information is used by advanced coverage tools. See unit @code{SCOs}
8604 in the compiler sources for details in files @code{scos.ads} and
8608 @geindex -fgnat-encodings (gcc)
8613 @item @code{-fgnat-encodings=[all|gdb|minimal]}
8615 This switch controls the balance between GNAT encodings and standard DWARF
8616 emitted in the debug information.
8619 @geindex -flto (gcc)
8624 @item @code{-flto[=@emph{n}]}
8626 Enables Link Time Optimization. This switch must be used in conjunction
8627 with the @code{-Ox} switches (but not with the @code{-gnatn} switch
8628 since it is a full replacement for the latter) and instructs the compiler
8629 to defer most optimizations until the link stage. The advantage of this
8630 approach is that the compiler can do a whole-program analysis and choose
8631 the best interprocedural optimization strategy based on a complete view
8632 of the program, instead of a fragmentary view with the usual approach.
8633 This can also speed up the compilation of big programs and reduce the
8634 size of the executable, compared with a traditional per-unit compilation
8635 with inlining across units enabled by the @code{-gnatn} switch.
8636 The drawback of this approach is that it may require more memory and that
8637 the debugging information generated by -g with it might be hardly usable.
8638 The switch, as well as the accompanying @code{-Ox} switches, must be
8639 specified both for the compilation and the link phases.
8640 If the @code{n} parameter is specified, the optimization and final code
8641 generation at link time are executed using @code{n} parallel jobs by
8642 means of an installed @code{make} program.
8645 @geindex -fno-inline (gcc)
8650 @item @code{-fno-inline}
8652 Suppresses all inlining, unless requested with pragma @code{Inline_Always}. The
8653 effect is enforced regardless of other optimization or inlining switches.
8654 Note that inlining can also be suppressed on a finer-grained basis with
8655 pragma @code{No_Inline}.
8658 @geindex -fno-inline-functions (gcc)
8663 @item @code{-fno-inline-functions}
8665 Suppresses automatic inlining of subprograms, which is enabled
8666 if @code{-O3} is used.
8669 @geindex -fno-inline-small-functions (gcc)
8674 @item @code{-fno-inline-small-functions}
8676 Suppresses automatic inlining of small subprograms, which is enabled
8677 if @code{-O2} is used.
8680 @geindex -fno-inline-functions-called-once (gcc)
8685 @item @code{-fno-inline-functions-called-once}
8687 Suppresses inlining of subprograms local to the unit and called once
8688 from within it, which is enabled if @code{-O1} is used.
8691 @geindex -fno-ivopts (gcc)
8696 @item @code{-fno-ivopts}
8698 Suppresses high-level loop induction variable optimizations, which are
8699 enabled if @code{-O1} is used. These optimizations are generally
8700 profitable but, for some specific cases of loops with numerous uses
8701 of the iteration variable that follow a common pattern, they may end
8702 up destroying the regularity that could be exploited at a lower level
8703 and thus producing inferior code.
8706 @geindex -fno-strict-aliasing (gcc)
8711 @item @code{-fno-strict-aliasing}
8713 Causes the compiler to avoid assumptions regarding non-aliasing
8714 of objects of different types. See
8715 @ref{e4,,Optimization and Strict Aliasing} for details.
8718 @geindex -fno-strict-overflow (gcc)
8723 @item @code{-fno-strict-overflow}
8725 Causes the compiler to avoid assumptions regarding the rules of signed
8726 integer overflow. These rules specify that signed integer overflow will
8727 result in a Constraint_Error exception at run time and are enforced in
8728 default mode by the compiler, so this switch should not be necessary in
8729 normal operating mode. It might be useful in conjunction with @code{-gnato0}
8730 for very peculiar cases of low-level programming.
8733 @geindex -fstack-check (gcc)
8738 @item @code{-fstack-check}
8740 Activates stack checking.
8741 See @ref{e5,,Stack Overflow Checking} for details.
8744 @geindex -fstack-usage (gcc)
8749 @item @code{-fstack-usage}
8751 Makes the compiler output stack usage information for the program, on a
8752 per-subprogram basis. See @ref{e6,,Static Stack Usage Analysis} for details.
8762 Generate debugging information. This information is stored in the object
8763 file and copied from there to the final executable file by the linker,
8764 where it can be read by the debugger. You must use the
8765 @code{-g} switch if you plan on using the debugger.
8768 @geindex -gnat05 (gcc)
8773 @item @code{-gnat05}
8775 Allow full Ada 2005 features.
8778 @geindex -gnat12 (gcc)
8783 @item @code{-gnat12}
8785 Allow full Ada 2012 features.
8788 @geindex -gnat83 (gcc)
8790 @geindex -gnat2005 (gcc)
8795 @item @code{-gnat2005}
8797 Allow full Ada 2005 features (same as @code{-gnat05})
8800 @geindex -gnat2012 (gcc)
8805 @item @code{-gnat2012}
8807 Allow full Ada 2012 features (same as @code{-gnat12})
8810 @geindex -gnat2022 (gcc)
8815 @item @code{-gnat2022}
8817 Allow full Ada 2022 features
8819 @item @code{-gnat83}
8821 Enforce Ada 83 restrictions.
8824 @geindex -gnat95 (gcc)
8829 @item @code{-gnat95}
8831 Enforce Ada 95 restrictions.
8833 Note: for compatibility with some Ada 95 compilers which support only
8834 the @code{overriding} keyword of Ada 2005, the @code{-gnatd.D} switch can
8835 be used along with @code{-gnat95} to achieve a similar effect with GNAT.
8837 @code{-gnatd.D} instructs GNAT to consider @code{overriding} as a keyword
8838 and handle its associated semantic checks, even in Ada 95 mode.
8841 @geindex -gnata (gcc)
8848 Assertions enabled. @code{Pragma Assert} and @code{pragma Debug} to be
8849 activated. Note that these pragmas can also be controlled using the
8850 configuration pragmas @code{Assertion_Policy} and @code{Debug_Policy}.
8851 It also activates pragmas @code{Check}, @code{Precondition}, and
8852 @code{Postcondition}. Note that these pragmas can also be controlled
8853 using the configuration pragma @code{Check_Policy}. In Ada 2012, it
8854 also activates all assertions defined in the RM as aspects: preconditions,
8855 postconditions, type invariants and (sub)type predicates. In all Ada modes,
8856 corresponding pragmas for type invariants and (sub)type predicates are
8857 also activated. The default is that all these assertions are disabled,
8858 and have no effect, other than being checked for syntactic validity, and
8859 in the case of subtype predicates, constructions such as membership tests
8860 still test predicates even if assertions are turned off.
8863 @geindex -gnatA (gcc)
8870 Avoid processing @code{gnat.adc}. If a @code{gnat.adc} file is present,
8874 @geindex -gnatb (gcc)
8881 Generate brief messages to @code{stderr} even if verbose mode set.
8884 @geindex -gnatB (gcc)
8891 Assume no invalid (bad) values except for ‘Valid attribute use
8892 (@ref{e7,,Validity Checking}).
8895 @geindex -gnatc (gcc)
8902 Check syntax and semantics only (no code generation attempted). When the
8903 compiler is invoked by @code{gnatmake}, if the switch @code{-gnatc} is
8904 only given to the compiler (after @code{-cargs} or in package Compiler of
8905 the project file, @code{gnatmake} will fail because it will not find the
8906 object file after compilation. If @code{gnatmake} is called with
8907 @code{-gnatc} as a builder switch (before @code{-cargs} or in package
8908 Builder of the project file) then @code{gnatmake} will not fail because
8909 it will not look for the object files after compilation, and it will not try
8913 @geindex -gnatC (gcc)
8920 Generate CodePeer intermediate format (no code generation attempted).
8921 This switch will generate an intermediate representation suitable for
8922 use by CodePeer (@code{.scil} files). This switch is not compatible with
8923 code generation (it will, among other things, disable some switches such
8924 as -gnatn, and enable others such as -gnata).
8927 @geindex -gnatd (gcc)
8934 Specify debug options for the compiler. The string of characters after
8935 the @code{-gnatd} specifies the specific debug options. The possible
8936 characters are 0-9, a-z, A-Z, optionally preceded by a dot or underscore.
8937 See compiler source file @code{debug.adb} for details of the implemented
8938 debug options. Certain debug options are relevant to applications
8939 programmers, and these are documented at appropriate points in this
8943 @geindex -gnatD[nn] (gcc)
8950 Create expanded source files for source level debugging. This switch
8951 also suppresses generation of cross-reference information
8952 (see @code{-gnatx}). Note that this switch is not allowed if a previous
8953 -gnatR switch has been given, since these two switches are not compatible.
8956 @geindex -gnateA (gcc)
8961 @item @code{-gnateA}
8963 Check that the actual parameters of a subprogram call are not aliases of one
8964 another. To qualify as aliasing, the actuals must denote objects of a composite
8965 type, their memory locations must be identical or overlapping, and at least one
8966 of the corresponding formal parameters must be of mode OUT or IN OUT.
8969 type Rec_Typ is record
8970 Data : Integer := 0;
8973 function Self (Val : Rec_Typ) return Rec_Typ is
8978 procedure Detect_Aliasing (Val_1 : in out Rec_Typ; Val_2 : Rec_Typ) is
8981 end Detect_Aliasing;
8985 Detect_Aliasing (Obj, Obj);
8986 Detect_Aliasing (Obj, Self (Obj));
8989 In the example above, the first call to @code{Detect_Aliasing} fails with a
8990 @code{Program_Error} at run time because the actuals for @code{Val_1} and
8991 @code{Val_2} denote the same object. The second call executes without raising
8992 an exception because @code{Self(Obj)} produces an anonymous object which does
8993 not share the memory location of @code{Obj}.
8996 @geindex -gnateb (gcc)
9001 @item @code{-gnateb}
9003 Store configuration files by their basename in ALI files. This switch is
9004 used for instance by gprbuild for distributed builds in order to prevent
9005 issues where machine-specific absolute paths could end up being stored in
9009 @geindex -gnatec (gcc)
9014 @item @code{-gnatec=@emph{path}}
9016 Specify a configuration pragma file
9017 (the equal sign is optional)
9018 (@ref{63,,The Configuration Pragmas Files}).
9021 @geindex -gnateC (gcc)
9026 @item @code{-gnateC}
9028 Generate CodePeer messages in a compiler-like format. This switch is only
9029 effective if @code{-gnatcC} is also specified and requires an installation
9033 @geindex -gnated (gcc)
9038 @item @code{-gnated}
9040 Disable atomic synchronization
9043 @geindex -gnateD (gcc)
9048 @item @code{-gnateDsymbol[=@emph{value}]}
9050 Defines a symbol, associated with @code{value}, for preprocessing.
9051 (@ref{90,,Integrated Preprocessing}).
9054 @geindex -gnateE (gcc)
9059 @item @code{-gnateE}
9061 Generate extra information in exception messages. In particular, display
9062 extra column information and the value and range associated with index and
9063 range check failures, and extra column information for access checks.
9064 In cases where the compiler is able to determine at compile time that
9065 a check will fail, it gives a warning, and the extra information is not
9066 produced at run time.
9069 @geindex -gnatef (gcc)
9074 @item @code{-gnatef}
9076 Display full source path name in brief error messages.
9079 @geindex -gnateF (gcc)
9084 @item @code{-gnateF}
9086 Check for overflow on all floating-point operations, including those
9087 for unconstrained predefined types. See description of pragma
9088 @code{Check_Float_Overflow} in GNAT RM.
9091 @geindex -gnateg (gcc)
9098 The @code{-gnatc} switch must always be specified before this switch, e.g.
9099 @code{-gnatceg}. Generate a C header from the Ada input file. See
9100 @ref{b7,,Generating C Headers for Ada Specifications} for more
9104 @geindex -gnateG (gcc)
9109 @item @code{-gnateG}
9111 Save result of preprocessing in a text file.
9114 @geindex -gnatei (gcc)
9119 @item @code{-gnatei@emph{nnn}}
9121 Set maximum number of instantiations during compilation of a single unit to
9122 @code{nnn}. This may be useful in increasing the default maximum of 8000 for
9123 the rare case when a single unit legitimately exceeds this limit.
9126 @geindex -gnateI (gcc)
9131 @item @code{-gnateI@emph{nnn}}
9133 Indicates that the source is a multi-unit source and that the index of the
9134 unit to compile is @code{nnn}. @code{nnn} needs to be a positive number and need
9135 to be a valid index in the multi-unit source.
9138 @geindex -gnatel (gcc)
9143 @item @code{-gnatel}
9145 This switch can be used with the static elaboration model to issue info
9147 where implicit @code{pragma Elaborate} and @code{pragma Elaborate_All}
9148 are generated. This is useful in diagnosing elaboration circularities
9149 caused by these implicit pragmas when using the static elaboration
9150 model. See See the section in this guide on elaboration checking for
9151 further details. These messages are not generated by default, and are
9152 intended only for temporary use when debugging circularity problems.
9155 @geindex -gnatel (gcc)
9160 @item @code{-gnateL}
9162 This switch turns off the info messages about implicit elaboration pragmas.
9165 @geindex -gnatem (gcc)
9170 @item @code{-gnatem=@emph{path}}
9172 Specify a mapping file
9173 (the equal sign is optional)
9174 (@ref{e8,,Units to Sources Mapping Files}).
9177 @geindex -gnatep (gcc)
9182 @item @code{-gnatep=@emph{file}}
9184 Specify a preprocessing data file
9185 (the equal sign is optional)
9186 (@ref{90,,Integrated Preprocessing}).
9189 @geindex -gnateP (gcc)
9194 @item @code{-gnateP}
9196 Turn categorization dependency errors into warnings.
9197 Ada requires that units that WITH one another have compatible categories, for
9198 example a Pure unit cannot WITH a Preelaborate unit. If this switch is used,
9199 these errors become warnings (which can be ignored, or suppressed in the usual
9200 manner). This can be useful in some specialized circumstances such as the
9201 temporary use of special test software.
9204 @geindex -gnateS (gcc)
9209 @item @code{-gnateS}
9211 Synonym of @code{-fdump-scos}, kept for backwards compatibility.
9214 @geindex -gnatet=file (gcc)
9219 @item @code{-gnatet=@emph{path}}
9221 Generate target dependent information. The format of the output file is
9222 described in the section about switch @code{-gnateT}.
9225 @geindex -gnateT (gcc)
9230 @item @code{-gnateT=@emph{path}}
9232 Read target dependent information, such as endianness or sizes and alignments
9233 of base type. If this switch is passed, the default target dependent
9234 information of the compiler is replaced by the one read from the input file.
9235 This is used by tools other than the compiler, e.g. to do
9236 semantic analysis of programs that will run on some other target than
9237 the machine on which the tool is run.
9239 The following target dependent values should be defined,
9240 where @code{Nat} denotes a natural integer value, @code{Pos} denotes a
9241 positive integer value, and fields marked with a question mark are
9242 boolean fields, where a value of 0 is False, and a value of 1 is True:
9245 Bits_BE : Nat; -- Bits stored big-endian?
9246 Bits_Per_Unit : Pos; -- Bits in a storage unit
9247 Bits_Per_Word : Pos; -- Bits in a word
9248 Bytes_BE : Nat; -- Bytes stored big-endian?
9249 Char_Size : Pos; -- Standard.Character'Size
9250 Double_Float_Alignment : Nat; -- Alignment of double float
9251 Double_Scalar_Alignment : Nat; -- Alignment of double length scalar
9252 Double_Size : Pos; -- Standard.Long_Float'Size
9253 Float_Size : Pos; -- Standard.Float'Size
9254 Float_Words_BE : Nat; -- Float words stored big-endian?
9255 Int_Size : Pos; -- Standard.Integer'Size
9256 Long_Double_Size : Pos; -- Standard.Long_Long_Float'Size
9257 Long_Long_Size : Pos; -- Standard.Long_Long_Integer'Size
9258 Long_Size : Pos; -- Standard.Long_Integer'Size
9259 Maximum_Alignment : Pos; -- Maximum permitted alignment
9260 Max_Unaligned_Field : Pos; -- Maximum size for unaligned bit field
9261 Pointer_Size : Pos; -- System.Address'Size
9262 Short_Enums : Nat; -- Foreign enums use short size?
9263 Short_Size : Pos; -- Standard.Short_Integer'Size
9264 Strict_Alignment : Nat; -- Strict alignment?
9265 System_Allocator_Alignment : Nat; -- Alignment for malloc calls
9266 Wchar_T_Size : Pos; -- Interfaces.C.wchar_t'Size
9267 Words_BE : Nat; -- Words stored big-endian?
9270 @code{Bits_Per_Unit} is the number of bits in a storage unit, the equivalent of
9271 GCC macro @code{BITS_PER_UNIT} documented as follows: @cite{Define this macro to be the number of bits in an addressable storage unit (byte); normally 8.}
9273 @code{Bits_Per_Word} is the number of bits in a machine word, the equivalent of
9274 GCC macro @code{BITS_PER_WORD} documented as follows: @cite{Number of bits in a word; normally 32.}
9276 @code{Double_Float_Alignment}, if not zero, is the maximum alignment that the
9277 compiler can choose by default for a 64-bit floating-point type or object.
9279 @code{Double_Scalar_Alignment}, if not zero, is the maximum alignment that the
9280 compiler can choose by default for a 64-bit or larger scalar type or object.
9282 @code{Maximum_Alignment} is the maximum alignment that the compiler can choose
9283 by default for a type or object, which is also the maximum alignment that can
9284 be specified in GNAT. It is computed for GCC backends as @code{BIGGEST_ALIGNMENT
9285 / BITS_PER_UNIT} where GCC macro @code{BIGGEST_ALIGNMENT} is documented as
9286 follows: @cite{Biggest alignment that any data type can require on this machine@comma{} in bits.}
9288 @code{Max_Unaligned_Field} is the maximum size for unaligned bit field, which is
9289 64 for the majority of GCC targets (but can be different on some targets).
9291 @code{Strict_Alignment} is the equivalent of GCC macro @code{STRICT_ALIGNMENT}
9292 documented as follows: @cite{Define this macro to be the value 1 if instructions will fail to work if given data not on the nominal alignment. If instructions will merely go slower in that case@comma{} define this macro as 0.}
9294 @code{System_Allocator_Alignment} is the guaranteed alignment of data returned
9295 by calls to @code{malloc}.
9297 The format of the input file is as follows. First come the values of
9298 the variables defined above, with one line per value:
9304 where @code{name} is the name of the parameter, spelled out in full,
9305 and cased as in the above list, and @code{value} is an unsigned decimal
9306 integer. Two or more blanks separates the name from the value.
9308 All the variables must be present, in alphabetical order (i.e. the
9309 same order as the list above).
9311 Then there is a blank line to separate the two parts of the file. Then
9312 come the lines showing the floating-point types to be registered, with
9313 one line per registered mode:
9316 name digs float_rep size alignment
9319 where @code{name} is the string name of the type (which can have
9320 single spaces embedded in the name (e.g. long double), @code{digs} is
9321 the number of digits for the floating-point type, @code{float_rep} is
9322 the float representation (I for IEEE-754-Binary, which is
9323 the only one supported at this time),
9324 @code{size} is the size in bits, @code{alignment} is the
9325 alignment in bits. The name is followed by at least two blanks, fields
9326 are separated by at least one blank, and a LF character immediately
9327 follows the alignment field.
9329 Here is an example of a target parameterization file:
9337 Double_Float_Alignment 0
9338 Double_Scalar_Alignment 0
9343 Long_Double_Size 128
9346 Maximum_Alignment 16
9347 Max_Unaligned_Field 64
9351 System_Allocator_Alignment 16
9357 long double 18 I 80 128
9362 @geindex -gnateu (gcc)
9367 @item @code{-gnateu}
9369 Ignore unrecognized validity, warning, and style switches that
9370 appear after this switch is given. This may be useful when
9371 compiling sources developed on a later version of the compiler
9372 with an earlier version. Of course the earlier version must
9373 support this switch.
9376 @geindex -gnateV (gcc)
9381 @item @code{-gnateV}
9383 Check that all actual parameters of a subprogram call are valid according to
9384 the rules of validity checking (@ref{e7,,Validity Checking}).
9387 @geindex -gnateY (gcc)
9392 @item @code{-gnateY}
9394 Ignore all STYLE_CHECKS pragmas. Full legality checks
9395 are still carried out, but the pragmas have no effect
9396 on what style checks are active. This allows all style
9397 checking options to be controlled from the command line.
9400 @geindex -gnatE (gcc)
9407 Dynamic elaboration checking mode enabled. For further details see
9408 @ref{f,,Elaboration Order Handling in GNAT}.
9411 @geindex -gnatf (gcc)
9418 Full errors. Multiple errors per line, all undefined references, do not
9419 attempt to suppress cascaded errors.
9422 @geindex -gnatF (gcc)
9429 Externals names are folded to all uppercase.
9432 @geindex -gnatg (gcc)
9439 Internal GNAT implementation mode. This should not be used for applications
9440 programs, it is intended only for use by the compiler and its run-time
9441 library. For documentation, see the GNAT sources. Note that @code{-gnatg}
9442 implies @code{-gnatw.ge} and @code{-gnatyg} so that all standard
9443 warnings and all standard style options are turned on. All warnings and style
9444 messages are treated as errors.
9447 @geindex -gnatG[nn] (gcc)
9452 @item @code{-gnatG=nn}
9454 List generated expanded code in source form.
9457 @geindex -gnath (gcc)
9464 Output usage information. The output is written to @code{stdout}.
9467 @geindex -gnatH (gcc)
9474 Legacy elaboration-checking mode enabled. When this switch is in effect,
9475 the pre-18.x access-before-elaboration model becomes the de facto model.
9476 For further details see @ref{f,,Elaboration Order Handling in GNAT}.
9479 @geindex -gnati (gcc)
9484 @item @code{-gnati@emph{c}}
9486 Identifier character set (@code{c} = 1/2/3/4/5/9/p/8/f/n/w).
9487 For details of the possible selections for @code{c},
9488 see @ref{31,,Character Set Control}.
9491 @geindex -gnatI (gcc)
9498 Ignore representation clauses. When this switch is used,
9499 representation clauses are treated as comments. This is useful
9500 when initially porting code where you want to ignore rep clause
9501 problems, and also for compiling foreign code (particularly
9502 for use with ASIS). The representation clauses that are ignored
9503 are: enumeration_representation_clause, record_representation_clause,
9504 and attribute_definition_clause for the following attributes:
9505 Address, Alignment, Bit_Order, Component_Size, Machine_Radix,
9506 Object_Size, Scalar_Storage_Order, Size, Small, Stream_Size,
9507 and Value_Size. Pragma Default_Scalar_Storage_Order is also ignored.
9508 Note that this option should be used only for compiling – the
9509 code is likely to malfunction at run time.
9512 @geindex -gnatjnn (gcc)
9517 @item @code{-gnatj@emph{nn}}
9519 Reformat error messages to fit on @code{nn} character lines
9522 @geindex -gnatJ (gcc)
9529 Permissive elaboration-checking mode enabled. When this switch is in effect,
9530 the post-18.x access-before-elaboration model ignores potential issues with:
9539 Activations of tasks defined in instances
9545 Calls from within an instance to its enclosing context
9548 Calls through generic formal parameters
9551 Calls to subprograms defined in instances
9557 Indirect calls using ‘Access
9566 Synchronous task suspension
9569 and does not emit compile-time diagnostics or run-time checks. For further
9570 details see @ref{f,,Elaboration Order Handling in GNAT}.
9573 @geindex -gnatk (gcc)
9578 @item @code{-gnatk=@emph{n}}
9580 Limit file names to @code{n} (1-999) characters (@code{k} = krunch).
9583 @geindex -gnatl (gcc)
9590 Output full source listing with embedded error messages.
9593 @geindex -gnatL (gcc)
9600 Used in conjunction with -gnatG or -gnatD to intersperse original
9601 source lines (as comment lines with line numbers) in the expanded
9605 @geindex -gnatm (gcc)
9610 @item @code{-gnatm=@emph{n}}
9612 Limit number of detected error or warning messages to @code{n}
9613 where @code{n} is in the range 1..999999. The default setting if
9614 no switch is given is 9999. If the number of warnings reaches this
9615 limit, then a message is output and further warnings are suppressed,
9616 but the compilation is continued. If the number of error messages
9617 reaches this limit, then a message is output and the compilation
9618 is abandoned. The equal sign here is optional. A value of zero
9619 means that no limit applies.
9622 @geindex -gnatn (gcc)
9627 @item @code{-gnatn[12]}
9629 Activate inlining across units for subprograms for which pragma @code{Inline}
9630 is specified. This inlining is performed by the GCC back-end. An optional
9631 digit sets the inlining level: 1 for moderate inlining across units
9632 or 2 for full inlining across units. If no inlining level is specified,
9633 the compiler will pick it based on the optimization level.
9636 @geindex -gnatN (gcc)
9643 Activate front end inlining for subprograms for which
9644 pragma @code{Inline} is specified. This inlining is performed
9645 by the front end and will be visible in the
9646 @code{-gnatG} output.
9648 When using a gcc-based back end, then the use of
9649 @code{-gnatN} is deprecated, and the use of @code{-gnatn} is preferred.
9650 Historically front end inlining was more extensive than the gcc back end
9651 inlining, but that is no longer the case.
9654 @geindex -gnato0 (gcc)
9659 @item @code{-gnato0}
9661 Suppresses overflow checking. This causes the behavior of the compiler to
9662 match the default for older versions where overflow checking was suppressed
9663 by default. This is equivalent to having
9664 @code{pragma Suppress (Overflow_Check)} in a configuration pragma file.
9667 @geindex -gnato?? (gcc)
9672 @item @code{-gnato??}
9674 Set default mode for handling generation of code to avoid intermediate
9675 arithmetic overflow. Here @code{??} is two digits, a
9676 single digit, or nothing. Each digit is one of the digits @code{1}
9680 @multitable {xxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
9695 All intermediate overflows checked against base type (@code{STRICT})
9703 Minimize intermediate overflows (@code{MINIMIZED})
9711 Eliminate intermediate overflows (@code{ELIMINATED})
9716 If only one digit appears, then it applies to all
9717 cases; if two digits are given, then the first applies outside
9718 assertions, pre/postconditions, and type invariants, and the second
9719 applies within assertions, pre/postconditions, and type invariants.
9721 If no digits follow the @code{-gnato}, then it is equivalent to
9723 causing all intermediate overflows to be handled in strict
9726 This switch also causes arithmetic overflow checking to be performed
9727 (as though @code{pragma Unsuppress (Overflow_Check)} had been specified).
9729 The default if no option @code{-gnato} is given is that overflow handling
9730 is in @code{STRICT} mode (computations done using the base type), and that
9731 overflow checking is enabled.
9733 Note that division by zero is a separate check that is not
9734 controlled by this switch (divide-by-zero checking is on by default).
9736 See also @ref{e9,,Specifying the Desired Mode}.
9739 @geindex -gnatp (gcc)
9746 Suppress all checks. See @ref{ea,,Run-Time Checks} for details. This switch
9747 has no effect if cancelled by a subsequent @code{-gnat-p} switch.
9750 @geindex -gnat-p (gcc)
9755 @item @code{-gnat-p}
9757 Cancel effect of previous @code{-gnatp} switch.
9760 @geindex -gnatq (gcc)
9767 Don’t quit. Try semantics, even if parse errors.
9770 @geindex -gnatQ (gcc)
9777 Don’t quit. Generate @code{ALI} and tree files even if illegalities.
9778 Note that code generation is still suppressed in the presence of any
9779 errors, so even with @code{-gnatQ} no object file is generated.
9782 @geindex -gnatr (gcc)
9789 Treat pragma Restrictions as Restriction_Warnings.
9792 @geindex -gnatR (gcc)
9797 @item @code{-gnatR[0|1|2|3|4][e][j][m][s]}
9799 Output representation information for declared types, objects and
9800 subprograms. Note that this switch is not allowed if a previous
9801 @code{-gnatD} switch has been given, since these two switches
9805 @geindex -gnats (gcc)
9815 @geindex -gnatS (gcc)
9822 Print package Standard.
9825 @geindex -gnatT (gcc)
9830 @item @code{-gnatT@emph{nnn}}
9832 All compiler tables start at @code{nnn} times usual starting size.
9835 @geindex -gnatu (gcc)
9842 List units for this compilation.
9845 @geindex -gnatU (gcc)
9852 Tag all error messages with the unique string ‘error:’
9855 @geindex -gnatv (gcc)
9862 Verbose mode. Full error output with source lines to @code{stdout}.
9865 @geindex -gnatV (gcc)
9872 Control level of validity checking (@ref{e7,,Validity Checking}).
9875 @geindex -gnatw (gcc)
9880 @item @code{-gnatw@emph{xxx}}
9883 @code{xxx} is a string of option letters that denotes
9884 the exact warnings that
9885 are enabled or disabled (@ref{eb,,Warning Message Control}).
9888 @geindex -gnatW (gcc)
9893 @item @code{-gnatW@emph{e}}
9895 Wide character encoding method
9896 (@code{e}=n/h/u/s/e/8).
9899 @geindex -gnatx (gcc)
9906 Suppress generation of cross-reference information.
9909 @geindex -gnatX (gcc)
9916 Enable GNAT implementation extensions and latest Ada version.
9919 @geindex -gnaty (gcc)
9926 Enable built-in style checks (@ref{ec,,Style Checking}).
9929 @geindex -gnatz (gcc)
9934 @item @code{-gnatz@emph{m}}
9936 Distribution stub generation and compilation
9937 (@code{m}=r/c for receiver/caller stubs).
9945 @item @code{-I@emph{dir}}
9949 Direct GNAT to search the @code{dir} directory for source files needed by
9950 the current compilation
9951 (see @ref{73,,Search Paths and the Run-Time Library (RTL)}).
9963 Except for the source file named in the command line, do not look for source
9964 files in the directory containing the source file named in the command line
9965 (see @ref{73,,Search Paths and the Run-Time Library (RTL)}).
9973 @item @code{-o @emph{file}}
9975 This switch is used in @code{gcc} to redirect the generated object file
9976 and its associated ALI file. Beware of this switch with GNAT, because it may
9977 cause the object file and ALI file to have different names which in turn
9978 may confuse the binder and the linker.
9981 @geindex -nostdinc (gcc)
9986 @item @code{-nostdinc}
9988 Inhibit the search of the default location for the GNAT Run Time
9989 Library (RTL) source files.
9992 @geindex -nostdlib (gcc)
9997 @item @code{-nostdlib}
9999 Inhibit the search of the default location for the GNAT Run Time
10000 Library (RTL) ALI files.
10008 @item @code{-O[@emph{n}]}
10010 @code{n} controls the optimization level:
10013 @multitable {xxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
10028 No optimization, the default setting if no @code{-O} appears
10036 Normal optimization, the default if you specify @code{-O} without an
10037 operand. A good compromise between code quality and compilation
10046 Extensive optimization, may improve execution time, possibly at
10047 the cost of substantially increased compilation time.
10055 Same as @code{-O2}, and also includes inline expansion for small
10056 subprograms in the same unit.
10064 Optimize space usage
10069 See also @ref{ed,,Optimization Levels}.
10072 @geindex -pass-exit-codes (gcc)
10077 @item @code{-pass-exit-codes}
10079 Catch exit codes from the compiler and use the most meaningful as
10083 @geindex --RTS (gcc)
10088 @item @code{--RTS=@emph{rts-path}}
10090 Specifies the default location of the run-time library. Same meaning as the
10091 equivalent @code{gnatmake} flag (@ref{ce,,Switches for gnatmake}).
10101 Used in place of @code{-c} to
10102 cause the assembler source file to be
10103 generated, using @code{.s} as the extension,
10104 instead of the object file.
10105 This may be useful if you need to examine the generated assembly code.
10108 @geindex -fverbose-asm (gcc)
10113 @item @code{-fverbose-asm}
10115 Used in conjunction with @code{-S}
10116 to cause the generated assembly code file to be annotated with variable
10117 names, making it significantly easier to follow.
10127 Show commands generated by the @code{gcc} driver. Normally used only for
10128 debugging purposes or if you need to be sure what version of the
10129 compiler you are executing.
10137 @item @code{-V @emph{ver}}
10139 Execute @code{ver} version of the compiler. This is the @code{gcc}
10140 version, not the GNAT version.
10150 Turn off warnings generated by the back end of the compiler. Use of
10151 this switch also causes the default for front end warnings to be set
10152 to suppress (as though @code{-gnatws} had appeared at the start of
10156 @geindex Combining GNAT switches
10158 You may combine a sequence of GNAT switches into a single switch. For
10159 example, the combined switch
10168 is equivalent to specifying the following sequence of switches:
10173 -gnato -gnatf -gnati3
10177 The following restrictions apply to the combination of switches
10184 The switch @code{-gnatc} if combined with other switches must come
10185 first in the string.
10188 The switch @code{-gnats} if combined with other switches must come
10189 first in the string.
10193 @code{-gnatzc} and @code{-gnatzr} may not be combined with any other
10194 switches, and only one of them may appear in the command line.
10197 The switch @code{-gnat-p} may not be combined with any other switch.
10200 Once a ‘y’ appears in the string (that is a use of the @code{-gnaty}
10201 switch), then all further characters in the switch are interpreted
10202 as style modifiers (see description of @code{-gnaty}).
10205 Once a ‘d’ appears in the string (that is a use of the @code{-gnatd}
10206 switch), then all further characters in the switch are interpreted
10207 as debug flags (see description of @code{-gnatd}).
10210 Once a ‘w’ appears in the string (that is a use of the @code{-gnatw}
10211 switch), then all further characters in the switch are interpreted
10212 as warning mode modifiers (see description of @code{-gnatw}).
10215 Once a ‘V’ appears in the string (that is a use of the @code{-gnatV}
10216 switch), then all further characters in the switch are interpreted
10217 as validity checking options (@ref{e7,,Validity Checking}).
10220 Option ‘em’, ‘ec’, ‘ep’, ‘l=’ and ‘R’ must be the last options in
10221 a combined list of options.
10224 @node Output and Error Message Control,Warning Message Control,Alphabetical List of All Switches,Compiler Switches
10225 @anchor{gnat_ugn/building_executable_programs_with_gnat id14}@anchor{ee}@anchor{gnat_ugn/building_executable_programs_with_gnat output-and-error-message-control}@anchor{ef}
10226 @subsection Output and Error Message Control
10231 The standard default format for error messages is called ‘brief format’.
10232 Brief format messages are written to @code{stderr} (the standard error
10233 file) and have the following form:
10236 e.adb:3:04: Incorrect spelling of keyword "function"
10237 e.adb:4:20: ";" should be "is"
10240 The first integer after the file name is the line number in the file,
10241 and the second integer is the column number within the line.
10242 @code{GNAT Studio} can parse the error messages
10243 and point to the referenced character.
10244 The following switches provide control over the error message
10247 @geindex -gnatv (gcc)
10252 @item @code{-gnatv}
10254 The @code{v} stands for verbose.
10255 The effect of this setting is to write long-format error
10256 messages to @code{stdout} (the standard output file.
10257 The same program compiled with the
10258 @code{-gnatv} switch would generate:
10261 3. funcion X (Q : Integer)
10263 >>> Incorrect spelling of keyword "function"
10266 >>> ";" should be "is"
10269 The vertical bar indicates the location of the error, and the @code{>>>}
10270 prefix can be used to search for error messages. When this switch is
10271 used the only source lines output are those with errors.
10274 @geindex -gnatl (gcc)
10279 @item @code{-gnatl}
10281 The @code{l} stands for list.
10282 This switch causes a full listing of
10283 the file to be generated. In the case where a body is
10284 compiled, the corresponding spec is also listed, along
10285 with any subunits. Typical output from compiling a package
10286 body @code{p.adb} might look like:
10291 1. package body p is
10293 3. procedure a is separate;
10304 2. pragma Elaborate_Body
10325 When you specify the @code{-gnatv} or @code{-gnatl} switches and
10326 standard output is redirected, a brief summary is written to
10327 @code{stderr} (standard error) giving the number of error messages and
10328 warning messages generated.
10331 @geindex -gnatl=fname (gcc)
10336 @item @code{-gnatl=@emph{fname}}
10338 This has the same effect as @code{-gnatl} except that the output is
10339 written to a file instead of to standard output. If the given name
10340 @code{fname} does not start with a period, then it is the full name
10341 of the file to be written. If @code{fname} is an extension, it is
10342 appended to the name of the file being compiled. For example, if
10343 file @code{xyz.adb} is compiled with @code{-gnatl=.lst},
10344 then the output is written to file xyz.adb.lst.
10347 @geindex -gnatU (gcc)
10352 @item @code{-gnatU}
10354 This switch forces all error messages to be preceded by the unique
10355 string ‘error:’. This means that error messages take a few more
10356 characters in space, but allows easy searching for and identification
10360 @geindex -gnatb (gcc)
10365 @item @code{-gnatb}
10367 The @code{b} stands for brief.
10368 This switch causes GNAT to generate the
10369 brief format error messages to @code{stderr} (the standard error
10370 file) as well as the verbose
10371 format message or full listing (which as usual is written to
10372 @code{stdout} (the standard output file).
10375 @geindex -gnatm (gcc)
10380 @item @code{-gnatm=@emph{n}}
10382 The @code{m} stands for maximum.
10383 @code{n} is a decimal integer in the
10384 range of 1 to 999999 and limits the number of error or warning
10385 messages to be generated. For example, using
10386 @code{-gnatm2} might yield
10389 e.adb:3:04: Incorrect spelling of keyword "function"
10390 e.adb:5:35: missing ".."
10391 fatal error: maximum number of errors detected
10392 compilation abandoned
10395 The default setting if
10396 no switch is given is 9999. If the number of warnings reaches this
10397 limit, then a message is output and further warnings are suppressed,
10398 but the compilation is continued. If the number of error messages
10399 reaches this limit, then a message is output and the compilation
10400 is abandoned. A value of zero means that no limit applies.
10402 Note that the equal sign is optional, so the switches
10403 @code{-gnatm2} and @code{-gnatm=2} are equivalent.
10406 @geindex -gnatf (gcc)
10411 @item @code{-gnatf}
10413 @geindex Error messages
10414 @geindex suppressing
10416 The @code{f} stands for full.
10417 Normally, the compiler suppresses error messages that are likely to be
10418 redundant. This switch causes all error
10419 messages to be generated. In particular, in the case of
10420 references to undefined variables. If a given variable is referenced
10421 several times, the normal format of messages is
10424 e.adb:7:07: "V" is undefined (more references follow)
10427 where the parenthetical comment warns that there are additional
10428 references to the variable @code{V}. Compiling the same program with the
10429 @code{-gnatf} switch yields
10432 e.adb:7:07: "V" is undefined
10433 e.adb:8:07: "V" is undefined
10434 e.adb:8:12: "V" is undefined
10435 e.adb:8:16: "V" is undefined
10436 e.adb:9:07: "V" is undefined
10437 e.adb:9:12: "V" is undefined
10440 The @code{-gnatf} switch also generates additional information for
10441 some error messages. Some examples are:
10447 Details on possibly non-portable unchecked conversion
10450 List possible interpretations for ambiguous calls
10453 Additional details on incorrect parameters
10457 @geindex -gnatjnn (gcc)
10462 @item @code{-gnatjnn}
10464 In normal operation mode (or if @code{-gnatj0} is used), then error messages
10465 with continuation lines are treated as though the continuation lines were
10466 separate messages (and so a warning with two continuation lines counts as
10467 three warnings, and is listed as three separate messages).
10469 If the @code{-gnatjnn} switch is used with a positive value for nn, then
10470 messages are output in a different manner. A message and all its continuation
10471 lines are treated as a unit, and count as only one warning or message in the
10472 statistics totals. Furthermore, the message is reformatted so that no line
10473 is longer than nn characters.
10476 @geindex -gnatq (gcc)
10481 @item @code{-gnatq}
10483 The @code{q} stands for quit (really ‘don’t quit’).
10484 In normal operation mode, the compiler first parses the program and
10485 determines if there are any syntax errors. If there are, appropriate
10486 error messages are generated and compilation is immediately terminated.
10488 GNAT to continue with semantic analysis even if syntax errors have been
10489 found. This may enable the detection of more errors in a single run. On
10490 the other hand, the semantic analyzer is more likely to encounter some
10491 internal fatal error when given a syntactically invalid tree.
10494 @geindex -gnatQ (gcc)
10499 @item @code{-gnatQ}
10501 In normal operation mode, the @code{ALI} file is not generated if any
10502 illegalities are detected in the program. The use of @code{-gnatQ} forces
10503 generation of the @code{ALI} file. This file is marked as being in
10504 error, so it cannot be used for binding purposes, but it does contain
10505 reasonably complete cross-reference information, and thus may be useful
10506 for use by tools (e.g., semantic browsing tools or integrated development
10507 environments) that are driven from the @code{ALI} file. This switch
10508 implies @code{-gnatq}, since the semantic phase must be run to get a
10509 meaningful ALI file.
10511 When @code{-gnatQ} is used and the generated @code{ALI} file is marked as
10512 being in error, @code{gnatmake} will attempt to recompile the source when it
10513 finds such an @code{ALI} file, including with switch @code{-gnatc}.
10515 Note that @code{-gnatQ} has no effect if @code{-gnats} is specified,
10516 since ALI files are never generated if @code{-gnats} is set.
10519 @node Warning Message Control,Debugging and Assertion Control,Output and Error Message Control,Compiler Switches
10520 @anchor{gnat_ugn/building_executable_programs_with_gnat id15}@anchor{f0}@anchor{gnat_ugn/building_executable_programs_with_gnat warning-message-control}@anchor{eb}
10521 @subsection Warning Message Control
10524 @geindex Warning messages
10526 In addition to error messages, which correspond to illegalities as defined
10527 in the Ada Reference Manual, the compiler detects two kinds of warning
10530 First, the compiler considers some constructs suspicious and generates a
10531 warning message to alert you to a possible error. Second, if the
10532 compiler detects a situation that is sure to raise an exception at
10533 run time, it generates a warning message. The following shows an example
10534 of warning messages:
10537 e.adb:4:24: warning: creation of object may raise Storage_Error
10538 e.adb:10:17: warning: static value out of range
10539 e.adb:10:17: warning: "Constraint_Error" will be raised at run time
10542 GNAT considers a large number of situations as appropriate
10543 for the generation of warning messages. As always, warnings are not
10544 definite indications of errors. For example, if you do an out-of-range
10545 assignment with the deliberate intention of raising a
10546 @code{Constraint_Error} exception, then the warning that may be
10547 issued does not indicate an error. Some of the situations for which GNAT
10548 issues warnings (at least some of the time) are given in the following
10549 list. This list is not complete, and new warnings are often added to
10550 subsequent versions of GNAT. The list is intended to give a general idea
10551 of the kinds of warnings that are generated.
10557 Possible infinitely recursive calls
10560 Out-of-range values being assigned
10563 Possible order of elaboration problems
10566 Size not a multiple of alignment for a record type
10569 Assertions (pragma Assert) that are sure to fail
10575 Address clauses with possibly unaligned values, or where an attempt is
10576 made to overlay a smaller variable with a larger one.
10579 Fixed-point type declarations with a null range
10582 Direct_IO or Sequential_IO instantiated with a type that has access values
10585 Variables that are never assigned a value
10588 Variables that are referenced before being initialized
10591 Task entries with no corresponding @code{accept} statement
10594 Duplicate accepts for the same task entry in a @code{select}
10597 Objects that take too much storage
10600 Unchecked conversion between types of differing sizes
10603 Missing @code{return} statement along some execution path in a function
10606 Incorrect (unrecognized) pragmas
10609 Incorrect external names
10612 Allocation from empty storage pool
10615 Potentially blocking operation in protected type
10618 Suspicious parenthesization of expressions
10621 Mismatching bounds in an aggregate
10624 Attempt to return local value by reference
10627 Premature instantiation of a generic body
10630 Attempt to pack aliased components
10633 Out of bounds array subscripts
10636 Wrong length on string assignment
10639 Violations of style rules if style checking is enabled
10642 Unused @emph{with} clauses
10645 @code{Bit_Order} usage that does not have any effect
10648 @code{Standard.Duration} used to resolve universal fixed expression
10651 Dereference of possibly null value
10654 Declaration that is likely to cause storage error
10657 Internal GNAT unit @emph{with}ed by application unit
10660 Values known to be out of range at compile time
10663 Unreferenced or unmodified variables. Note that a special
10664 exemption applies to variables which contain any of the substrings
10665 @code{DISCARD, DUMMY, IGNORE, JUNK, UNUSED}, in any casing. Such variables
10666 are considered likely to be intentionally used in a situation where
10667 otherwise a warning would be given, so warnings of this kind are
10668 always suppressed for such variables.
10671 Address overlays that could clobber memory
10674 Unexpected initialization when address clause present
10677 Bad alignment for address clause
10680 Useless type conversions
10683 Redundant assignment statements and other redundant constructs
10686 Useless exception handlers
10689 Accidental hiding of name by child unit
10692 Access before elaboration detected at compile time
10695 A range in a @code{for} loop that is known to be null or might be null
10698 The following section lists compiler switches that are available
10699 to control the handling of warning messages. It is also possible
10700 to exercise much finer control over what warnings are issued and
10701 suppressed using the GNAT pragma Warnings (see the description
10702 of the pragma in the @cite{GNAT_Reference_manual}).
10704 @geindex -gnatwa (gcc)
10709 @item @code{-gnatwa}
10711 @emph{Activate most optional warnings.}
10713 This switch activates most optional warning messages. See the remaining list
10714 in this section for details on optional warning messages that can be
10715 individually controlled. The warnings that are not turned on by this
10722 @code{-gnatwd} (implicit dereferencing)
10725 @code{-gnatw.d} (tag warnings with -gnatw switch)
10728 @code{-gnatwh} (hiding)
10731 @code{-gnatw.h} (holes in record layouts)
10734 @code{-gnatw.j} (late primitives of tagged types)
10737 @code{-gnatw.k} (redefinition of names in standard)
10740 @code{-gnatwl} (elaboration warnings)
10743 @code{-gnatw.l} (inherited aspects)
10746 @code{-gnatw.n} (atomic synchronization)
10749 @code{-gnatwo} (address clause overlay)
10752 @code{-gnatw.o} (values set by out parameters ignored)
10755 @code{-gnatw.q} (questionable layout of record types)
10758 @code{-gnatw_r} (out-of-order record representation clauses)
10761 @code{-gnatw.s} (overridden size clause)
10764 @code{-gnatwt} (tracking of deleted conditional code)
10767 @code{-gnatw.u} (unordered enumeration)
10770 @code{-gnatw.w} (use of Warnings Off)
10773 @code{-gnatw.y} (reasons for package needing body)
10776 All other optional warnings are turned on.
10779 @geindex -gnatwA (gcc)
10784 @item @code{-gnatwA}
10786 @emph{Suppress all optional errors.}
10788 This switch suppresses all optional warning messages, see remaining list
10789 in this section for details on optional warning messages that can be
10790 individually controlled. Note that unlike switch @code{-gnatws}, the
10791 use of switch @code{-gnatwA} does not suppress warnings that are
10792 normally given unconditionally and cannot be individually controlled
10793 (for example, the warning about a missing exit path in a function).
10794 Also, again unlike switch @code{-gnatws}, warnings suppressed by
10795 the use of switch @code{-gnatwA} can be individually turned back
10796 on. For example the use of switch @code{-gnatwA} followed by
10797 switch @code{-gnatwd} will suppress all optional warnings except
10798 the warnings for implicit dereferencing.
10801 @geindex -gnatw.a (gcc)
10806 @item @code{-gnatw.a}
10808 @emph{Activate warnings on failing assertions.}
10810 @geindex Assert failures
10812 This switch activates warnings for assertions where the compiler can tell at
10813 compile time that the assertion will fail. Note that this warning is given
10814 even if assertions are disabled. The default is that such warnings are
10818 @geindex -gnatw.A (gcc)
10823 @item @code{-gnatw.A}
10825 @emph{Suppress warnings on failing assertions.}
10827 @geindex Assert failures
10829 This switch suppresses warnings for assertions where the compiler can tell at
10830 compile time that the assertion will fail.
10838 @item @code{-gnatw_a}
10840 @emph{Activate warnings on anonymous allocators.}
10842 @geindex Anonymous allocators
10844 This switch activates warnings for allocators of anonymous access types,
10845 which can involve run-time accessibility checks and lead to unexpected
10846 accessibility violations. For more details on the rules involved, see
10855 @item @code{-gnatw_A}
10857 @emph{Supress warnings on anonymous allocators.}
10859 @geindex Anonymous allocators
10861 This switch suppresses warnings for anonymous access type allocators.
10864 @geindex -gnatwb (gcc)
10869 @item @code{-gnatwb}
10871 @emph{Activate warnings on bad fixed values.}
10873 @geindex Bad fixed values
10875 @geindex Fixed-point Small value
10877 @geindex Small value
10879 This switch activates warnings for static fixed-point expressions whose
10880 value is not an exact multiple of Small. Such values are implementation
10881 dependent, since an implementation is free to choose either of the multiples
10882 that surround the value. GNAT always chooses the closer one, but this is not
10883 required behavior, and it is better to specify a value that is an exact
10884 multiple, ensuring predictable execution. The default is that such warnings
10888 @geindex -gnatwB (gcc)
10893 @item @code{-gnatwB}
10895 @emph{Suppress warnings on bad fixed values.}
10897 This switch suppresses warnings for static fixed-point expressions whose
10898 value is not an exact multiple of Small.
10901 @geindex -gnatw.b (gcc)
10906 @item @code{-gnatw.b}
10908 @emph{Activate warnings on biased representation.}
10910 @geindex Biased representation
10912 This switch activates warnings when a size clause, value size clause, component
10913 clause, or component size clause forces the use of biased representation for an
10914 integer type (e.g. representing a range of 10..11 in a single bit by using 0/1
10915 to represent 10/11). The default is that such warnings are generated.
10918 @geindex -gnatwB (gcc)
10923 @item @code{-gnatw.B}
10925 @emph{Suppress warnings on biased representation.}
10927 This switch suppresses warnings for representation clauses that force the use
10928 of biased representation.
10931 @geindex -gnatwc (gcc)
10936 @item @code{-gnatwc}
10938 @emph{Activate warnings on conditionals.}
10940 @geindex Conditionals
10943 This switch activates warnings for conditional expressions used in
10944 tests that are known to be True or False at compile time. The default
10945 is that such warnings are not generated.
10946 Note that this warning does
10947 not get issued for the use of boolean variables or constants whose
10948 values are known at compile time, since this is a standard technique
10949 for conditional compilation in Ada, and this would generate too many
10950 false positive warnings.
10952 This warning option also activates a special test for comparisons using
10953 the operators ‘>=’ and’ <=’.
10954 If the compiler can tell that only the equality condition is possible,
10955 then it will warn that the ‘>’ or ‘<’ part of the test
10956 is useless and that the operator could be replaced by ‘=’.
10957 An example would be comparing a @code{Natural} variable <= 0.
10959 This warning option also generates warnings if
10960 one or both tests is optimized away in a membership test for integer
10961 values if the result can be determined at compile time. Range tests on
10962 enumeration types are not included, since it is common for such tests
10963 to include an end point.
10965 This warning can also be turned on using @code{-gnatwa}.
10968 @geindex -gnatwC (gcc)
10973 @item @code{-gnatwC}
10975 @emph{Suppress warnings on conditionals.}
10977 This switch suppresses warnings for conditional expressions used in
10978 tests that are known to be True or False at compile time.
10981 @geindex -gnatw.c (gcc)
10986 @item @code{-gnatw.c}
10988 @emph{Activate warnings on missing component clauses.}
10990 @geindex Component clause
10993 This switch activates warnings for record components where a record
10994 representation clause is present and has component clauses for the
10995 majority, but not all, of the components. A warning is given for each
10996 component for which no component clause is present.
10999 @geindex -gnatw.C (gcc)
11004 @item @code{-gnatw.C}
11006 @emph{Suppress warnings on missing component clauses.}
11008 This switch suppresses warnings for record components that are
11009 missing a component clause in the situation described above.
11012 @geindex -gnatw_c (gcc)
11017 @item @code{-gnatw_c}
11019 @emph{Activate warnings on unknown condition in Compile_Time_Warning.}
11021 @geindex Compile_Time_Warning
11023 @geindex Compile_Time_Error
11025 This switch activates warnings on a pragma Compile_Time_Warning
11026 or Compile_Time_Error whose condition has a value that is not
11027 known at compile time.
11028 The default is that such warnings are generated.
11031 @geindex -gnatw_C (gcc)
11036 @item @code{-gnatw_C}
11038 @emph{Suppress warnings on unknown condition in Compile_Time_Warning.}
11040 This switch supresses warnings on a pragma Compile_Time_Warning
11041 or Compile_Time_Error whose condition has a value that is not
11042 known at compile time.
11045 @geindex -gnatwd (gcc)
11050 @item @code{-gnatwd}
11052 @emph{Activate warnings on implicit dereferencing.}
11054 If this switch is set, then the use of a prefix of an access type
11055 in an indexed component, slice, or selected component without an
11056 explicit @code{.all} will generate a warning. With this warning
11057 enabled, access checks occur only at points where an explicit
11058 @code{.all} appears in the source code (assuming no warnings are
11059 generated as a result of this switch). The default is that such
11060 warnings are not generated.
11063 @geindex -gnatwD (gcc)
11068 @item @code{-gnatwD}
11070 @emph{Suppress warnings on implicit dereferencing.}
11072 @geindex Implicit dereferencing
11074 @geindex Dereferencing
11077 This switch suppresses warnings for implicit dereferences in
11078 indexed components, slices, and selected components.
11081 @geindex -gnatw.d (gcc)
11086 @item @code{-gnatw.d}
11088 @emph{Activate tagging of warning and info messages.}
11090 If this switch is set, then warning messages are tagged, with one of the
11100 Used to tag warnings controlled by the switch @code{-gnatwx} where x
11105 Used to tag warnings controlled by the switch @code{-gnatw.x} where x
11110 Used to tag elaboration information (info) messages generated when the
11111 static model of elaboration is used and the @code{-gnatel} switch is set.
11114 @emph{[restriction warning]}
11115 Used to tag warning messages for restriction violations, activated by use
11116 of the pragma @code{Restriction_Warnings}.
11119 @emph{[warning-as-error]}
11120 Used to tag warning messages that have been converted to error messages by
11121 use of the pragma Warning_As_Error. Note that such warnings are prefixed by
11122 the string “error: ” rather than “warning: “.
11125 @emph{[enabled by default]}
11126 Used to tag all other warnings that are always given by default, unless
11127 warnings are completely suppressed using pragma @emph{Warnings(Off)} or
11128 the switch @code{-gnatws}.
11133 @geindex -gnatw.d (gcc)
11138 @item @code{-gnatw.D}
11140 @emph{Deactivate tagging of warning and info messages messages.}
11142 If this switch is set, then warning messages return to the default
11143 mode in which warnings and info messages are not tagged as described above for
11147 @geindex -gnatwe (gcc)
11150 @geindex treat as error
11155 @item @code{-gnatwe}
11157 @emph{Treat warnings and style checks as errors.}
11159 This switch causes warning messages and style check messages to be
11161 The warning string still appears, but the warning messages are counted
11162 as errors, and prevent the generation of an object file. Note that this
11163 is the only -gnatw switch that affects the handling of style check messages.
11164 Note also that this switch has no effect on info (information) messages, which
11165 are not treated as errors if this switch is present.
11168 @geindex -gnatw.e (gcc)
11173 @item @code{-gnatw.e}
11175 @emph{Activate every optional warning.}
11178 @geindex activate every optional warning
11180 This switch activates all optional warnings, including those which
11181 are not activated by @code{-gnatwa}. The use of this switch is not
11182 recommended for normal use. If you turn this switch on, it is almost
11183 certain that you will get large numbers of useless warnings. The
11184 warnings that are excluded from @code{-gnatwa} are typically highly
11185 specialized warnings that are suitable for use only in code that has
11186 been specifically designed according to specialized coding rules.
11189 @geindex -gnatwE (gcc)
11192 @geindex treat as error
11197 @item @code{-gnatwE}
11199 @emph{Treat all run-time exception warnings as errors.}
11201 This switch causes warning messages regarding errors that will be raised
11202 during run-time execution to be treated as errors.
11205 @geindex -gnatwf (gcc)
11210 @item @code{-gnatwf}
11212 @emph{Activate warnings on unreferenced formals.}
11215 @geindex unreferenced
11217 This switch causes a warning to be generated if a formal parameter
11218 is not referenced in the body of the subprogram. This warning can
11219 also be turned on using @code{-gnatwu}. The
11220 default is that these warnings are not generated.
11223 @geindex -gnatwF (gcc)
11228 @item @code{-gnatwF}
11230 @emph{Suppress warnings on unreferenced formals.}
11232 This switch suppresses warnings for unreferenced formal
11233 parameters. Note that the
11234 combination @code{-gnatwu} followed by @code{-gnatwF} has the
11235 effect of warning on unreferenced entities other than subprogram
11239 @geindex -gnatwg (gcc)
11244 @item @code{-gnatwg}
11246 @emph{Activate warnings on unrecognized pragmas.}
11249 @geindex unrecognized
11251 This switch causes a warning to be generated if an unrecognized
11252 pragma is encountered. Apart from issuing this warning, the
11253 pragma is ignored and has no effect. The default
11254 is that such warnings are issued (satisfying the Ada Reference
11255 Manual requirement that such warnings appear).
11258 @geindex -gnatwG (gcc)
11263 @item @code{-gnatwG}
11265 @emph{Suppress warnings on unrecognized pragmas.}
11267 This switch suppresses warnings for unrecognized pragmas.
11270 @geindex -gnatw.g (gcc)
11275 @item @code{-gnatw.g}
11277 @emph{Warnings used for GNAT sources.}
11279 This switch sets the warning categories that are used by the standard
11280 GNAT style. Currently this is equivalent to
11281 @code{-gnatwAao.q.s.CI.V.X.Z}
11282 but more warnings may be added in the future without advanced notice.
11285 @geindex -gnatwh (gcc)
11290 @item @code{-gnatwh}
11292 @emph{Activate warnings on hiding.}
11294 @geindex Hiding of Declarations
11296 This switch activates warnings on hiding declarations that are considered
11297 potentially confusing. Not all cases of hiding cause warnings; for example an
11298 overriding declaration hides an implicit declaration, which is just normal
11299 code. The default is that warnings on hiding are not generated.
11302 @geindex -gnatwH (gcc)
11307 @item @code{-gnatwH}
11309 @emph{Suppress warnings on hiding.}
11311 This switch suppresses warnings on hiding declarations.
11314 @geindex -gnatw.h (gcc)
11319 @item @code{-gnatw.h}
11321 @emph{Activate warnings on holes/gaps in records.}
11323 @geindex Record Representation (gaps)
11325 This switch activates warnings on component clauses in record
11326 representation clauses that leave holes (gaps) in the record layout.
11327 If this warning option is active, then record representation clauses
11328 should specify a contiguous layout, adding unused fill fields if needed.
11331 @geindex -gnatw.H (gcc)
11336 @item @code{-gnatw.H}
11338 @emph{Suppress warnings on holes/gaps in records.}
11340 This switch suppresses warnings on component clauses in record
11341 representation clauses that leave holes (haps) in the record layout.
11344 @geindex -gnatwi (gcc)
11349 @item @code{-gnatwi}
11351 @emph{Activate warnings on implementation units.}
11353 This switch activates warnings for a @emph{with} of an internal GNAT
11354 implementation unit, defined as any unit from the @code{Ada},
11355 @code{Interfaces}, @code{GNAT},
11357 hierarchies that is not
11358 documented in either the Ada Reference Manual or the GNAT
11359 Programmer’s Reference Manual. Such units are intended only
11360 for internal implementation purposes and should not be @emph{with}ed
11361 by user programs. The default is that such warnings are generated
11364 @geindex -gnatwI (gcc)
11369 @item @code{-gnatwI}
11371 @emph{Disable warnings on implementation units.}
11373 This switch disables warnings for a @emph{with} of an internal GNAT
11374 implementation unit.
11377 @geindex -gnatw.i (gcc)
11382 @item @code{-gnatw.i}
11384 @emph{Activate warnings on overlapping actuals.}
11386 This switch enables a warning on statically detectable overlapping actuals in
11387 a subprogram call, when one of the actuals is an in-out parameter, and the
11388 types of the actuals are not by-copy types. This warning is off by default.
11391 @geindex -gnatw.I (gcc)
11396 @item @code{-gnatw.I}
11398 @emph{Disable warnings on overlapping actuals.}
11400 This switch disables warnings on overlapping actuals in a call.
11403 @geindex -gnatwj (gcc)
11408 @item @code{-gnatwj}
11410 @emph{Activate warnings on obsolescent features (Annex J).}
11413 @geindex obsolescent
11415 @geindex Obsolescent features
11417 If this warning option is activated, then warnings are generated for
11418 calls to subprograms marked with @code{pragma Obsolescent} and
11419 for use of features in Annex J of the Ada Reference Manual. In the
11420 case of Annex J, not all features are flagged. In particular use
11421 of the renamed packages (like @code{Text_IO}) and use of package
11422 @code{ASCII} are not flagged, since these are very common and
11423 would generate many annoying positive warnings. The default is that
11424 such warnings are not generated.
11426 In addition to the above cases, warnings are also generated for
11427 GNAT features that have been provided in past versions but which
11428 have been superseded (typically by features in the new Ada standard).
11429 For example, @code{pragma Ravenscar} will be flagged since its
11430 function is replaced by @code{pragma Profile(Ravenscar)}, and
11431 @code{pragma Interface_Name} will be flagged since its function
11432 is replaced by @code{pragma Import}.
11434 Note that this warning option functions differently from the
11435 restriction @code{No_Obsolescent_Features} in two respects.
11436 First, the restriction applies only to annex J features.
11437 Second, the restriction does flag uses of package @code{ASCII}.
11440 @geindex -gnatwJ (gcc)
11445 @item @code{-gnatwJ}
11447 @emph{Suppress warnings on obsolescent features (Annex J).}
11449 This switch disables warnings on use of obsolescent features.
11452 @geindex -gnatw.j (gcc)
11457 @item @code{-gnatw.j}
11459 @emph{Activate warnings on late declarations of tagged type primitives.}
11461 This switch activates warnings on visible primitives added to a
11462 tagged type after deriving a private extension from it.
11465 @geindex -gnatw.J (gcc)
11470 @item @code{-gnatw.J}
11472 @emph{Suppress warnings on late declarations of tagged type primitives.}
11474 This switch suppresses warnings on visible primitives added to a
11475 tagged type after deriving a private extension from it.
11478 @geindex -gnatwk (gcc)
11483 @item @code{-gnatwk}
11485 @emph{Activate warnings on variables that could be constants.}
11487 This switch activates warnings for variables that are initialized but
11488 never modified, and then could be declared constants. The default is that
11489 such warnings are not given.
11492 @geindex -gnatwK (gcc)
11497 @item @code{-gnatwK}
11499 @emph{Suppress warnings on variables that could be constants.}
11501 This switch disables warnings on variables that could be declared constants.
11504 @geindex -gnatw.k (gcc)
11509 @item @code{-gnatw.k}
11511 @emph{Activate warnings on redefinition of names in standard.}
11513 This switch activates warnings for declarations that declare a name that
11514 is defined in package Standard. Such declarations can be confusing,
11515 especially since the names in package Standard continue to be directly
11516 visible, meaning that use visibiliy on such redeclared names does not
11517 work as expected. Names of discriminants and components in records are
11518 not included in this check.
11521 @geindex -gnatwK (gcc)
11526 @item @code{-gnatw.K}
11528 @emph{Suppress warnings on redefinition of names in standard.}
11530 This switch disables warnings for declarations that declare a name that
11531 is defined in package Standard.
11534 @geindex -gnatwl (gcc)
11539 @item @code{-gnatwl}
11541 @emph{Activate warnings for elaboration pragmas.}
11543 @geindex Elaboration
11546 This switch activates warnings for possible elaboration problems,
11547 including suspicious use
11548 of @code{Elaborate} pragmas, when using the static elaboration model, and
11549 possible situations that may raise @code{Program_Error} when using the
11550 dynamic elaboration model.
11551 See the section in this guide on elaboration checking for further details.
11552 The default is that such warnings
11556 @geindex -gnatwL (gcc)
11561 @item @code{-gnatwL}
11563 @emph{Suppress warnings for elaboration pragmas.}
11565 This switch suppresses warnings for possible elaboration problems.
11568 @geindex -gnatw.l (gcc)
11573 @item @code{-gnatw.l}
11575 @emph{List inherited aspects.}
11577 This switch causes the compiler to list inherited invariants,
11578 preconditions, and postconditions from Type_Invariant’Class, Invariant’Class,
11579 Pre’Class, and Post’Class aspects. Also list inherited subtype predicates.
11582 @geindex -gnatw.L (gcc)
11587 @item @code{-gnatw.L}
11589 @emph{Suppress listing of inherited aspects.}
11591 This switch suppresses listing of inherited aspects.
11594 @geindex -gnatwm (gcc)
11599 @item @code{-gnatwm}
11601 @emph{Activate warnings on modified but unreferenced variables.}
11603 This switch activates warnings for variables that are assigned (using
11604 an initialization value or with one or more assignment statements) but
11605 whose value is never read. The warning is suppressed for volatile
11606 variables and also for variables that are renamings of other variables
11607 or for which an address clause is given.
11608 The default is that these warnings are not given.
11611 @geindex -gnatwM (gcc)
11616 @item @code{-gnatwM}
11618 @emph{Disable warnings on modified but unreferenced variables.}
11620 This switch disables warnings for variables that are assigned or
11621 initialized, but never read.
11624 @geindex -gnatw.m (gcc)
11629 @item @code{-gnatw.m}
11631 @emph{Activate warnings on suspicious modulus values.}
11633 This switch activates warnings for modulus values that seem suspicious.
11634 The cases caught are where the size is the same as the modulus (e.g.
11635 a modulus of 7 with a size of 7 bits), and modulus values of 32 or 64
11636 with no size clause. The guess in both cases is that 2**x was intended
11637 rather than x. In addition expressions of the form 2*x for small x
11638 generate a warning (the almost certainly accurate guess being that
11639 2**x was intended). This switch also activates warnings for negative
11640 literal values of a modular type, which are interpreted as large positive
11641 integers after wrap-around. The default is that these warnings are given.
11644 @geindex -gnatw.M (gcc)
11649 @item @code{-gnatw.M}
11651 @emph{Disable warnings on suspicious modulus values.}
11653 This switch disables warnings for suspicious modulus values.
11656 @geindex -gnatwn (gcc)
11661 @item @code{-gnatwn}
11663 @emph{Set normal warnings mode.}
11665 This switch sets normal warning mode, in which enabled warnings are
11666 issued and treated as warnings rather than errors. This is the default
11667 mode. the switch @code{-gnatwn} can be used to cancel the effect of
11668 an explicit @code{-gnatws} or
11669 @code{-gnatwe}. It also cancels the effect of the
11670 implicit @code{-gnatwe} that is activated by the
11671 use of @code{-gnatg}.
11674 @geindex -gnatw.n (gcc)
11676 @geindex Atomic Synchronization
11682 @item @code{-gnatw.n}
11684 @emph{Activate warnings on atomic synchronization.}
11686 This switch actives warnings when an access to an atomic variable
11687 requires the generation of atomic synchronization code. These
11688 warnings are off by default.
11691 @geindex -gnatw.N (gcc)
11696 @item @code{-gnatw.N}
11698 @emph{Suppress warnings on atomic synchronization.}
11700 @geindex Atomic Synchronization
11703 This switch suppresses warnings when an access to an atomic variable
11704 requires the generation of atomic synchronization code.
11707 @geindex -gnatwo (gcc)
11709 @geindex Address Clauses
11715 @item @code{-gnatwo}
11717 @emph{Activate warnings on address clause overlays.}
11719 This switch activates warnings for possibly unintended initialization
11720 effects of defining address clauses that cause one variable to overlap
11721 another. The default is that such warnings are generated.
11724 @geindex -gnatwO (gcc)
11729 @item @code{-gnatwO}
11731 @emph{Suppress warnings on address clause overlays.}
11733 This switch suppresses warnings on possibly unintended initialization
11734 effects of defining address clauses that cause one variable to overlap
11738 @geindex -gnatw.o (gcc)
11743 @item @code{-gnatw.o}
11745 @emph{Activate warnings on modified but unreferenced out parameters.}
11747 This switch activates warnings for variables that are modified by using
11748 them as actuals for a call to a procedure with an out mode formal, where
11749 the resulting assigned value is never read. It is applicable in the case
11750 where there is more than one out mode formal. If there is only one out
11751 mode formal, the warning is issued by default (controlled by -gnatwu).
11752 The warning is suppressed for volatile
11753 variables and also for variables that are renamings of other variables
11754 or for which an address clause is given.
11755 The default is that these warnings are not given.
11758 @geindex -gnatw.O (gcc)
11763 @item @code{-gnatw.O}
11765 @emph{Disable warnings on modified but unreferenced out parameters.}
11767 This switch suppresses warnings for variables that are modified by using
11768 them as actuals for a call to a procedure with an out mode formal, where
11769 the resulting assigned value is never read.
11772 @geindex -gnatwp (gcc)
11780 @item @code{-gnatwp}
11782 @emph{Activate warnings on ineffective pragma Inlines.}
11784 This switch activates warnings for failure of front end inlining
11785 (activated by @code{-gnatN}) to inline a particular call. There are
11786 many reasons for not being able to inline a call, including most
11787 commonly that the call is too complex to inline. The default is
11788 that such warnings are not given.
11789 Warnings on ineffective inlining by the gcc back-end can be activated
11790 separately, using the gcc switch -Winline.
11793 @geindex -gnatwP (gcc)
11798 @item @code{-gnatwP}
11800 @emph{Suppress warnings on ineffective pragma Inlines.}
11802 This switch suppresses warnings on ineffective pragma Inlines. If the
11803 inlining mechanism cannot inline a call, it will simply ignore the
11807 @geindex -gnatw.p (gcc)
11809 @geindex Parameter order
11815 @item @code{-gnatw.p}
11817 @emph{Activate warnings on parameter ordering.}
11819 This switch activates warnings for cases of suspicious parameter
11820 ordering when the list of arguments are all simple identifiers that
11821 match the names of the formals, but are in a different order. The
11822 warning is suppressed if any use of named parameter notation is used,
11823 so this is the appropriate way to suppress a false positive (and
11824 serves to emphasize that the “misordering” is deliberate). The
11825 default is that such warnings are not given.
11828 @geindex -gnatw.P (gcc)
11833 @item @code{-gnatw.P}
11835 @emph{Suppress warnings on parameter ordering.}
11837 This switch suppresses warnings on cases of suspicious parameter
11841 @geindex -gnatwq (gcc)
11843 @geindex Parentheses
11849 @item @code{-gnatwq}
11851 @emph{Activate warnings on questionable missing parentheses.}
11853 This switch activates warnings for cases where parentheses are not used and
11854 the result is potential ambiguity from a readers point of view. For example
11855 (not a > b) when a and b are modular means ((not a) > b) and very likely the
11856 programmer intended (not (a > b)). Similarly (-x mod 5) means (-(x mod 5)) and
11857 quite likely ((-x) mod 5) was intended. In such situations it seems best to
11858 follow the rule of always parenthesizing to make the association clear, and
11859 this warning switch warns if such parentheses are not present. The default
11860 is that these warnings are given.
11863 @geindex -gnatwQ (gcc)
11868 @item @code{-gnatwQ}
11870 @emph{Suppress warnings on questionable missing parentheses.}
11872 This switch suppresses warnings for cases where the association is not
11873 clear and the use of parentheses is preferred.
11876 @geindex -gnatw.q (gcc)
11884 @item @code{-gnatw.q}
11886 @emph{Activate warnings on questionable layout of record types.}
11888 This switch activates warnings for cases where the default layout of
11889 a record type, that is to say the layout of its components in textual
11890 order of the source code, would very likely cause inefficiencies in
11891 the code generated by the compiler, both in terms of space and speed
11892 during execution. One warning is issued for each problematic component
11893 without representation clause in the nonvariant part and then in each
11894 variant recursively, if any.
11896 The purpose of these warnings is neither to prescribe an optimal layout
11897 nor to force the use of representation clauses, but rather to get rid of
11898 the most blatant inefficiencies in the layout. Therefore, the default
11899 layout is matched against the following synthetic ordered layout and
11900 the deviations are flagged on a component-by-component basis:
11906 first all components or groups of components whose length is fixed
11907 and a multiple of the storage unit,
11910 then the remaining components whose length is fixed and not a multiple
11911 of the storage unit,
11914 then the remaining components whose length doesn’t depend on discriminants
11915 (that is to say, with variable but uniform length for all objects),
11918 then all components whose length depends on discriminants,
11921 finally the variant part (if any),
11924 for the nonvariant part and for each variant recursively, if any.
11926 The exact wording of the warning depends on whether the compiler is allowed
11927 to reorder the components in the record type or precluded from doing it by
11928 means of pragma @code{No_Component_Reordering}.
11930 The default is that these warnings are not given.
11933 @geindex -gnatw.Q (gcc)
11938 @item @code{-gnatw.Q}
11940 @emph{Suppress warnings on questionable layout of record types.}
11942 This switch suppresses warnings for cases where the default layout of
11943 a record type would very likely cause inefficiencies.
11946 @geindex -gnatwr (gcc)
11951 @item @code{-gnatwr}
11953 @emph{Activate warnings on redundant constructs.}
11955 This switch activates warnings for redundant constructs. The following
11956 is the current list of constructs regarded as redundant:
11962 Assignment of an item to itself.
11965 Type conversion that converts an expression to its own type.
11968 Use of the attribute @code{Base} where @code{typ'Base} is the same
11972 Use of pragma @code{Pack} when all components are placed by a record
11973 representation clause.
11976 Exception handler containing only a reraise statement (raise with no
11977 operand) which has no effect.
11980 Use of the operator abs on an operand that is known at compile time
11984 Comparison of an object or (unary or binary) operation of boolean type to
11985 an explicit True value.
11988 The default is that warnings for redundant constructs are not given.
11991 @geindex -gnatwR (gcc)
11996 @item @code{-gnatwR}
11998 @emph{Suppress warnings on redundant constructs.}
12000 This switch suppresses warnings for redundant constructs.
12003 @geindex -gnatw.r (gcc)
12008 @item @code{-gnatw.r}
12010 @emph{Activate warnings for object renaming function.}
12012 This switch activates warnings for an object renaming that renames a
12013 function call, which is equivalent to a constant declaration (as
12014 opposed to renaming the function itself). The default is that these
12015 warnings are given.
12018 @geindex -gnatw.R (gcc)
12023 @item @code{-gnatw.R}
12025 @emph{Suppress warnings for object renaming function.}
12027 This switch suppresses warnings for object renaming function.
12030 @geindex -gnatw_r (gcc)
12035 @item @code{-gnatw_r}
12037 @emph{Activate warnings for out-of-order record representation clauses.}
12039 This switch activates warnings for record representation clauses,
12040 if the order of component declarations, component clauses,
12041 and bit-level layout do not all agree.
12042 The default is that these warnings are not given.
12045 @geindex -gnatw_R (gcc)
12050 @item @code{-gnatw_R}
12052 @emph{Suppress warnings for out-of-order record representation clauses.}
12055 @geindex -gnatws (gcc)
12060 @item @code{-gnatws}
12062 @emph{Suppress all warnings.}
12064 This switch completely suppresses the
12065 output of all warning messages from the GNAT front end, including
12066 both warnings that can be controlled by switches described in this
12067 section, and those that are normally given unconditionally. The
12068 effect of this suppress action can only be cancelled by a subsequent
12069 use of the switch @code{-gnatwn}.
12071 Note that switch @code{-gnatws} does not suppress
12072 warnings from the @code{gcc} back end.
12073 To suppress these back end warnings as well, use the switch @code{-w}
12074 in addition to @code{-gnatws}. Also this switch has no effect on the
12075 handling of style check messages.
12078 @geindex -gnatw.s (gcc)
12080 @geindex Record Representation (component sizes)
12085 @item @code{-gnatw.s}
12087 @emph{Activate warnings on overridden size clauses.}
12089 This switch activates warnings on component clauses in record
12090 representation clauses where the length given overrides that
12091 specified by an explicit size clause for the component type. A
12092 warning is similarly given in the array case if a specified
12093 component size overrides an explicit size clause for the array
12097 @geindex -gnatw.S (gcc)
12102 @item @code{-gnatw.S}
12104 @emph{Suppress warnings on overridden size clauses.}
12106 This switch suppresses warnings on component clauses in record
12107 representation clauses that override size clauses, and similar
12108 warnings when an array component size overrides a size clause.
12111 @geindex -gnatwt (gcc)
12113 @geindex Deactivated code
12116 @geindex Deleted code
12122 @item @code{-gnatwt}
12124 @emph{Activate warnings for tracking of deleted conditional code.}
12126 This switch activates warnings for tracking of code in conditionals (IF and
12127 CASE statements) that is detected to be dead code which cannot be executed, and
12128 which is removed by the front end. This warning is off by default. This may be
12129 useful for detecting deactivated code in certified applications.
12132 @geindex -gnatwT (gcc)
12137 @item @code{-gnatwT}
12139 @emph{Suppress warnings for tracking of deleted conditional code.}
12141 This switch suppresses warnings for tracking of deleted conditional code.
12144 @geindex -gnatw.t (gcc)
12149 @item @code{-gnatw.t}
12151 @emph{Activate warnings on suspicious contracts.}
12153 This switch activates warnings on suspicious contracts. This includes
12154 warnings on suspicious postconditions (whether a pragma @code{Postcondition} or a
12155 @code{Post} aspect in Ada 2012) and suspicious contract cases (pragma or aspect
12156 @code{Contract_Cases}). A function postcondition or contract case is suspicious
12157 when no postcondition or contract case for this function mentions the result
12158 of the function. A procedure postcondition or contract case is suspicious
12159 when it only refers to the pre-state of the procedure, because in that case
12160 it should rather be expressed as a precondition. This switch also controls
12161 warnings on suspicious cases of expressions typically found in contracts like
12162 quantified expressions and uses of Update attribute. The default is that such
12163 warnings are generated.
12166 @geindex -gnatw.T (gcc)
12171 @item @code{-gnatw.T}
12173 @emph{Suppress warnings on suspicious contracts.}
12175 This switch suppresses warnings on suspicious contracts.
12178 @geindex -gnatwu (gcc)
12183 @item @code{-gnatwu}
12185 @emph{Activate warnings on unused entities.}
12187 This switch activates warnings to be generated for entities that
12188 are declared but not referenced, and for units that are @emph{with}ed
12190 referenced. In the case of packages, a warning is also generated if
12191 no entities in the package are referenced. This means that if a with’ed
12192 package is referenced but the only references are in @code{use}
12193 clauses or @code{renames}
12194 declarations, a warning is still generated. A warning is also generated
12195 for a generic package that is @emph{with}ed but never instantiated.
12196 In the case where a package or subprogram body is compiled, and there
12197 is a @emph{with} on the corresponding spec
12198 that is only referenced in the body,
12199 a warning is also generated, noting that the
12200 @emph{with} can be moved to the body. The default is that
12201 such warnings are not generated.
12202 This switch also activates warnings on unreferenced formals
12203 (it includes the effect of @code{-gnatwf}).
12206 @geindex -gnatwU (gcc)
12211 @item @code{-gnatwU}
12213 @emph{Suppress warnings on unused entities.}
12215 This switch suppresses warnings for unused entities and packages.
12216 It also turns off warnings on unreferenced formals (and thus includes
12217 the effect of @code{-gnatwF}).
12220 @geindex -gnatw.u (gcc)
12225 @item @code{-gnatw.u}
12227 @emph{Activate warnings on unordered enumeration types.}
12229 This switch causes enumeration types to be considered as conceptually
12230 unordered, unless an explicit pragma @code{Ordered} is given for the type.
12231 The effect is to generate warnings in clients that use explicit comparisons
12232 or subranges, since these constructs both treat objects of the type as
12233 ordered. (A @emph{client} is defined as a unit that is other than the unit in
12234 which the type is declared, or its body or subunits.) Please refer to
12235 the description of pragma @code{Ordered} in the
12236 @cite{GNAT Reference Manual} for further details.
12237 The default is that such warnings are not generated.
12240 @geindex -gnatw.U (gcc)
12245 @item @code{-gnatw.U}
12247 @emph{Deactivate warnings on unordered enumeration types.}
12249 This switch causes all enumeration types to be considered as ordered, so
12250 that no warnings are given for comparisons or subranges for any type.
12253 @geindex -gnatwv (gcc)
12255 @geindex Unassigned variable warnings
12260 @item @code{-gnatwv}
12262 @emph{Activate warnings on unassigned variables.}
12264 This switch activates warnings for access to variables which
12265 may not be properly initialized. The default is that
12266 such warnings are generated. This switch will also be emitted when
12267 initializing an array or record object via the following aggregate:
12270 Array_Or_Record : XXX := (others => <>);
12273 unless the relevant type fully initializes all components.
12276 @geindex -gnatwV (gcc)
12281 @item @code{-gnatwV}
12283 @emph{Suppress warnings on unassigned variables.}
12285 This switch suppresses warnings for access to variables which
12286 may not be properly initialized.
12289 @geindex -gnatw.v (gcc)
12291 @geindex bit order warnings
12296 @item @code{-gnatw.v}
12298 @emph{Activate info messages for non-default bit order.}
12300 This switch activates messages (labeled “info”, they are not warnings,
12301 just informational messages) about the effects of non-default bit-order
12302 on records to which a component clause is applied. The effect of specifying
12303 non-default bit ordering is a bit subtle (and changed with Ada 2005), so
12304 these messages, which are given by default, are useful in understanding the
12305 exact consequences of using this feature.
12308 @geindex -gnatw.V (gcc)
12313 @item @code{-gnatw.V}
12315 @emph{Suppress info messages for non-default bit order.}
12317 This switch suppresses information messages for the effects of specifying
12318 non-default bit order on record components with component clauses.
12321 @geindex -gnatww (gcc)
12323 @geindex String indexing warnings
12328 @item @code{-gnatww}
12330 @emph{Activate warnings on wrong low bound assumption.}
12332 This switch activates warnings for indexing an unconstrained string parameter
12333 with a literal or S’Length. This is a case where the code is assuming that the
12334 low bound is one, which is in general not true (for example when a slice is
12335 passed). The default is that such warnings are generated.
12338 @geindex -gnatwW (gcc)
12343 @item @code{-gnatwW}
12345 @emph{Suppress warnings on wrong low bound assumption.}
12347 This switch suppresses warnings for indexing an unconstrained string parameter
12348 with a literal or S’Length. Note that this warning can also be suppressed
12349 in a particular case by adding an assertion that the lower bound is 1,
12350 as shown in the following example:
12353 procedure K (S : String) is
12354 pragma Assert (S'First = 1);
12359 @geindex -gnatw.w (gcc)
12361 @geindex Warnings Off control
12366 @item @code{-gnatw.w}
12368 @emph{Activate warnings on Warnings Off pragmas.}
12370 This switch activates warnings for use of @code{pragma Warnings (Off, entity)}
12371 where either the pragma is entirely useless (because it suppresses no
12372 warnings), or it could be replaced by @code{pragma Unreferenced} or
12373 @code{pragma Unmodified}.
12374 Also activates warnings for the case of
12375 Warnings (Off, String), where either there is no matching
12376 Warnings (On, String), or the Warnings (Off) did not suppress any warning.
12377 The default is that these warnings are not given.
12380 @geindex -gnatw.W (gcc)
12385 @item @code{-gnatw.W}
12387 @emph{Suppress warnings on unnecessary Warnings Off pragmas.}
12389 This switch suppresses warnings for use of @code{pragma Warnings (Off, ...)}.
12392 @geindex -gnatwx (gcc)
12394 @geindex Export/Import pragma warnings
12399 @item @code{-gnatwx}
12401 @emph{Activate warnings on Export/Import pragmas.}
12403 This switch activates warnings on Export/Import pragmas when
12404 the compiler detects a possible conflict between the Ada and
12405 foreign language calling sequences. For example, the use of
12406 default parameters in a convention C procedure is dubious
12407 because the C compiler cannot supply the proper default, so
12408 a warning is issued. The default is that such warnings are
12412 @geindex -gnatwX (gcc)
12417 @item @code{-gnatwX}
12419 @emph{Suppress warnings on Export/Import pragmas.}
12421 This switch suppresses warnings on Export/Import pragmas.
12422 The sense of this is that you are telling the compiler that
12423 you know what you are doing in writing the pragma, and it
12424 should not complain at you.
12427 @geindex -gnatwm (gcc)
12432 @item @code{-gnatw.x}
12434 @emph{Activate warnings for No_Exception_Propagation mode.}
12436 This switch activates warnings for exception usage when pragma Restrictions
12437 (No_Exception_Propagation) is in effect. Warnings are given for implicit or
12438 explicit exception raises which are not covered by a local handler, and for
12439 exception handlers which do not cover a local raise. The default is that
12440 these warnings are given for units that contain exception handlers.
12442 @item @code{-gnatw.X}
12444 @emph{Disable warnings for No_Exception_Propagation mode.}
12446 This switch disables warnings for exception usage when pragma Restrictions
12447 (No_Exception_Propagation) is in effect.
12450 @geindex -gnatwy (gcc)
12452 @geindex Ada compatibility issues warnings
12457 @item @code{-gnatwy}
12459 @emph{Activate warnings for Ada compatibility issues.}
12461 For the most part, newer versions of Ada are upwards compatible
12462 with older versions. For example, Ada 2005 programs will almost
12463 always work when compiled as Ada 2012.
12464 However there are some exceptions (for example the fact that
12465 @code{some} is now a reserved word in Ada 2012). This
12466 switch activates several warnings to help in identifying
12467 and correcting such incompatibilities. The default is that
12468 these warnings are generated. Note that at one point Ada 2005
12469 was called Ada 0Y, hence the choice of character.
12472 @geindex -gnatwY (gcc)
12474 @geindex Ada compatibility issues warnings
12479 @item @code{-gnatwY}
12481 @emph{Disable warnings for Ada compatibility issues.}
12483 This switch suppresses the warnings intended to help in identifying
12484 incompatibilities between Ada language versions.
12487 @geindex -gnatw.y (gcc)
12489 @geindex Package spec needing body
12494 @item @code{-gnatw.y}
12496 @emph{Activate information messages for why package spec needs body.}
12498 There are a number of cases in which a package spec needs a body.
12499 For example, the use of pragma Elaborate_Body, or the declaration
12500 of a procedure specification requiring a completion. This switch
12501 causes information messages to be output showing why a package
12502 specification requires a body. This can be useful in the case of
12503 a large package specification which is unexpectedly requiring a
12504 body. The default is that such information messages are not output.
12507 @geindex -gnatw.Y (gcc)
12509 @geindex No information messages for why package spec needs body
12514 @item @code{-gnatw.Y}
12516 @emph{Disable information messages for why package spec needs body.}
12518 This switch suppresses the output of information messages showing why
12519 a package specification needs a body.
12522 @geindex -gnatwz (gcc)
12524 @geindex Unchecked_Conversion warnings
12529 @item @code{-gnatwz}
12531 @emph{Activate warnings on unchecked conversions.}
12533 This switch activates warnings for unchecked conversions
12534 where the types are known at compile time to have different
12535 sizes. The default is that such warnings are generated. Warnings are also
12536 generated for subprogram pointers with different conventions.
12539 @geindex -gnatwZ (gcc)
12544 @item @code{-gnatwZ}
12546 @emph{Suppress warnings on unchecked conversions.}
12548 This switch suppresses warnings for unchecked conversions
12549 where the types are known at compile time to have different
12550 sizes or conventions.
12553 @geindex -gnatw.z (gcc)
12555 @geindex Size/Alignment warnings
12560 @item @code{-gnatw.z}
12562 @emph{Activate warnings for size not a multiple of alignment.}
12564 This switch activates warnings for cases of array and record types
12565 with specified @code{Size} and @code{Alignment} attributes where the
12566 size is not a multiple of the alignment, resulting in an object
12567 size that is greater than the specified size. The default
12568 is that such warnings are generated.
12571 @geindex -gnatw.Z (gcc)
12573 @geindex Size/Alignment warnings
12578 @item @code{-gnatw.Z}
12580 @emph{Suppress warnings for size not a multiple of alignment.}
12582 This switch suppresses warnings for cases of array and record types
12583 with specified @code{Size} and @code{Alignment} attributes where the
12584 size is not a multiple of the alignment, resulting in an object
12585 size that is greater than the specified size. The warning can also
12586 be suppressed by giving an explicit @code{Object_Size} value.
12589 @geindex -Wunused (gcc)
12594 @item @code{-Wunused}
12596 The warnings controlled by the @code{-gnatw} switch are generated by
12597 the front end of the compiler. The GCC back end can provide
12598 additional warnings and they are controlled by the @code{-W} switch.
12599 For example, @code{-Wunused} activates back end
12600 warnings for entities that are declared but not referenced.
12603 @geindex -Wuninitialized (gcc)
12608 @item @code{-Wuninitialized}
12610 Similarly, @code{-Wuninitialized} activates
12611 the back end warning for uninitialized variables. This switch must be
12612 used in conjunction with an optimization level greater than zero.
12615 @geindex -Wstack-usage (gcc)
12620 @item @code{-Wstack-usage=@emph{len}}
12622 Warn if the stack usage of a subprogram might be larger than @code{len} bytes.
12623 See @ref{e6,,Static Stack Usage Analysis} for details.
12626 @geindex -Wall (gcc)
12633 This switch enables most warnings from the GCC back end.
12634 The code generator detects a number of warning situations that are missed
12635 by the GNAT front end, and this switch can be used to activate them.
12636 The use of this switch also sets the default front end warning mode to
12637 @code{-gnatwa}, that is, most front end warnings activated as well.
12647 Conversely, this switch suppresses warnings from the GCC back end.
12648 The use of this switch also sets the default front end warning mode to
12649 @code{-gnatws}, that is, front end warnings suppressed as well.
12652 @geindex -Werror (gcc)
12657 @item @code{-Werror}
12659 This switch causes warnings from the GCC back end to be treated as
12660 errors. The warning string still appears, but the warning messages are
12661 counted as errors, and prevent the generation of an object file.
12664 A string of warning parameters can be used in the same parameter. For example:
12670 will turn on all optional warnings except for unrecognized pragma warnings,
12671 and also specify that warnings should be treated as errors.
12673 When no switch @code{-gnatw} is used, this is equivalent to:
12820 @node Debugging and Assertion Control,Validity Checking,Warning Message Control,Compiler Switches
12821 @anchor{gnat_ugn/building_executable_programs_with_gnat debugging-and-assertion-control}@anchor{f1}@anchor{gnat_ugn/building_executable_programs_with_gnat id16}@anchor{f2}
12822 @subsection Debugging and Assertion Control
12825 @geindex -gnata (gcc)
12830 @item @code{-gnata}
12836 @geindex Assertions
12838 @geindex Precondition
12840 @geindex Postcondition
12842 @geindex Type invariants
12844 @geindex Subtype predicates
12846 The @code{-gnata} option is equivalent to the following @code{Assertion_Policy} pragma:
12849 pragma Assertion_Policy (Check);
12852 Which is a shorthand for:
12855 pragma Assertion_Policy
12857 Static_Predicate => Check,
12858 Dynamic_Predicate => Check,
12860 Pre'Class => Check,
12862 Post'Class => Check,
12863 Type_Invariant => Check,
12864 Type_Invariant'Class => Check);
12867 The pragmas @code{Assert} and @code{Debug} normally have no effect and
12868 are ignored. This switch, where @code{a} stands for ‘assert’, causes
12869 pragmas @code{Assert} and @code{Debug} to be activated. This switch also
12870 causes preconditions, postconditions, subtype predicates, and
12871 type invariants to be activated.
12873 The pragmas have the form:
12876 pragma Assert (<Boolean-expression> [, <static-string-expression>])
12877 pragma Debug (<procedure call>)
12878 pragma Type_Invariant (<type-local-name>, <Boolean-expression>)
12879 pragma Predicate (<type-local-name>, <Boolean-expression>)
12880 pragma Precondition (<Boolean-expression>, <string-expression>)
12881 pragma Postcondition (<Boolean-expression>, <string-expression>)
12884 The aspects have the form:
12887 with [Pre|Post|Type_Invariant|Dynamic_Predicate|Static_Predicate]
12888 => <Boolean-expression>;
12891 The @code{Assert} pragma causes @code{Boolean-expression} to be tested.
12892 If the result is @code{True}, the pragma has no effect (other than
12893 possible side effects from evaluating the expression). If the result is
12894 @code{False}, the exception @code{Assert_Failure} declared in the package
12895 @code{System.Assertions} is raised (passing @code{static-string-expression}, if
12896 present, as the message associated with the exception). If no string
12897 expression is given, the default is a string containing the file name and
12898 line number of the pragma.
12900 The @code{Debug} pragma causes @code{procedure} to be called. Note that
12901 @code{pragma Debug} may appear within a declaration sequence, allowing
12902 debugging procedures to be called between declarations.
12904 For the aspect specification, the @code{Boolean-expression} is evaluated.
12905 If the result is @code{True}, the aspect has no effect. If the result
12906 is @code{False}, the exception @code{Assert_Failure} is raised.
12909 @node Validity Checking,Style Checking,Debugging and Assertion Control,Compiler Switches
12910 @anchor{gnat_ugn/building_executable_programs_with_gnat id17}@anchor{f3}@anchor{gnat_ugn/building_executable_programs_with_gnat validity-checking}@anchor{e7}
12911 @subsection Validity Checking
12914 @geindex Validity Checking
12916 The Ada Reference Manual defines the concept of invalid values (see
12917 RM 13.9.1). The primary source of invalid values is uninitialized
12918 variables. A scalar variable that is left uninitialized may contain
12919 an invalid value; the concept of invalid does not apply to access or
12922 It is an error to read an invalid value, but the RM does not require
12923 run-time checks to detect such errors, except for some minimal
12924 checking to prevent erroneous execution (i.e. unpredictable
12925 behavior). This corresponds to the @code{-gnatVd} switch below,
12926 which is the default. For example, by default, if the expression of a
12927 case statement is invalid, it will raise Constraint_Error rather than
12928 causing a wild jump, and if an array index on the left-hand side of an
12929 assignment is invalid, it will raise Constraint_Error rather than
12930 overwriting an arbitrary memory location.
12932 The @code{-gnatVa} may be used to enable additional validity checks,
12933 which are not required by the RM. These checks are often very
12934 expensive (which is why the RM does not require them). These checks
12935 are useful in tracking down uninitialized variables, but they are
12936 not usually recommended for production builds, and in particular
12937 we do not recommend using these extra validity checking options in
12938 combination with optimization, since this can confuse the optimizer.
12939 If performance is a consideration, leading to the need to optimize,
12940 then the validity checking options should not be used.
12942 The other @code{-gnatV@emph{x}} switches below allow finer-grained
12943 control; you can enable whichever validity checks you desire. However,
12944 for most debugging purposes, @code{-gnatVa} is sufficient, and the
12945 default @code{-gnatVd} (i.e. standard Ada behavior) is usually
12946 sufficient for non-debugging use.
12948 The @code{-gnatB} switch tells the compiler to assume that all
12949 values are valid (that is, within their declared subtype range)
12950 except in the context of a use of the Valid attribute. This means
12951 the compiler can generate more efficient code, since the range
12952 of values is better known at compile time. However, an uninitialized
12953 variable can cause wild jumps and memory corruption in this mode.
12955 The @code{-gnatV@emph{x}} switch allows control over the validity
12956 checking mode as described below.
12957 The @code{x} argument is a string of letters that
12958 indicate validity checks that are performed or not performed in addition
12959 to the default checks required by Ada as described above.
12961 @geindex -gnatVa (gcc)
12966 @item @code{-gnatVa}
12968 @emph{All validity checks.}
12970 All validity checks are turned on.
12971 That is, @code{-gnatVa} is
12972 equivalent to @code{gnatVcdfimoprst}.
12975 @geindex -gnatVc (gcc)
12980 @item @code{-gnatVc}
12982 @emph{Validity checks for copies.}
12984 The right hand side of assignments, and the initializing values of
12985 object declarations are validity checked.
12988 @geindex -gnatVd (gcc)
12993 @item @code{-gnatVd}
12995 @emph{Default (RM) validity checks.}
12997 Some validity checks are done by default following normal Ada semantics
12998 (RM 13.9.1 (9-11)).
12999 A check is done in case statements that the expression is within the range
13000 of the subtype. If it is not, Constraint_Error is raised.
13001 For assignments to array components, a check is done that the expression used
13002 as index is within the range. If it is not, Constraint_Error is raised.
13003 Both these validity checks may be turned off using switch @code{-gnatVD}.
13004 They are turned on by default. If @code{-gnatVD} is specified, a subsequent
13005 switch @code{-gnatVd} will leave the checks turned on.
13006 Switch @code{-gnatVD} should be used only if you are sure that all such
13007 expressions have valid values. If you use this switch and invalid values
13008 are present, then the program is erroneous, and wild jumps or memory
13009 overwriting may occur.
13012 @geindex -gnatVe (gcc)
13017 @item @code{-gnatVe}
13019 @emph{Validity checks for elementary components.}
13021 In the absence of this switch, assignments to record or array components are
13022 not validity checked, even if validity checks for assignments generally
13023 (@code{-gnatVc}) are turned on. In Ada, assignment of composite values do not
13024 require valid data, but assignment of individual components does. So for
13025 example, there is a difference between copying the elements of an array with a
13026 slice assignment, compared to assigning element by element in a loop. This
13027 switch allows you to turn off validity checking for components, even when they
13028 are assigned component by component.
13031 @geindex -gnatVf (gcc)
13036 @item @code{-gnatVf}
13038 @emph{Validity checks for floating-point values.}
13040 In the absence of this switch, validity checking occurs only for discrete
13041 values. If @code{-gnatVf} is specified, then validity checking also applies
13042 for floating-point values, and NaNs and infinities are considered invalid,
13043 as well as out of range values for constrained types. Note that this means
13044 that standard IEEE infinity mode is not allowed. The exact contexts
13045 in which floating-point values are checked depends on the setting of other
13046 options. For example, @code{-gnatVif} or @code{-gnatVfi}
13047 (the order does not matter) specifies that floating-point parameters of mode
13048 @code{in} should be validity checked.
13051 @geindex -gnatVi (gcc)
13056 @item @code{-gnatVi}
13058 @emph{Validity checks for `@w{`}in`@w{`} mode parameters.}
13060 Arguments for parameters of mode @code{in} are validity checked in function
13061 and procedure calls at the point of call.
13064 @geindex -gnatVm (gcc)
13069 @item @code{-gnatVm}
13071 @emph{Validity checks for `@w{`}in out`@w{`} mode parameters.}
13073 Arguments for parameters of mode @code{in out} are validity checked in
13074 procedure calls at the point of call. The @code{'m'} here stands for
13075 modify, since this concerns parameters that can be modified by the call.
13076 Note that there is no specific option to test @code{out} parameters,
13077 but any reference within the subprogram will be tested in the usual
13078 manner, and if an invalid value is copied back, any reference to it
13079 will be subject to validity checking.
13082 @geindex -gnatVn (gcc)
13087 @item @code{-gnatVn}
13089 @emph{No validity checks.}
13091 This switch turns off all validity checking, including the default checking
13092 for case statements and left hand side subscripts. Note that the use of
13093 the switch @code{-gnatp} suppresses all run-time checks, including
13094 validity checks, and thus implies @code{-gnatVn}. When this switch
13095 is used, it cancels any other @code{-gnatV} previously issued.
13098 @geindex -gnatVo (gcc)
13103 @item @code{-gnatVo}
13105 @emph{Validity checks for operator and attribute operands.}
13107 Arguments for predefined operators and attributes are validity checked.
13108 This includes all operators in package @code{Standard},
13109 the shift operators defined as intrinsic in package @code{Interfaces}
13110 and operands for attributes such as @code{Pos}. Checks are also made
13111 on individual component values for composite comparisons, and on the
13112 expressions in type conversions and qualified expressions. Checks are
13113 also made on explicit ranges using @code{..} (e.g., slices, loops etc).
13116 @geindex -gnatVp (gcc)
13121 @item @code{-gnatVp}
13123 @emph{Validity checks for parameters.}
13125 This controls the treatment of parameters within a subprogram (as opposed
13126 to @code{-gnatVi} and @code{-gnatVm} which control validity testing
13127 of parameters on a call. If either of these call options is used, then
13128 normally an assumption is made within a subprogram that the input arguments
13129 have been validity checking at the point of call, and do not need checking
13130 again within a subprogram). If @code{-gnatVp} is set, then this assumption
13131 is not made, and parameters are not assumed to be valid, so their validity
13132 will be checked (or rechecked) within the subprogram.
13135 @geindex -gnatVr (gcc)
13140 @item @code{-gnatVr}
13142 @emph{Validity checks for function returns.}
13144 The expression in @code{return} statements in functions is validity
13148 @geindex -gnatVs (gcc)
13153 @item @code{-gnatVs}
13155 @emph{Validity checks for subscripts.}
13157 All subscripts expressions are checked for validity, whether they appear
13158 on the right side or left side (in default mode only left side subscripts
13159 are validity checked).
13162 @geindex -gnatVt (gcc)
13167 @item @code{-gnatVt}
13169 @emph{Validity checks for tests.}
13171 Expressions used as conditions in @code{if}, @code{while} or @code{exit}
13172 statements are checked, as well as guard expressions in entry calls.
13175 The @code{-gnatV} switch may be followed by a string of letters
13176 to turn on a series of validity checking options.
13177 For example, @code{-gnatVcr}
13178 specifies that in addition to the default validity checking, copies and
13179 function return expressions are to be validity checked.
13180 In order to make it easier to specify the desired combination of effects,
13181 the upper case letters @code{CDFIMORST} may
13182 be used to turn off the corresponding lower case option.
13183 Thus @code{-gnatVaM} turns on all validity checking options except for
13184 checking of @code{in out} parameters.
13186 The specification of additional validity checking generates extra code (and
13187 in the case of @code{-gnatVa} the code expansion can be substantial).
13188 However, these additional checks can be very useful in detecting
13189 uninitialized variables, incorrect use of unchecked conversion, and other
13190 errors leading to invalid values. The use of pragma @code{Initialize_Scalars}
13191 is useful in conjunction with the extra validity checking, since this
13192 ensures that wherever possible uninitialized variables have invalid values.
13194 See also the pragma @code{Validity_Checks} which allows modification of
13195 the validity checking mode at the program source level, and also allows for
13196 temporary disabling of validity checks.
13198 @node Style Checking,Run-Time Checks,Validity Checking,Compiler Switches
13199 @anchor{gnat_ugn/building_executable_programs_with_gnat id18}@anchor{f4}@anchor{gnat_ugn/building_executable_programs_with_gnat style-checking}@anchor{ec}
13200 @subsection Style Checking
13203 @geindex Style checking
13205 @geindex -gnaty (gcc)
13207 The @code{-gnatyx} switch causes the compiler to
13208 enforce specified style rules. A limited set of style rules has been used
13209 in writing the GNAT sources themselves. This switch allows user programs
13210 to activate all or some of these checks. If the source program fails a
13211 specified style check, an appropriate message is given, preceded by
13212 the character sequence ‘(style)’. This message does not prevent
13213 successful compilation (unless the @code{-gnatwe} switch is used).
13215 Note that this is by no means intended to be a general facility for
13216 checking arbitrary coding standards. It is simply an embedding of the
13217 style rules we have chosen for the GNAT sources. If you are starting
13218 a project which does not have established style standards, you may
13219 find it useful to adopt the entire set of GNAT coding standards, or
13220 some subset of them.
13223 The string @code{x} is a sequence of letters or digits
13224 indicating the particular style
13225 checks to be performed. The following checks are defined:
13227 @geindex -gnaty[0-9] (gcc)
13232 @item @code{-gnaty0}
13234 @emph{Specify indentation level.}
13236 If a digit from 1-9 appears
13237 in the string after @code{-gnaty}
13238 then proper indentation is checked, with the digit indicating the
13239 indentation level required. A value of zero turns off this style check.
13240 The rule checks that the following constructs start on a column that is
13241 a multiple of the alignment level:
13247 beginnings of declarations (except record component declarations)
13251 beginnings of the structural components of compound statements;
13254 @code{end} keyword that completes the declaration of a program unit declaration
13255 or body or that completes a compound statement.
13258 Full line comments must be
13259 aligned with the @code{--} starting on a column that is a multiple of
13260 the alignment level, or they may be aligned the same way as the following
13261 non-blank line (this is useful when full line comments appear in the middle
13262 of a statement, or they may be aligned with the source line on the previous
13266 @geindex -gnatya (gcc)
13271 @item @code{-gnatya}
13273 @emph{Check attribute casing.}
13275 Attribute names, including the case of keywords such as @code{digits}
13276 used as attributes names, must be written in mixed case, that is, the
13277 initial letter and any letter following an underscore must be uppercase.
13278 All other letters must be lowercase.
13281 @geindex -gnatyA (gcc)
13286 @item @code{-gnatyA}
13288 @emph{Use of array index numbers in array attributes.}
13290 When using the array attributes First, Last, Range,
13291 or Length, the index number must be omitted for one-dimensional arrays
13292 and is required for multi-dimensional arrays.
13295 @geindex -gnatyb (gcc)
13300 @item @code{-gnatyb}
13302 @emph{Blanks not allowed at statement end.}
13304 Trailing blanks are not allowed at the end of statements. The purpose of this
13305 rule, together with h (no horizontal tabs), is to enforce a canonical format
13306 for the use of blanks to separate source tokens.
13309 @geindex -gnatyB (gcc)
13314 @item @code{-gnatyB}
13316 @emph{Check Boolean operators.}
13318 The use of AND/OR operators is not permitted except in the cases of modular
13319 operands, array operands, and simple stand-alone boolean variables or
13320 boolean constants. In all other cases @code{and then}/@cite{or else} are
13324 @geindex -gnatyc (gcc)
13329 @item @code{-gnatyc}
13331 @emph{Check comments, double space.}
13333 Comments must meet the following set of rules:
13339 The @code{--} that starts the column must either start in column one,
13340 or else at least one blank must precede this sequence.
13343 Comments that follow other tokens on a line must have at least one blank
13344 following the @code{--} at the start of the comment.
13347 Full line comments must have at least two blanks following the
13348 @code{--} that starts the comment, with the following exceptions.
13351 A line consisting only of the @code{--} characters, possibly preceded
13352 by blanks is permitted.
13355 A comment starting with @code{--x} where @code{x} is a special character
13357 This allows proper processing of the output from specialized tools
13358 such as @code{gnatprep} (where @code{--!} is used) and in earlier versions of the SPARK
13360 language (where @code{--#} is used). For the purposes of this rule, a
13361 special character is defined as being in one of the ASCII ranges
13362 @code{16#21#...16#2F#} or @code{16#3A#...16#3F#}.
13363 Note that this usage is not permitted
13364 in GNAT implementation units (i.e., when @code{-gnatg} is used).
13367 A line consisting entirely of minus signs, possibly preceded by blanks, is
13368 permitted. This allows the construction of box comments where lines of minus
13369 signs are used to form the top and bottom of the box.
13372 A comment that starts and ends with @code{--} is permitted as long as at
13373 least one blank follows the initial @code{--}. Together with the preceding
13374 rule, this allows the construction of box comments, as shown in the following
13378 ---------------------------
13379 -- This is a box comment --
13380 -- with two text lines. --
13381 ---------------------------
13386 @geindex -gnatyC (gcc)
13391 @item @code{-gnatyC}
13393 @emph{Check comments, single space.}
13395 This is identical to @code{c} except that only one space
13396 is required following the @code{--} of a comment instead of two.
13399 @geindex -gnatyd (gcc)
13404 @item @code{-gnatyd}
13406 @emph{Check no DOS line terminators present.}
13408 All lines must be terminated by a single ASCII.LF
13409 character (in particular the DOS line terminator sequence CR/LF is not
13413 @geindex -gnatyD (gcc)
13418 @item @code{-gnatyD}
13420 @emph{Check declared identifiers in mixed case.}
13422 Declared identifiers must be in mixed case, as in
13423 This_Is_An_Identifier. Use -gnatyr in addition to ensure
13424 that references match declarations.
13427 @geindex -gnatye (gcc)
13432 @item @code{-gnatye}
13434 @emph{Check end/exit labels.}
13436 Optional labels on @code{end} statements ending subprograms and on
13437 @code{exit} statements exiting named loops, are required to be present.
13440 @geindex -gnatyf (gcc)
13445 @item @code{-gnatyf}
13447 @emph{No form feeds or vertical tabs.}
13449 Neither form feeds nor vertical tab characters are permitted
13450 in the source text.
13453 @geindex -gnatyg (gcc)
13458 @item @code{-gnatyg}
13460 @emph{GNAT style mode.}
13462 The set of style check switches is set to match that used by the GNAT sources.
13463 This may be useful when developing code that is eventually intended to be
13464 incorporated into GNAT. Currently this is equivalent to @code{-gnatyydISux})
13465 but additional style switches may be added to this set in the future without
13469 @geindex -gnatyh (gcc)
13474 @item @code{-gnatyh}
13476 @emph{No horizontal tabs.}
13478 Horizontal tab characters are not permitted in the source text.
13479 Together with the b (no blanks at end of line) check, this
13480 enforces a canonical form for the use of blanks to separate
13484 @geindex -gnatyi (gcc)
13489 @item @code{-gnatyi}
13491 @emph{Check if-then layout.}
13493 The keyword @code{then} must appear either on the same
13494 line as corresponding @code{if}, or on a line on its own, lined
13495 up under the @code{if}.
13498 @geindex -gnatyI (gcc)
13503 @item @code{-gnatyI}
13505 @emph{check mode IN keywords.}
13507 Mode @code{in} (the default mode) is not
13508 allowed to be given explicitly. @code{in out} is fine,
13509 but not @code{in} on its own.
13512 @geindex -gnatyk (gcc)
13517 @item @code{-gnatyk}
13519 @emph{Check keyword casing.}
13521 All keywords must be in lower case (with the exception of keywords
13522 such as @code{digits} used as attribute names to which this check
13523 does not apply). A single error is reported for each line breaking
13524 this rule even if multiple casing issues exist on a same line.
13527 @geindex -gnatyl (gcc)
13532 @item @code{-gnatyl}
13534 @emph{Check layout.}
13536 Layout of statement and declaration constructs must follow the
13537 recommendations in the Ada Reference Manual, as indicated by the
13538 form of the syntax rules. For example an @code{else} keyword must
13539 be lined up with the corresponding @code{if} keyword.
13541 There are two respects in which the style rule enforced by this check
13542 option are more liberal than those in the Ada Reference Manual. First
13543 in the case of record declarations, it is permissible to put the
13544 @code{record} keyword on the same line as the @code{type} keyword, and
13545 then the @code{end} in @code{end record} must line up under @code{type}.
13546 This is also permitted when the type declaration is split on two lines.
13547 For example, any of the following three layouts is acceptable:
13568 Second, in the case of a block statement, a permitted alternative
13569 is to put the block label on the same line as the @code{declare} or
13570 @code{begin} keyword, and then line the @code{end} keyword up under
13571 the block label. For example both the following are permitted:
13588 The same alternative format is allowed for loops. For example, both of
13589 the following are permitted:
13592 Clear : while J < 10 loop
13603 @geindex -gnatyLnnn (gcc)
13608 @item @code{-gnatyL}
13610 @emph{Set maximum nesting level.}
13612 The maximum level of nesting of constructs (including subprograms, loops,
13613 blocks, packages, and conditionals) may not exceed the given value
13614 @emph{nnn}. A value of zero disconnects this style check.
13617 @geindex -gnatym (gcc)
13622 @item @code{-gnatym}
13624 @emph{Check maximum line length.}
13626 The length of source lines must not exceed 79 characters, including
13627 any trailing blanks. The value of 79 allows convenient display on an
13628 80 character wide device or window, allowing for possible special
13629 treatment of 80 character lines. Note that this count is of
13630 characters in the source text. This means that a tab character counts
13631 as one character in this count and a wide character sequence counts as
13632 a single character (however many bytes are needed in the encoding).
13635 @geindex -gnatyMnnn (gcc)
13640 @item @code{-gnatyM}
13642 @emph{Set maximum line length.}
13644 The length of lines must not exceed the
13645 given value @emph{nnn}. The maximum value that can be specified is 32767.
13646 If neither style option for setting the line length is used, then the
13647 default is 255. This also controls the maximum length of lexical elements,
13648 where the only restriction is that they must fit on a single line.
13651 @geindex -gnatyn (gcc)
13656 @item @code{-gnatyn}
13658 @emph{Check casing of entities in Standard.}
13660 Any identifier from Standard must be cased
13661 to match the presentation in the Ada Reference Manual (for example,
13662 @code{Integer} and @code{ASCII.NUL}).
13665 @geindex -gnatyN (gcc)
13670 @item @code{-gnatyN}
13672 @emph{Turn off all style checks.}
13674 All style check options are turned off.
13677 @geindex -gnatyo (gcc)
13682 @item @code{-gnatyo}
13684 @emph{Check order of subprogram bodies.}
13686 All subprogram bodies in a given scope
13687 (e.g., a package body) must be in alphabetical order. The ordering
13688 rule uses normal Ada rules for comparing strings, ignoring casing
13689 of letters, except that if there is a trailing numeric suffix, then
13690 the value of this suffix is used in the ordering (e.g., Junk2 comes
13694 @geindex -gnatyO (gcc)
13699 @item @code{-gnatyO}
13701 @emph{Check that overriding subprograms are explicitly marked as such.}
13703 This applies to all subprograms of a derived type that override a primitive
13704 operation of the type, for both tagged and untagged types. In particular,
13705 the declaration of a primitive operation of a type extension that overrides
13706 an inherited operation must carry an overriding indicator. Another case is
13707 the declaration of a function that overrides a predefined operator (such
13708 as an equality operator).
13711 @geindex -gnatyp (gcc)
13716 @item @code{-gnatyp}
13718 @emph{Check pragma casing.}
13720 Pragma names must be written in mixed case, that is, the
13721 initial letter and any letter following an underscore must be uppercase.
13722 All other letters must be lowercase. An exception is that SPARK_Mode is
13723 allowed as an alternative for Spark_Mode.
13726 @geindex -gnatyr (gcc)
13731 @item @code{-gnatyr}
13733 @emph{Check references.}
13735 All identifier references must be cased in the same way as the
13736 corresponding declaration. No specific casing style is imposed on
13737 identifiers. The only requirement is for consistency of references
13741 @geindex -gnatys (gcc)
13746 @item @code{-gnatys}
13748 @emph{Check separate specs.}
13750 Separate declarations (‘specs’) are required for subprograms (a
13751 body is not allowed to serve as its own declaration). The only
13752 exception is that parameterless library level procedures are
13753 not required to have a separate declaration. This exception covers
13754 the most frequent form of main program procedures.
13757 @geindex -gnatyS (gcc)
13762 @item @code{-gnatyS}
13764 @emph{Check no statements after then/else.}
13766 No statements are allowed
13767 on the same line as a @code{then} or @code{else} keyword following the
13768 keyword in an @code{if} statement. @code{or else} and @code{and then} are not
13769 affected, and a special exception allows a pragma to appear after @code{else}.
13772 @geindex -gnatyt (gcc)
13777 @item @code{-gnatyt}
13779 @emph{Check token spacing.}
13781 The following token spacing rules are enforced:
13787 The keywords @code{abs} and @code{not} must be followed by a space.
13790 The token @code{=>} must be surrounded by spaces.
13793 The token @code{<>} must be preceded by a space or a left parenthesis.
13796 Binary operators other than @code{**} must be surrounded by spaces.
13797 There is no restriction on the layout of the @code{**} binary operator.
13800 Colon must be surrounded by spaces.
13803 Colon-equal (assignment, initialization) must be surrounded by spaces.
13806 Comma must be the first non-blank character on the line, or be
13807 immediately preceded by a non-blank character, and must be followed
13811 If the token preceding a left parenthesis ends with a letter or digit, then
13812 a space must separate the two tokens.
13815 If the token following a right parenthesis starts with a letter or digit, then
13816 a space must separate the two tokens.
13819 A right parenthesis must either be the first non-blank character on
13820 a line, or it must be preceded by a non-blank character.
13823 A semicolon must not be preceded by a space, and must not be followed by
13824 a non-blank character.
13827 A unary plus or minus may not be followed by a space.
13830 A vertical bar must be surrounded by spaces.
13833 Exactly one blank (and no other white space) must appear between
13834 a @code{not} token and a following @code{in} token.
13837 @geindex -gnatyu (gcc)
13842 @item @code{-gnatyu}
13844 @emph{Check unnecessary blank lines.}
13846 Unnecessary blank lines are not allowed. A blank line is considered
13847 unnecessary if it appears at the end of the file, or if more than
13848 one blank line occurs in sequence.
13851 @geindex -gnatyx (gcc)
13856 @item @code{-gnatyx}
13858 @emph{Check extra parentheses.}
13860 Unnecessary extra level of parentheses (C-style) are not allowed
13861 around conditions in @code{if} statements, @code{while} statements and
13862 @code{exit} statements.
13865 @geindex -gnatyy (gcc)
13870 @item @code{-gnatyy}
13872 @emph{Set all standard style check options.}
13874 This is equivalent to @code{gnaty3aAbcefhiklmnprst}, that is all checking
13875 options enabled with the exception of @code{-gnatyB}, @code{-gnatyd},
13876 @code{-gnatyI}, @code{-gnatyLnnn}, @code{-gnatyo}, @code{-gnatyO},
13877 @code{-gnatyS}, @code{-gnatyu}, and @code{-gnatyx}.
13880 @geindex -gnaty- (gcc)
13885 @item @code{-gnaty-}
13887 @emph{Remove style check options.}
13889 This causes any subsequent options in the string to act as canceling the
13890 corresponding style check option. To cancel maximum nesting level control,
13891 use the @code{L} parameter without any integer value after that, because any
13892 digit following @emph{-} in the parameter string of the @code{-gnaty}
13893 option will be treated as canceling the indentation check. The same is true
13894 for the @code{M} parameter. @code{y} and @code{N} parameters are not
13895 allowed after @emph{-}.
13898 @geindex -gnaty+ (gcc)
13903 @item @code{-gnaty+}
13905 @emph{Enable style check options.}
13907 This causes any subsequent options in the string to enable the corresponding
13908 style check option. That is, it cancels the effect of a previous -,
13912 @c end of switch description (leave this comment to ease automatic parsing for
13916 In the above rules, appearing in column one is always permitted, that is,
13917 counts as meeting either a requirement for a required preceding space,
13918 or as meeting a requirement for no preceding space.
13920 Appearing at the end of a line is also always permitted, that is, counts
13921 as meeting either a requirement for a following space, or as meeting
13922 a requirement for no following space.
13924 If any of these style rules is violated, a message is generated giving
13925 details on the violation. The initial characters of such messages are
13926 always ‘@cite{(style)}’. Note that these messages are treated as warning
13927 messages, so they normally do not prevent the generation of an object
13928 file. The @code{-gnatwe} switch can be used to treat warning messages,
13929 including style messages, as fatal errors.
13931 The switch @code{-gnaty} on its own (that is not
13932 followed by any letters or digits) is equivalent
13933 to the use of @code{-gnatyy} as described above, that is all
13934 built-in standard style check options are enabled.
13936 The switch @code{-gnatyN} clears any previously set style checks.
13938 @node Run-Time Checks,Using gcc for Syntax Checking,Style Checking,Compiler Switches
13939 @anchor{gnat_ugn/building_executable_programs_with_gnat id19}@anchor{f5}@anchor{gnat_ugn/building_executable_programs_with_gnat run-time-checks}@anchor{ea}
13940 @subsection Run-Time Checks
13943 @geindex Division by zero
13945 @geindex Access before elaboration
13948 @geindex division by zero
13951 @geindex access before elaboration
13954 @geindex stack overflow checking
13956 By default, the following checks are suppressed: stack overflow
13957 checks, and checks for access before elaboration on subprogram
13958 calls. All other checks, including overflow checks, range checks and
13959 array bounds checks, are turned on by default. The following @code{gcc}
13960 switches refine this default behavior.
13962 @geindex -gnatp (gcc)
13967 @item @code{-gnatp}
13969 @geindex Suppressing checks
13972 @geindex suppressing
13974 This switch causes the unit to be compiled
13975 as though @code{pragma Suppress (All_checks)}
13976 had been present in the source. Validity checks are also eliminated (in
13977 other words @code{-gnatp} also implies @code{-gnatVn}.
13978 Use this switch to improve the performance
13979 of the code at the expense of safety in the presence of invalid data or
13982 Note that when checks are suppressed, the compiler is allowed, but not
13983 required, to omit the checking code. If the run-time cost of the
13984 checking code is zero or near-zero, the compiler will generate it even
13985 if checks are suppressed. In particular, if the compiler can prove
13986 that a certain check will necessarily fail, it will generate code to
13987 do an unconditional ‘raise’, even if checks are suppressed. The
13988 compiler warns in this case. Another case in which checks may not be
13989 eliminated is when they are embedded in certain run-time routines such
13990 as math library routines.
13992 Of course, run-time checks are omitted whenever the compiler can prove
13993 that they will not fail, whether or not checks are suppressed.
13995 Note that if you suppress a check that would have failed, program
13996 execution is erroneous, which means the behavior is totally
13997 unpredictable. The program might crash, or print wrong answers, or
13998 do anything else. It might even do exactly what you wanted it to do
13999 (and then it might start failing mysteriously next week or next
14000 year). The compiler will generate code based on the assumption that
14001 the condition being checked is true, which can result in erroneous
14002 execution if that assumption is wrong.
14004 The checks subject to suppression include all the checks defined by the Ada
14005 standard, the additional implementation defined checks @code{Alignment_Check},
14006 @code{Duplicated_Tag_Check}, @code{Predicate_Check}, @code{Container_Checks}, @code{Tampering_Check},
14007 and @code{Validity_Check}, as well as any checks introduced using @code{pragma Check_Name}.
14008 Note that @code{Atomic_Synchronization} is not automatically suppressed by use of this option.
14010 If the code depends on certain checks being active, you can use
14011 pragma @code{Unsuppress} either as a configuration pragma or as
14012 a local pragma to make sure that a specified check is performed
14013 even if @code{gnatp} is specified.
14015 The @code{-gnatp} switch has no effect if a subsequent
14016 @code{-gnat-p} switch appears.
14019 @geindex -gnat-p (gcc)
14021 @geindex Suppressing checks
14024 @geindex suppressing
14031 @item @code{-gnat-p}
14033 This switch cancels the effect of a previous @code{gnatp} switch.
14036 @geindex -gnato?? (gcc)
14038 @geindex Overflow checks
14040 @geindex Overflow mode
14048 @item @code{-gnato??}
14050 This switch controls the mode used for computing intermediate
14051 arithmetic integer operations, and also enables overflow checking.
14052 For a full description of overflow mode and checking control, see
14053 the ‘Overflow Check Handling in GNAT’ appendix in this
14056 Overflow checks are always enabled by this switch. The argument
14057 controls the mode, using the codes
14062 @item @emph{1 = STRICT}
14064 In STRICT mode, intermediate operations are always done using the
14065 base type, and overflow checking ensures that the result is within
14066 the base type range.
14068 @item @emph{2 = MINIMIZED}
14070 In MINIMIZED mode, overflows in intermediate operations are avoided
14071 where possible by using a larger integer type for the computation
14072 (typically @code{Long_Long_Integer}). Overflow checking ensures that
14073 the result fits in this larger integer type.
14075 @item @emph{3 = ELIMINATED}
14077 In ELIMINATED mode, overflows in intermediate operations are avoided
14078 by using multi-precision arithmetic. In this case, overflow checking
14079 has no effect on intermediate operations (since overflow is impossible).
14082 If two digits are present after @code{-gnato} then the first digit
14083 sets the mode for expressions outside assertions, and the second digit
14084 sets the mode for expressions within assertions. Here assertions is used
14085 in the technical sense (which includes for example precondition and
14086 postcondition expressions).
14088 If one digit is present, the corresponding mode is applicable to both
14089 expressions within and outside assertion expressions.
14091 If no digits are present, the default is to enable overflow checks
14092 and set STRICT mode for both kinds of expressions. This is compatible
14093 with the use of @code{-gnato} in previous versions of GNAT.
14095 @geindex Machine_Overflows
14097 Note that the @code{-gnato??} switch does not affect the code generated
14098 for any floating-point operations; it applies only to integer semantics.
14099 For floating-point, GNAT has the @code{Machine_Overflows}
14100 attribute set to @code{False} and the normal mode of operation is to
14101 generate IEEE NaN and infinite values on overflow or invalid operations
14102 (such as dividing 0.0 by 0.0).
14104 The reason that we distinguish overflow checking from other kinds of
14105 range constraint checking is that a failure of an overflow check, unlike
14106 for example the failure of a range check, can result in an incorrect
14107 value, but cannot cause random memory destruction (like an out of range
14108 subscript), or a wild jump (from an out of range case value). Overflow
14109 checking is also quite expensive in time and space, since in general it
14110 requires the use of double length arithmetic.
14112 Note again that the default is @code{-gnato11} (equivalent to @code{-gnato1}),
14113 so overflow checking is performed in STRICT mode by default.
14116 @geindex -gnatE (gcc)
14118 @geindex Elaboration checks
14121 @geindex elaboration
14126 @item @code{-gnatE}
14128 Enables dynamic checks for access-before-elaboration
14129 on subprogram calls and generic instantiations.
14130 Note that @code{-gnatE} is not necessary for safety, because in the
14131 default mode, GNAT ensures statically that the checks would not fail.
14132 For full details of the effect and use of this switch,
14133 @ref{c7,,Compiling with gcc}.
14136 @geindex -fstack-check (gcc)
14138 @geindex Stack Overflow Checking
14141 @geindex stack overflow checking
14146 @item @code{-fstack-check}
14148 Activates stack overflow checking. For full details of the effect and use of
14149 this switch see @ref{e5,,Stack Overflow Checking}.
14152 @geindex Unsuppress
14154 The setting of these switches only controls the default setting of the
14155 checks. You may modify them using either @code{Suppress} (to remove
14156 checks) or @code{Unsuppress} (to add back suppressed checks) pragmas in
14157 the program source.
14159 @node Using gcc for Syntax Checking,Using gcc for Semantic Checking,Run-Time Checks,Compiler Switches
14160 @anchor{gnat_ugn/building_executable_programs_with_gnat id20}@anchor{f6}@anchor{gnat_ugn/building_executable_programs_with_gnat using-gcc-for-syntax-checking}@anchor{f7}
14161 @subsection Using @code{gcc} for Syntax Checking
14164 @geindex -gnats (gcc)
14169 @item @code{-gnats}
14171 The @code{s} stands for ‘syntax’.
14173 Run GNAT in syntax checking only mode. For
14174 example, the command
14177 $ gcc -c -gnats x.adb
14180 compiles file @code{x.adb} in syntax-check-only mode. You can check a
14181 series of files in a single command
14182 , and can use wildcards to specify such a group of files.
14183 Note that you must specify the @code{-c} (compile
14184 only) flag in addition to the @code{-gnats} flag.
14186 You may use other switches in conjunction with @code{-gnats}. In
14187 particular, @code{-gnatl} and @code{-gnatv} are useful to control the
14188 format of any generated error messages.
14190 When the source file is empty or contains only empty lines and/or comments,
14191 the output is a warning:
14194 $ gcc -c -gnats -x ada toto.txt
14195 toto.txt:1:01: warning: empty file, contains no compilation units
14199 Otherwise, the output is simply the error messages, if any. No object file or
14200 ALI file is generated by a syntax-only compilation. Also, no units other
14201 than the one specified are accessed. For example, if a unit @code{X}
14202 @emph{with}s a unit @code{Y}, compiling unit @code{X} in syntax
14203 check only mode does not access the source file containing unit
14206 @geindex Multiple units
14207 @geindex syntax checking
14209 Normally, GNAT allows only a single unit in a source file. However, this
14210 restriction does not apply in syntax-check-only mode, and it is possible
14211 to check a file containing multiple compilation units concatenated
14212 together. This is primarily used by the @code{gnatchop} utility
14213 (@ref{1d,,Renaming Files with gnatchop}).
14216 @node Using gcc for Semantic Checking,Compiling Different Versions of Ada,Using gcc for Syntax Checking,Compiler Switches
14217 @anchor{gnat_ugn/building_executable_programs_with_gnat id21}@anchor{f8}@anchor{gnat_ugn/building_executable_programs_with_gnat using-gcc-for-semantic-checking}@anchor{f9}
14218 @subsection Using @code{gcc} for Semantic Checking
14221 @geindex -gnatc (gcc)
14226 @item @code{-gnatc}
14228 The @code{c} stands for ‘check’.
14229 Causes the compiler to operate in semantic check mode,
14230 with full checking for all illegalities specified in the
14231 Ada Reference Manual, but without generation of any object code
14232 (no object file is generated).
14234 Because dependent files must be accessed, you must follow the GNAT
14235 semantic restrictions on file structuring to operate in this mode:
14241 The needed source files must be accessible
14242 (see @ref{73,,Search Paths and the Run-Time Library (RTL)}).
14245 Each file must contain only one compilation unit.
14248 The file name and unit name must match (@ref{3b,,File Naming Rules}).
14251 The output consists of error messages as appropriate. No object file is
14252 generated. An @code{ALI} file is generated for use in the context of
14253 cross-reference tools, but this file is marked as not being suitable
14254 for binding (since no object file is generated).
14255 The checking corresponds exactly to the notion of
14256 legality in the Ada Reference Manual.
14258 Any unit can be compiled in semantics-checking-only mode, including
14259 units that would not normally be compiled (subunits,
14260 and specifications where a separate body is present).
14263 @node Compiling Different Versions of Ada,Character Set Control,Using gcc for Semantic Checking,Compiler Switches
14264 @anchor{gnat_ugn/building_executable_programs_with_gnat compiling-different-versions-of-ada}@anchor{6}@anchor{gnat_ugn/building_executable_programs_with_gnat id22}@anchor{fa}
14265 @subsection Compiling Different Versions of Ada
14268 The switches described in this section allow you to explicitly specify
14269 the version of the Ada language that your programs are written in.
14270 The default mode is Ada 2012,
14271 but you can also specify Ada 95, Ada 2005 mode, or
14272 indicate Ada 83 compatibility mode.
14274 @geindex Compatibility with Ada 83
14276 @geindex -gnat83 (gcc)
14279 @geindex Ada 83 tests
14281 @geindex Ada 83 mode
14286 @item @code{-gnat83} (Ada 83 Compatibility Mode)
14288 Although GNAT is primarily an Ada 95 / Ada 2005 compiler, this switch
14289 specifies that the program is to be compiled in Ada 83 mode. With
14290 @code{-gnat83}, GNAT rejects most post-Ada 83 extensions and applies Ada 83
14291 semantics where this can be done easily.
14292 It is not possible to guarantee this switch does a perfect
14293 job; some subtle tests, such as are
14294 found in earlier ACVC tests (and that have been removed from the ACATS suite
14295 for Ada 95), might not compile correctly.
14296 Nevertheless, this switch may be useful in some circumstances, for example
14297 where, due to contractual reasons, existing code needs to be maintained
14298 using only Ada 83 features.
14300 With few exceptions (most notably the need to use @code{<>} on
14302 @geindex Generic formal parameters
14303 generic formal parameters,
14304 the use of the new Ada 95 / Ada 2005
14305 reserved words, and the use of packages
14306 with optional bodies), it is not necessary to specify the
14307 @code{-gnat83} switch when compiling Ada 83 programs, because, with rare
14308 exceptions, Ada 95 and Ada 2005 are upwardly compatible with Ada 83. Thus
14309 a correct Ada 83 program is usually also a correct program
14310 in these later versions of the language standard. For further information
14311 please refer to the @emph{Compatibility and Porting Guide} chapter in the
14312 @cite{GNAT Reference Manual}.
14315 @geindex -gnat95 (gcc)
14317 @geindex Ada 95 mode
14322 @item @code{-gnat95} (Ada 95 mode)
14324 This switch directs the compiler to implement the Ada 95 version of the
14326 Since Ada 95 is almost completely upwards
14327 compatible with Ada 83, Ada 83 programs may generally be compiled using
14328 this switch (see the description of the @code{-gnat83} switch for further
14329 information about Ada 83 mode).
14330 If an Ada 2005 program is compiled in Ada 95 mode,
14331 uses of the new Ada 2005 features will cause error
14332 messages or warnings.
14334 This switch also can be used to cancel the effect of a previous
14335 @code{-gnat83}, @code{-gnat05/2005}, or @code{-gnat12/2012}
14336 switch earlier in the command line.
14339 @geindex -gnat05 (gcc)
14341 @geindex -gnat2005 (gcc)
14343 @geindex Ada 2005 mode
14348 @item @code{-gnat05} or @code{-gnat2005} (Ada 2005 mode)
14350 This switch directs the compiler to implement the Ada 2005 version of the
14351 language, as documented in the official Ada standards document.
14352 Since Ada 2005 is almost completely upwards
14353 compatible with Ada 95 (and thus also with Ada 83), Ada 83 and Ada 95 programs
14354 may generally be compiled using this switch (see the description of the
14355 @code{-gnat83} and @code{-gnat95} switches for further
14359 @geindex -gnat12 (gcc)
14361 @geindex -gnat2012 (gcc)
14363 @geindex Ada 2012 mode
14368 @item @code{-gnat12} or @code{-gnat2012} (Ada 2012 mode)
14370 This switch directs the compiler to implement the Ada 2012 version of the
14371 language (also the default).
14372 Since Ada 2012 is almost completely upwards
14373 compatible with Ada 2005 (and thus also with Ada 83, and Ada 95),
14374 Ada 83 and Ada 95 programs
14375 may generally be compiled using this switch (see the description of the
14376 @code{-gnat83}, @code{-gnat95}, and @code{-gnat05/2005} switches
14377 for further information).
14380 @geindex -gnat2022 (gcc)
14382 @geindex Ada 2022 mode
14387 @item @code{-gnat2022} (Ada 2022 mode)
14389 This switch directs the compiler to implement the Ada 2022 version of the
14393 @geindex -gnatX (gcc)
14395 @geindex Ada language extensions
14397 @geindex GNAT extensions
14402 @item @code{-gnatX} (Enable GNAT Extensions)
14404 This switch directs the compiler to implement the latest version of the
14405 language (currently Ada 2022) and also to enable certain GNAT implementation
14406 extensions that are not part of any Ada standard. For a full list of these
14407 extensions, see the GNAT reference manual, @code{Pragma Extensions_Allowed}.
14410 @node Character Set Control,File Naming Control,Compiling Different Versions of Ada,Compiler Switches
14411 @anchor{gnat_ugn/building_executable_programs_with_gnat character-set-control}@anchor{31}@anchor{gnat_ugn/building_executable_programs_with_gnat id23}@anchor{fb}
14412 @subsection Character Set Control
14415 @geindex -gnati (gcc)
14420 @item @code{-gnati@emph{c}}
14422 Normally GNAT recognizes the Latin-1 character set in source program
14423 identifiers, as described in the Ada Reference Manual.
14425 GNAT to recognize alternate character sets in identifiers. @code{c} is a
14426 single character indicating the character set, as follows:
14429 @multitable {xxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
14436 ISO 8859-1 (Latin-1) identifiers
14444 ISO 8859-2 (Latin-2) letters allowed in identifiers
14452 ISO 8859-3 (Latin-3) letters allowed in identifiers
14460 ISO 8859-4 (Latin-4) letters allowed in identifiers
14468 ISO 8859-5 (Cyrillic) letters allowed in identifiers
14476 ISO 8859-15 (Latin-9) letters allowed in identifiers
14484 IBM PC letters (code page 437) allowed in identifiers
14492 IBM PC letters (code page 850) allowed in identifiers
14500 Full upper-half codes allowed in identifiers
14508 No upper-half codes allowed in identifiers
14516 Wide-character codes (that is, codes greater than 255)
14517 allowed in identifiers
14522 See @ref{23,,Foreign Language Representation} for full details on the
14523 implementation of these character sets.
14526 @geindex -gnatW (gcc)
14531 @item @code{-gnatW@emph{e}}
14533 Specify the method of encoding for wide characters.
14534 @code{e} is one of the following:
14537 @multitable {xxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
14544 Hex encoding (brackets coding also recognized)
14552 Upper half encoding (brackets encoding also recognized)
14560 Shift/JIS encoding (brackets encoding also recognized)
14568 EUC encoding (brackets encoding also recognized)
14576 UTF-8 encoding (brackets encoding also recognized)
14584 Brackets encoding only (default value)
14589 For full details on these encoding
14590 methods see @ref{37,,Wide_Character Encodings}.
14591 Note that brackets coding is always accepted, even if one of the other
14592 options is specified, so for example @code{-gnatW8} specifies that both
14593 brackets and UTF-8 encodings will be recognized. The units that are
14594 with’ed directly or indirectly will be scanned using the specified
14595 representation scheme, and so if one of the non-brackets scheme is
14596 used, it must be used consistently throughout the program. However,
14597 since brackets encoding is always recognized, it may be conveniently
14598 used in standard libraries, allowing these libraries to be used with
14599 any of the available coding schemes.
14601 Note that brackets encoding only applies to program text. Within comments,
14602 brackets are considered to be normal graphic characters, and bracket sequences
14603 are never recognized as wide characters.
14605 If no @code{-gnatW?} parameter is present, then the default
14606 representation is normally Brackets encoding only. However, if the
14607 first three characters of the file are 16#EF# 16#BB# 16#BF# (the standard
14608 byte order mark or BOM for UTF-8), then these three characters are
14609 skipped and the default representation for the file is set to UTF-8.
14611 Note that the wide character representation that is specified (explicitly
14612 or by default) for the main program also acts as the default encoding used
14613 for Wide_Text_IO files if not specifically overridden by a WCEM form
14617 When no @code{-gnatW?} is specified, then characters (other than wide
14618 characters represented using brackets notation) are treated as 8-bit
14619 Latin-1 codes. The codes recognized are the Latin-1 graphic characters,
14620 and ASCII format effectors (CR, LF, HT, VT). Other lower half control
14621 characters in the range 16#00#..16#1F# are not accepted in program text
14622 or in comments. Upper half control characters (16#80#..16#9F#) are rejected
14623 in program text, but allowed and ignored in comments. Note in particular
14624 that the Next Line (NEL) character whose encoding is 16#85# is not recognized
14625 as an end of line in this default mode. If your source program contains
14626 instances of the NEL character used as a line terminator,
14627 you must use UTF-8 encoding for the whole
14628 source program. In default mode, all lines must be ended by a standard
14629 end of line sequence (CR, CR/LF, or LF).
14631 Note that the convention of simply accepting all upper half characters in
14632 comments means that programs that use standard ASCII for program text, but
14633 UTF-8 encoding for comments are accepted in default mode, providing that the
14634 comments are ended by an appropriate (CR, or CR/LF, or LF) line terminator.
14635 This is a common mode for many programs with foreign language comments.
14637 @node File Naming Control,Subprogram Inlining Control,Character Set Control,Compiler Switches
14638 @anchor{gnat_ugn/building_executable_programs_with_gnat file-naming-control}@anchor{fc}@anchor{gnat_ugn/building_executable_programs_with_gnat id24}@anchor{fd}
14639 @subsection File Naming Control
14642 @geindex -gnatk (gcc)
14647 @item @code{-gnatk@emph{n}}
14649 Activates file name ‘krunching’. @code{n}, a decimal integer in the range
14650 1-999, indicates the maximum allowable length of a file name (not
14651 including the @code{.ads} or @code{.adb} extension). The default is not
14652 to enable file name krunching.
14654 For the source file naming rules, @ref{3b,,File Naming Rules}.
14657 @node Subprogram Inlining Control,Auxiliary Output Control,File Naming Control,Compiler Switches
14658 @anchor{gnat_ugn/building_executable_programs_with_gnat id25}@anchor{fe}@anchor{gnat_ugn/building_executable_programs_with_gnat subprogram-inlining-control}@anchor{ff}
14659 @subsection Subprogram Inlining Control
14662 @geindex -gnatn (gcc)
14667 @item @code{-gnatn[12]}
14669 The @code{n} here is intended to suggest the first syllable of the word ‘inline’.
14670 GNAT recognizes and processes @code{Inline} pragmas. However, for inlining to
14671 actually occur, optimization must be enabled and, by default, inlining of
14672 subprograms across units is not performed. If you want to additionally
14673 enable inlining of subprograms specified by pragma @code{Inline} across units,
14674 you must also specify this switch.
14676 In the absence of this switch, GNAT does not attempt inlining across units
14677 and does not access the bodies of subprograms for which @code{pragma Inline} is
14678 specified if they are not in the current unit.
14680 You can optionally specify the inlining level: 1 for moderate inlining across
14681 units, which is a good compromise between compilation times and performances
14682 at run time, or 2 for full inlining across units, which may bring about
14683 longer compilation times. If no inlining level is specified, the compiler will
14684 pick it based on the optimization level: 1 for @code{-O1}, @code{-O2} or
14685 @code{-Os} and 2 for @code{-O3}.
14687 If you specify this switch the compiler will access these bodies,
14688 creating an extra source dependency for the resulting object file, and
14689 where possible, the call will be inlined.
14690 For further details on when inlining is possible
14691 see @ref{100,,Inlining of Subprograms}.
14694 @geindex -gnatN (gcc)
14699 @item @code{-gnatN}
14701 This switch activates front-end inlining which also
14702 generates additional dependencies.
14704 When using a gcc-based back end, then the use of
14705 @code{-gnatN} is deprecated, and the use of @code{-gnatn} is preferred.
14706 Historically front end inlining was more extensive than the gcc back end
14707 inlining, but that is no longer the case.
14710 @node Auxiliary Output Control,Debugging Control,Subprogram Inlining Control,Compiler Switches
14711 @anchor{gnat_ugn/building_executable_programs_with_gnat auxiliary-output-control}@anchor{101}@anchor{gnat_ugn/building_executable_programs_with_gnat id26}@anchor{102}
14712 @subsection Auxiliary Output Control
14715 @geindex -gnatu (gcc)
14720 @item @code{-gnatu}
14722 Print a list of units required by this compilation on @code{stdout}.
14723 The listing includes all units on which the unit being compiled depends
14724 either directly or indirectly.
14727 @geindex -pass-exit-codes (gcc)
14732 @item @code{-pass-exit-codes}
14734 If this switch is not used, the exit code returned by @code{gcc} when
14735 compiling multiple files indicates whether all source files have
14736 been successfully used to generate object files or not.
14738 When @code{-pass-exit-codes} is used, @code{gcc} exits with an extended
14739 exit status and allows an integrated development environment to better
14740 react to a compilation failure. Those exit status are:
14743 @multitable {xxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
14750 There was an error in at least one source file.
14758 At least one source file did not generate an object file.
14766 The compiler died unexpectedly (internal error for example).
14774 An object file has been generated for every source file.
14780 @node Debugging Control,Exception Handling Control,Auxiliary Output Control,Compiler Switches
14781 @anchor{gnat_ugn/building_executable_programs_with_gnat debugging-control}@anchor{103}@anchor{gnat_ugn/building_executable_programs_with_gnat id27}@anchor{104}
14782 @subsection Debugging Control
14787 @geindex Debugging options
14790 @geindex -gnatd (gcc)
14795 @item @code{-gnatd@emph{x}}
14797 Activate internal debugging switches. @code{x} is a letter or digit, or
14798 string of letters or digits, which specifies the type of debugging
14799 outputs desired. Normally these are used only for internal development
14800 or system debugging purposes. You can find full documentation for these
14801 switches in the body of the @code{Debug} unit in the compiler source
14802 file @code{debug.adb}.
14805 @geindex -gnatG (gcc)
14810 @item @code{-gnatG[=@emph{nn}]}
14812 This switch causes the compiler to generate auxiliary output containing
14813 a pseudo-source listing of the generated expanded code. Like most Ada
14814 compilers, GNAT works by first transforming the high level Ada code into
14815 lower level constructs. For example, tasking operations are transformed
14816 into calls to the tasking run-time routines. A unique capability of GNAT
14817 is to list this expanded code in a form very close to normal Ada source.
14818 This is very useful in understanding the implications of various Ada
14819 usage on the efficiency of the generated code. There are many cases in
14820 Ada (e.g., the use of controlled types), where simple Ada statements can
14821 generate a lot of run-time code. By using @code{-gnatG} you can identify
14822 these cases, and consider whether it may be desirable to modify the coding
14823 approach to improve efficiency.
14825 The optional parameter @code{nn} if present after -gnatG specifies an
14826 alternative maximum line length that overrides the normal default of 72.
14827 This value is in the range 40-999999, values less than 40 being silently
14828 reset to 40. The equal sign is optional.
14830 The format of the output is very similar to standard Ada source, and is
14831 easily understood by an Ada programmer. The following special syntactic
14832 additions correspond to low level features used in the generated code that
14833 do not have any exact analogies in pure Ada source form. The following
14834 is a partial list of these special constructions. See the spec
14835 of package @code{Sprint} in file @code{sprint.ads} for a full list.
14837 @geindex -gnatL (gcc)
14839 If the switch @code{-gnatL} is used in conjunction with
14840 @code{-gnatG}, then the original source lines are interspersed
14841 in the expanded source (as comment lines with the original line number).
14846 @item @code{new @emph{xxx} [storage_pool = @emph{yyy}]}
14848 Shows the storage pool being used for an allocator.
14850 @item @code{at end @emph{procedure-name};}
14852 Shows the finalization (cleanup) procedure for a scope.
14854 @item @code{(if @emph{expr} then @emph{expr} else @emph{expr})}
14856 Conditional expression equivalent to the @code{x?y:z} construction in C.
14858 @item @code{@emph{target}^(@emph{source})}
14860 A conversion with floating-point truncation instead of rounding.
14862 @item @code{@emph{target}?(@emph{source})}
14864 A conversion that bypasses normal Ada semantic checking. In particular
14865 enumeration types and fixed-point types are treated simply as integers.
14867 @item @code{@emph{target}?^(@emph{source})}
14869 Combines the above two cases.
14872 @code{@emph{x} #/ @emph{y}}
14874 @code{@emph{x} #mod @emph{y}}
14876 @code{@emph{x} # @emph{y}}
14881 @item @code{@emph{x} #rem @emph{y}}
14883 A division or multiplication of fixed-point values which are treated as
14884 integers without any kind of scaling.
14886 @item @code{free @emph{expr} [storage_pool = @emph{xxx}]}
14888 Shows the storage pool associated with a @code{free} statement.
14890 @item @code{[subtype or type declaration]}
14892 Used to list an equivalent declaration for an internally generated
14893 type that is referenced elsewhere in the listing.
14895 @item @code{freeze @emph{type-name} [@emph{actions}]}
14897 Shows the point at which @code{type-name} is frozen, with possible
14898 associated actions to be performed at the freeze point.
14900 @item @code{reference @emph{itype}}
14902 Reference (and hence definition) to internal type @code{itype}.
14904 @item @code{@emph{function-name}! (@emph{arg}, @emph{arg}, @emph{arg})}
14906 Intrinsic function call.
14908 @item @code{@emph{label-name} : label}
14910 Declaration of label @code{labelname}.
14912 @item @code{#$ @emph{subprogram-name}}
14914 An implicit call to a run-time support routine
14915 (to meet the requirement of H.3.1(9) in a
14916 convenient manner).
14918 @item @code{@emph{expr} && @emph{expr} && @emph{expr} ... && @emph{expr}}
14920 A multiple concatenation (same effect as @code{expr} & @code{expr} &
14921 @code{expr}, but handled more efficiently).
14923 @item @code{[constraint_error]}
14925 Raise the @code{Constraint_Error} exception.
14927 @item @code{@emph{expression}'reference}
14929 A pointer to the result of evaluating @{expression@}.
14931 @item @code{@emph{target-type}!(@emph{source-expression})}
14933 An unchecked conversion of @code{source-expression} to @code{target-type}.
14935 @item @code{[@emph{numerator}/@emph{denominator}]}
14937 Used to represent internal real literals (that) have no exact
14938 representation in base 2-16 (for example, the result of compile time
14939 evaluation of the expression 1.0/27.0).
14943 @geindex -gnatD (gcc)
14948 @item @code{-gnatD[=nn]}
14950 When used in conjunction with @code{-gnatG}, this switch causes
14951 the expanded source, as described above for
14952 @code{-gnatG} to be written to files with names
14953 @code{xxx.dg}, where @code{xxx} is the normal file name,
14954 instead of to the standard output file. For
14955 example, if the source file name is @code{hello.adb}, then a file
14956 @code{hello.adb.dg} will be written. The debugging
14957 information generated by the @code{gcc} @code{-g} switch
14958 will refer to the generated @code{xxx.dg} file. This allows
14959 you to do source level debugging using the generated code which is
14960 sometimes useful for complex code, for example to find out exactly
14961 which part of a complex construction raised an exception. This switch
14962 also suppresses generation of cross-reference information (see
14963 @code{-gnatx}) since otherwise the cross-reference information
14964 would refer to the @code{.dg} file, which would cause
14965 confusion since this is not the original source file.
14967 Note that @code{-gnatD} actually implies @code{-gnatG}
14968 automatically, so it is not necessary to give both options.
14969 In other words @code{-gnatD} is equivalent to @code{-gnatDG}).
14971 @geindex -gnatL (gcc)
14973 If the switch @code{-gnatL} is used in conjunction with
14974 @code{-gnatDG}, then the original source lines are interspersed
14975 in the expanded source (as comment lines with the original line number).
14977 The optional parameter @code{nn} if present after -gnatD specifies an
14978 alternative maximum line length that overrides the normal default of 72.
14979 This value is in the range 40-999999, values less than 40 being silently
14980 reset to 40. The equal sign is optional.
14983 @geindex -gnatr (gcc)
14985 @geindex pragma Restrictions
14990 @item @code{-gnatr}
14992 This switch causes pragma Restrictions to be treated as Restriction_Warnings
14993 so that violation of restrictions causes warnings rather than illegalities.
14994 This is useful during the development process when new restrictions are added
14995 or investigated. The switch also causes pragma Profile to be treated as
14996 Profile_Warnings, and pragma Restricted_Run_Time and pragma Ravenscar set
14997 restriction warnings rather than restrictions.
15000 @geindex -gnatR (gcc)
15005 @item @code{-gnatR[0|1|2|3|4][e][j][m][s]}
15007 This switch controls output from the compiler of a listing showing
15008 representation information for declared types, objects and subprograms.
15009 For @code{-gnatR0}, no information is output (equivalent to omitting
15010 the @code{-gnatR} switch). For @code{-gnatR1} (which is the default,
15011 so @code{-gnatR} with no parameter has the same effect), size and
15012 alignment information is listed for declared array and record types.
15014 For @code{-gnatR2}, size and alignment information is listed for all
15015 declared types and objects. The @code{Linker_Section} is also listed for any
15016 entity for which the @code{Linker_Section} is set explicitly or implicitly (the
15017 latter case occurs for objects of a type for which a @code{Linker_Section}
15020 For @code{-gnatR3}, symbolic expressions for values that are computed
15021 at run time for records are included. These symbolic expressions have
15022 a mostly obvious format with #n being used to represent the value of the
15023 n’th discriminant. See source files @code{repinfo.ads/adb} in the
15024 GNAT sources for full details on the format of @code{-gnatR3} output.
15026 For @code{-gnatR4}, information for relevant compiler-generated types
15027 is also listed, i.e. when they are structurally part of other declared
15030 If the switch is followed by an @code{e} (e.g. @code{-gnatR2e}), then
15031 extended representation information for record sub-components of records
15034 If the switch is followed by an @code{m} (e.g. @code{-gnatRm}), then
15035 subprogram conventions and parameter passing mechanisms for all the
15036 subprograms are included.
15038 If the switch is followed by a @code{j} (e.g., @code{-gnatRj}), then
15039 the output is in the JSON data interchange format specified by the
15040 ECMA-404 standard. The semantic description of this JSON output is
15041 available in the specification of the Repinfo unit present in the
15044 If the switch is followed by an @code{s} (e.g., @code{-gnatR3s}), then
15045 the output is to a file with the name @code{file.rep} where @code{file} is
15046 the name of the corresponding source file, except if @code{j} is also
15047 specified, in which case the file name is @code{file.json}.
15049 Note that it is possible for record components to have zero size. In
15050 this case, the component clause uses an obvious extension of permitted
15051 Ada syntax, for example @code{at 0 range 0 .. -1}.
15054 @geindex -gnatS (gcc)
15059 @item @code{-gnatS}
15061 The use of the switch @code{-gnatS} for an
15062 Ada compilation will cause the compiler to output a
15063 representation of package Standard in a form very
15064 close to standard Ada. It is not quite possible to
15065 do this entirely in standard Ada (since new
15066 numeric base types cannot be created in standard
15067 Ada), but the output is easily
15068 readable to any Ada programmer, and is useful to
15069 determine the characteristics of target dependent
15070 types in package Standard.
15073 @geindex -gnatx (gcc)
15078 @item @code{-gnatx}
15080 Normally the compiler generates full cross-referencing information in
15081 the @code{ALI} file. This information is used by a number of tools,
15082 including @code{gnatfind} and @code{gnatxref}. The @code{-gnatx} switch
15083 suppresses this information. This saves some space and may slightly
15084 speed up compilation, but means that these tools cannot be used.
15087 @geindex -fgnat-encodings (gcc)
15092 @item @code{-fgnat-encodings=[all|gdb|minimal]}
15094 This switch controls the balance between GNAT encodings and standard DWARF
15095 emitted in the debug information.
15097 Historically, old debug formats like stabs were not powerful enough to
15098 express some Ada types (for instance, variant records or fixed-point types).
15099 To work around this, GNAT introduced proprietary encodings that embed the
15100 missing information (“GNAT encodings”).
15102 Recent versions of the DWARF debug information format are now able to
15103 correctly describe most of these Ada constructs (“standard DWARF”). As
15104 third-party tools started to use this format, GNAT has been enhanced to
15105 generate it. However, most tools (including GDB) are still relying on GNAT
15108 To support all tools, GNAT needs to be versatile about the balance between
15109 generation of GNAT encodings and standard DWARF. This is what
15110 @code{-fgnat-encodings} is about.
15116 @code{=all}: Emit all GNAT encodings, and then emit as much standard DWARF as
15117 possible so it does not conflict with GNAT encodings.
15120 @code{=gdb}: Emit as much standard DWARF as possible as long as the current
15121 GDB handles it. Emit GNAT encodings for the rest.
15124 @code{=minimal}: Emit as much standard DWARF as possible and emit GNAT
15125 encodings for the rest.
15129 @node Exception Handling Control,Units to Sources Mapping Files,Debugging Control,Compiler Switches
15130 @anchor{gnat_ugn/building_executable_programs_with_gnat exception-handling-control}@anchor{105}@anchor{gnat_ugn/building_executable_programs_with_gnat id28}@anchor{106}
15131 @subsection Exception Handling Control
15134 GNAT uses two methods for handling exceptions at run time. The
15135 @code{setjmp/longjmp} method saves the context when entering
15136 a frame with an exception handler. Then when an exception is
15137 raised, the context can be restored immediately, without the
15138 need for tracing stack frames. This method provides very fast
15139 exception propagation, but introduces significant overhead for
15140 the use of exception handlers, even if no exception is raised.
15142 The other approach is called ‘zero cost’ exception handling.
15143 With this method, the compiler builds static tables to describe
15144 the exception ranges. No dynamic code is required when entering
15145 a frame containing an exception handler. When an exception is
15146 raised, the tables are used to control a back trace of the
15147 subprogram invocation stack to locate the required exception
15148 handler. This method has considerably poorer performance for
15149 the propagation of exceptions, but there is no overhead for
15150 exception handlers if no exception is raised. Note that in this
15151 mode and in the context of mixed Ada and C/C++ programming,
15152 to propagate an exception through a C/C++ code, the C/C++ code
15153 must be compiled with the @code{-funwind-tables} GCC’s
15156 The following switches may be used to control which of the
15157 two exception handling methods is used.
15159 @geindex --RTS=sjlj (gnatmake)
15164 @item @code{--RTS=sjlj}
15166 This switch causes the setjmp/longjmp run-time (when available) to be used
15167 for exception handling. If the default
15168 mechanism for the target is zero cost exceptions, then
15169 this switch can be used to modify this default, and must be
15170 used for all units in the partition.
15171 This option is rarely used. One case in which it may be
15172 advantageous is if you have an application where exception
15173 raising is common and the overall performance of the
15174 application is improved by favoring exception propagation.
15177 @geindex --RTS=zcx (gnatmake)
15179 @geindex Zero Cost Exceptions
15184 @item @code{--RTS=zcx}
15186 This switch causes the zero cost approach to be used
15187 for exception handling. If this is the default mechanism for the
15188 target (see below), then this switch is unneeded. If the default
15189 mechanism for the target is setjmp/longjmp exceptions, then
15190 this switch can be used to modify this default, and must be
15191 used for all units in the partition.
15192 This option can only be used if the zero cost approach
15193 is available for the target in use, otherwise it will generate an error.
15196 The same option @code{--RTS} must be used both for @code{gcc}
15197 and @code{gnatbind}. Passing this option to @code{gnatmake}
15198 (@ref{ce,,Switches for gnatmake}) will ensure the required consistency
15199 through the compilation and binding steps.
15201 @node Units to Sources Mapping Files,Code Generation Control,Exception Handling Control,Compiler Switches
15202 @anchor{gnat_ugn/building_executable_programs_with_gnat id29}@anchor{107}@anchor{gnat_ugn/building_executable_programs_with_gnat units-to-sources-mapping-files}@anchor{e8}
15203 @subsection Units to Sources Mapping Files
15206 @geindex -gnatem (gcc)
15211 @item @code{-gnatem=@emph{path}}
15213 A mapping file is a way to communicate to the compiler two mappings:
15214 from unit names to file names (without any directory information) and from
15215 file names to path names (with full directory information). These mappings
15216 are used by the compiler to short-circuit the path search.
15218 The use of mapping files is not required for correct operation of the
15219 compiler, but mapping files can improve efficiency, particularly when
15220 sources are read over a slow network connection. In normal operation,
15221 you need not be concerned with the format or use of mapping files,
15222 and the @code{-gnatem} switch is not a switch that you would use
15223 explicitly. It is intended primarily for use by automatic tools such as
15224 @code{gnatmake} running under the project file facility. The
15225 description here of the format of mapping files is provided
15226 for completeness and for possible use by other tools.
15228 A mapping file is a sequence of sets of three lines. In each set, the
15229 first line is the unit name, in lower case, with @code{%s} appended
15230 for specs and @code{%b} appended for bodies; the second line is the
15231 file name; and the third line is the path name.
15238 /gnat/project1/sources/main.2.ada
15241 When the switch @code{-gnatem} is specified, the compiler will
15242 create in memory the two mappings from the specified file. If there is
15243 any problem (nonexistent file, truncated file or duplicate entries),
15244 no mapping will be created.
15246 Several @code{-gnatem} switches may be specified; however, only the
15247 last one on the command line will be taken into account.
15249 When using a project file, @code{gnatmake} creates a temporary
15250 mapping file and communicates it to the compiler using this switch.
15253 @node Code Generation Control,,Units to Sources Mapping Files,Compiler Switches
15254 @anchor{gnat_ugn/building_executable_programs_with_gnat code-generation-control}@anchor{108}@anchor{gnat_ugn/building_executable_programs_with_gnat id30}@anchor{109}
15255 @subsection Code Generation Control
15258 The GCC technology provides a wide range of target dependent
15259 @code{-m} switches for controlling
15260 details of code generation with respect to different versions of
15261 architectures. This includes variations in instruction sets (e.g.,
15262 different members of the power pc family), and different requirements
15263 for optimal arrangement of instructions (e.g., different members of
15264 the x86 family). The list of available @code{-m} switches may be
15265 found in the GCC documentation.
15267 Use of these @code{-m} switches may in some cases result in improved
15270 The GNAT technology is tested and qualified without any
15271 @code{-m} switches,
15272 so generally the most reliable approach is to avoid the use of these
15273 switches. However, we generally expect most of these switches to work
15274 successfully with GNAT, and many customers have reported successful
15275 use of these options.
15277 Our general advice is to avoid the use of @code{-m} switches unless
15278 special needs lead to requirements in this area. In particular,
15279 there is no point in using @code{-m} switches to improve performance
15280 unless you actually see a performance improvement.
15282 @node Linker Switches,Binding with gnatbind,Compiler Switches,Building Executable Programs with GNAT
15283 @anchor{gnat_ugn/building_executable_programs_with_gnat id31}@anchor{10a}@anchor{gnat_ugn/building_executable_programs_with_gnat linker-switches}@anchor{10b}
15284 @section Linker Switches
15287 Linker switches can be specified after @code{-largs} builder switch.
15289 @geindex -fuse-ld=name
15294 @item @code{-fuse-ld=@emph{name}}
15296 Linker to be used. The default is @code{bfd} for @code{ld.bfd},
15297 the alternative being @code{gold} for @code{ld.gold}. The later is
15298 a more recent and faster linker, but only available on GNU/Linux
15302 @node Binding with gnatbind,Linking with gnatlink,Linker Switches,Building Executable Programs with GNAT
15303 @anchor{gnat_ugn/building_executable_programs_with_gnat binding-with-gnatbind}@anchor{c8}@anchor{gnat_ugn/building_executable_programs_with_gnat id32}@anchor{10c}
15304 @section Binding with @code{gnatbind}
15309 This chapter describes the GNAT binder, @code{gnatbind}, which is used
15310 to bind compiled GNAT objects.
15312 The @code{gnatbind} program performs four separate functions:
15318 Checks that a program is consistent, in accordance with the rules in
15319 Chapter 10 of the Ada Reference Manual. In particular, error
15320 messages are generated if a program uses inconsistent versions of a
15324 Checks that an acceptable order of elaboration exists for the program
15325 and issues an error message if it cannot find an order of elaboration
15326 that satisfies the rules in Chapter 10 of the Ada Language Manual.
15329 Generates a main program incorporating the given elaboration order.
15330 This program is a small Ada package (body and spec) that
15331 must be subsequently compiled
15332 using the GNAT compiler. The necessary compilation step is usually
15333 performed automatically by @code{gnatlink}. The two most important
15334 functions of this program
15335 are to call the elaboration routines of units in an appropriate order
15336 and to call the main program.
15339 Determines the set of object files required by the given main program.
15340 This information is output in the forms of comments in the generated program,
15341 to be read by the @code{gnatlink} utility used to link the Ada application.
15345 * Running gnatbind::
15346 * Switches for gnatbind::
15347 * Command-Line Access::
15348 * Search Paths for gnatbind::
15349 * Examples of gnatbind Usage::
15353 @node Running gnatbind,Switches for gnatbind,,Binding with gnatbind
15354 @anchor{gnat_ugn/building_executable_programs_with_gnat id33}@anchor{10d}@anchor{gnat_ugn/building_executable_programs_with_gnat running-gnatbind}@anchor{10e}
15355 @subsection Running @code{gnatbind}
15358 The form of the @code{gnatbind} command is
15361 $ gnatbind [ switches ] mainprog[.ali] [ switches ]
15364 where @code{mainprog.adb} is the Ada file containing the main program
15365 unit body. @code{gnatbind} constructs an Ada
15366 package in two files whose names are
15367 @code{b~mainprog.ads}, and @code{b~mainprog.adb}.
15368 For example, if given the
15369 parameter @code{hello.ali}, for a main program contained in file
15370 @code{hello.adb}, the binder output files would be @code{b~hello.ads}
15371 and @code{b~hello.adb}.
15373 When doing consistency checking, the binder takes into consideration
15374 any source files it can locate. For example, if the binder determines
15375 that the given main program requires the package @code{Pack}, whose
15377 file is @code{pack.ali} and whose corresponding source spec file is
15378 @code{pack.ads}, it attempts to locate the source file @code{pack.ads}
15379 (using the same search path conventions as previously described for the
15380 @code{gcc} command). If it can locate this source file, it checks that
15382 or source checksums of the source and its references to in @code{ALI} files
15383 match. In other words, any @code{ALI} files that mentions this spec must have
15384 resulted from compiling this version of the source file (or in the case
15385 where the source checksums match, a version close enough that the
15386 difference does not matter).
15388 @geindex Source files
15389 @geindex use by binder
15391 The effect of this consistency checking, which includes source files, is
15392 that the binder ensures that the program is consistent with the latest
15393 version of the source files that can be located at bind time. Editing a
15394 source file without compiling files that depend on the source file cause
15395 error messages to be generated by the binder.
15397 For example, suppose you have a main program @code{hello.adb} and a
15398 package @code{P}, from file @code{p.ads} and you perform the following
15405 Enter @code{gcc -c hello.adb} to compile the main program.
15408 Enter @code{gcc -c p.ads} to compile package @code{P}.
15411 Edit file @code{p.ads}.
15414 Enter @code{gnatbind hello}.
15417 At this point, the file @code{p.ali} contains an out-of-date time stamp
15418 because the file @code{p.ads} has been edited. The attempt at binding
15419 fails, and the binder generates the following error messages:
15422 error: "hello.adb" must be recompiled ("p.ads" has been modified)
15423 error: "p.ads" has been modified and must be recompiled
15426 Now both files must be recompiled as indicated, and then the bind can
15427 succeed, generating a main program. You need not normally be concerned
15428 with the contents of this file, but for reference purposes a sample
15429 binder output file is given in @ref{e,,Example of Binder Output File}.
15431 In most normal usage, the default mode of @code{gnatbind} which is to
15432 generate the main package in Ada, as described in the previous section.
15433 In particular, this means that any Ada programmer can read and understand
15434 the generated main program. It can also be debugged just like any other
15435 Ada code provided the @code{-g} switch is used for
15436 @code{gnatbind} and @code{gnatlink}.
15438 @node Switches for gnatbind,Command-Line Access,Running gnatbind,Binding with gnatbind
15439 @anchor{gnat_ugn/building_executable_programs_with_gnat id34}@anchor{10f}@anchor{gnat_ugn/building_executable_programs_with_gnat switches-for-gnatbind}@anchor{110}
15440 @subsection Switches for @code{gnatbind}
15443 The following switches are available with @code{gnatbind}; details will
15444 be presented in subsequent sections.
15446 @geindex --version (gnatbind)
15451 @item @code{--version}
15453 Display Copyright and version, then exit disregarding all other options.
15456 @geindex --help (gnatbind)
15461 @item @code{--help}
15463 If @code{--version} was not used, display usage, then exit disregarding
15467 @geindex -a (gnatbind)
15474 Indicates that, if supported by the platform, the adainit procedure should
15475 be treated as an initialisation routine by the linker (a constructor). This
15476 is intended to be used by the Project Manager to automatically initialize
15477 shared Stand-Alone Libraries.
15480 @geindex -aO (gnatbind)
15487 Specify directory to be searched for ALI files.
15490 @geindex -aI (gnatbind)
15497 Specify directory to be searched for source file.
15500 @geindex -A (gnatbind)
15505 @item @code{-A[=@emph{filename}]}
15507 Output ALI list (to standard output or to the named file).
15510 @geindex -b (gnatbind)
15517 Generate brief messages to @code{stderr} even if verbose mode set.
15520 @geindex -c (gnatbind)
15527 Check only, no generation of binder output file.
15530 @geindex -dnn[k|m] (gnatbind)
15535 @item @code{-d@emph{nn}[k|m]}
15537 This switch can be used to change the default task stack size value
15538 to a specified size @code{nn}, which is expressed in bytes by default, or
15539 in kilobytes when suffixed with @code{k} or in megabytes when suffixed
15541 In the absence of a @code{[k|m]} suffix, this switch is equivalent,
15542 in effect, to completing all task specs with
15545 pragma Storage_Size (nn);
15548 When they do not already have such a pragma.
15551 @geindex -D (gnatbind)
15556 @item @code{-D@emph{nn}[k|m]}
15558 Set the default secondary stack size to @code{nn}. The suffix indicates whether
15559 the size is in bytes (no suffix), kilobytes (@code{k} suffix) or megabytes
15562 The secondary stack holds objects of unconstrained types that are returned by
15563 functions, for example unconstrained Strings. The size of the secondary stack
15564 can be dynamic or fixed depending on the target.
15566 For most targets, the secondary stack grows on demand and is implemented as
15567 a chain of blocks in the heap. In this case, the default secondary stack size
15568 determines the initial size of the secondary stack for each task and the
15569 smallest amount the secondary stack can grow by.
15571 For Ravenscar, ZFP, and Cert run-times the size of the secondary stack is
15572 fixed. This switch can be used to change the default size of these stacks.
15573 The default secondary stack size can be overridden on a per-task basis if
15574 individual tasks have different secondary stack requirements. This is
15575 achieved through the Secondary_Stack_Size aspect that takes the size of the
15576 secondary stack in bytes.
15579 @geindex -e (gnatbind)
15586 Output complete list of elaboration-order dependencies.
15589 @geindex -Ea (gnatbind)
15596 Store tracebacks in exception occurrences when the target supports it.
15597 The “a” is for “address”; tracebacks will contain hexadecimal addresses,
15598 unless symbolic tracebacks are enabled.
15600 See also the packages @code{GNAT.Traceback} and
15601 @code{GNAT.Traceback.Symbolic} for more information.
15602 Note that on x86 ports, you must not use @code{-fomit-frame-pointer}
15606 @geindex -Es (gnatbind)
15613 Store tracebacks in exception occurrences when the target supports it.
15614 The “s” is for “symbolic”; symbolic tracebacks are enabled.
15617 @geindex -E (gnatbind)
15624 Currently the same as @code{-Ea}.
15627 @geindex -f (gnatbind)
15632 @item @code{-f@emph{elab-order}}
15634 Force elaboration order. For further details see @ref{111,,Elaboration Control}
15635 and @ref{f,,Elaboration Order Handling in GNAT}.
15638 @geindex -F (gnatbind)
15645 Force the checks of elaboration flags. @code{gnatbind} does not normally
15646 generate checks of elaboration flags for the main executable, except when
15647 a Stand-Alone Library is used. However, there are cases when this cannot be
15648 detected by gnatbind. An example is importing an interface of a Stand-Alone
15649 Library through a pragma Import and only specifying through a linker switch
15650 this Stand-Alone Library. This switch is used to guarantee that elaboration
15651 flag checks are generated.
15654 @geindex -h (gnatbind)
15661 Output usage (help) information.
15664 @geindex -H (gnatbind)
15671 Legacy elaboration order model enabled. For further details see
15672 @ref{f,,Elaboration Order Handling in GNAT}.
15675 @geindex -H32 (gnatbind)
15682 Use 32-bit allocations for @code{__gnat_malloc} (and thus for access types).
15683 For further details see @ref{112,,Dynamic Allocation Control}.
15686 @geindex -H64 (gnatbind)
15688 @geindex __gnat_malloc
15695 Use 64-bit allocations for @code{__gnat_malloc} (and thus for access types).
15696 For further details see @ref{112,,Dynamic Allocation Control}.
15698 @geindex -I (gnatbind)
15702 Specify directory to be searched for source and ALI files.
15704 @geindex -I- (gnatbind)
15708 Do not look for sources in the current directory where @code{gnatbind} was
15709 invoked, and do not look for ALI files in the directory containing the
15710 ALI file named in the @code{gnatbind} command line.
15712 @geindex -l (gnatbind)
15716 Output chosen elaboration order.
15718 @geindex -L (gnatbind)
15720 @item @code{-L@emph{xxx}}
15722 Bind the units for library building. In this case the @code{adainit} and
15723 @code{adafinal} procedures (@ref{a0,,Binding with Non-Ada Main Programs})
15724 are renamed to @code{@emph{xxx}init} and
15725 @code{@emph{xxx}final}.
15727 (@ref{2a,,GNAT and Libraries}, for more details.)
15729 @geindex -M (gnatbind)
15731 @item @code{-M@emph{xyz}}
15733 Rename generated main program from main to xyz. This option is
15734 supported on cross environments only.
15736 @geindex -m (gnatbind)
15738 @item @code{-m@emph{n}}
15740 Limit number of detected errors or warnings to @code{n}, where @code{n} is
15741 in the range 1..999999. The default value if no switch is
15742 given is 9999. If the number of warnings reaches this limit, then a
15743 message is output and further warnings are suppressed, the bind
15744 continues in this case. If the number of errors reaches this
15745 limit, then a message is output and the bind is abandoned.
15746 A value of zero means that no limit is enforced. The equal
15749 @geindex -minimal (gnatbind)
15751 @item @code{-minimal}
15753 Generate a binder file suitable for space-constrained applications. When
15754 active, binder-generated objects not required for program operation are no
15755 longer generated. @strong{Warning:} this option comes with the following
15762 Starting the program’s execution in the debugger will cause it to
15763 stop at the start of the @code{main} function instead of the main subprogram.
15764 This can be worked around by manually inserting a breakpoint on that
15765 subprogram and resuming the program’s execution until reaching that breakpoint.
15768 Programs using GNAT.Compiler_Version will not link.
15771 @geindex -n (gnatbind)
15777 @geindex -nostdinc (gnatbind)
15779 @item @code{-nostdinc}
15781 Do not look for sources in the system default directory.
15783 @geindex -nostdlib (gnatbind)
15785 @item @code{-nostdlib}
15787 Do not look for library files in the system default directory.
15789 @geindex --RTS (gnatbind)
15791 @item @code{--RTS=@emph{rts-path}}
15793 Specifies the default location of the run-time library. Same meaning as the
15794 equivalent @code{gnatmake} flag (@ref{ce,,Switches for gnatmake}).
15796 @geindex -o (gnatbind)
15798 @item @code{-o @emph{file}}
15800 Name the output file @code{file} (default is @code{b~`xxx}.adb`).
15801 Note that if this option is used, then linking must be done manually,
15802 gnatlink cannot be used.
15804 @geindex -O (gnatbind)
15806 @item @code{-O[=@emph{filename}]}
15808 Output object list (to standard output or to the named file).
15810 @geindex -p (gnatbind)
15814 Pessimistic (worst-case) elaboration order.
15816 @geindex -P (gnatbind)
15820 Generate binder file suitable for CodePeer.
15822 @geindex -R (gnatbind)
15826 Output closure source list, which includes all non-run-time units that are
15827 included in the bind.
15829 @geindex -Ra (gnatbind)
15833 Like @code{-R} but the list includes run-time units.
15835 @geindex -s (gnatbind)
15839 Require all source files to be present.
15841 @geindex -S (gnatbind)
15843 @item @code{-S@emph{xxx}}
15845 Specifies the value to be used when detecting uninitialized scalar
15846 objects with pragma Initialize_Scalars.
15847 The @code{xxx} string specified with the switch is one of:
15853 @code{in} for an invalid value.
15855 If zero is invalid for the discrete type in question,
15856 then the scalar value is set to all zero bits.
15857 For signed discrete types, the largest possible negative value of
15858 the underlying scalar is set (i.e. a one bit followed by all zero bits).
15859 For unsigned discrete types, the underlying scalar value is set to all
15860 one bits. For floating-point types, a NaN value is set
15861 (see body of package System.Scalar_Values for exact values).
15864 @code{lo} for low value.
15866 If zero is invalid for the discrete type in question,
15867 then the scalar value is set to all zero bits.
15868 For signed discrete types, the largest possible negative value of
15869 the underlying scalar is set (i.e. a one bit followed by all zero bits).
15870 For unsigned discrete types, the underlying scalar value is set to all
15871 zero bits. For floating-point, a small value is set
15872 (see body of package System.Scalar_Values for exact values).
15875 @code{hi} for high value.
15877 If zero is invalid for the discrete type in question,
15878 then the scalar value is set to all one bits.
15879 For signed discrete types, the largest possible positive value of
15880 the underlying scalar is set (i.e. a zero bit followed by all one bits).
15881 For unsigned discrete types, the underlying scalar value is set to all
15882 one bits. For floating-point, a large value is set
15883 (see body of package System.Scalar_Values for exact values).
15886 @code{xx} for hex value (two hex digits).
15888 The underlying scalar is set to a value consisting of repeated bytes, whose
15889 value corresponds to the given value. For example if @code{BF} is given,
15890 then a 32-bit scalar value will be set to the bit patterm @code{16#BFBFBFBF#}.
15893 @geindex GNAT_INIT_SCALARS
15895 In addition, you can specify @code{-Sev} to indicate that the value is
15896 to be set at run time. In this case, the program will look for an environment
15897 variable of the form @code{GNAT_INIT_SCALARS=@emph{yy}}, where @code{yy} is one
15898 of @code{in/lo/hi/@emph{xx}} with the same meanings as above.
15899 If no environment variable is found, or if it does not have a valid value,
15900 then the default is @code{in} (invalid values).
15903 @geindex -static (gnatbind)
15908 @item @code{-static}
15910 Link against a static GNAT run-time.
15912 @geindex -shared (gnatbind)
15914 @item @code{-shared}
15916 Link against a shared GNAT run-time when available.
15918 @geindex -t (gnatbind)
15922 Tolerate time stamp and other consistency errors.
15924 @geindex -T (gnatbind)
15926 @item @code{-T@emph{n}}
15928 Set the time slice value to @code{n} milliseconds. If the system supports
15929 the specification of a specific time slice value, then the indicated value
15930 is used. If the system does not support specific time slice values, but
15931 does support some general notion of round-robin scheduling, then any
15932 nonzero value will activate round-robin scheduling.
15934 A value of zero is treated specially. It turns off time
15935 slicing, and in addition, indicates to the tasking run-time that the
15936 semantics should match as closely as possible the Annex D
15937 requirements of the Ada RM, and in particular sets the default
15938 scheduling policy to @code{FIFO_Within_Priorities}.
15940 @geindex -u (gnatbind)
15942 @item @code{-u@emph{n}}
15944 Enable dynamic stack usage, with @code{n} results stored and displayed
15945 at program termination. A result is generated when a task
15946 terminates. Results that can’t be stored are displayed on the fly, at
15947 task termination. This option is currently not supported on Itanium
15948 platforms. (See @ref{113,,Dynamic Stack Usage Analysis} for details.)
15950 @geindex -v (gnatbind)
15954 Verbose mode. Write error messages, header, summary output to
15957 @geindex -V (gnatbind)
15959 @item @code{-V@emph{key}=@emph{value}}
15961 Store the given association of @code{key} to @code{value} in the bind environment.
15962 Values stored this way can be retrieved at run time using
15963 @code{GNAT.Bind_Environment}.
15965 @geindex -w (gnatbind)
15967 @item @code{-w@emph{x}}
15969 Warning mode; @code{x} = s/e for suppress/treat as error.
15971 @geindex -Wx (gnatbind)
15973 @item @code{-Wx@emph{e}}
15975 Override default wide character encoding for standard Text_IO files.
15977 @geindex -x (gnatbind)
15981 Exclude source files (check object consistency only).
15983 @geindex -xdr (gnatbind)
15987 Use the target-independent XDR protocol for stream oriented attributes
15988 instead of the default implementation which is based on direct binary
15989 representations and is therefore target-and endianness-dependent.
15990 However it does not support 128-bit integer types and the exception
15991 @code{Ada.IO_Exceptions.Device_Error} is raised if any attempt is made
15992 at streaming 128-bit integer types with it.
15994 @geindex -Xnnn (gnatbind)
15996 @item @code{-X@emph{nnn}}
15998 Set default exit status value, normally 0 for POSIX compliance.
16000 @geindex -y (gnatbind)
16004 Enable leap seconds support in @code{Ada.Calendar} and its children.
16006 @geindex -z (gnatbind)
16010 No main subprogram.
16013 You may obtain this listing of switches by running @code{gnatbind} with
16017 * Consistency-Checking Modes::
16018 * Binder Error Message Control::
16019 * Elaboration Control::
16021 * Dynamic Allocation Control::
16022 * Binding with Non-Ada Main Programs::
16023 * Binding Programs with No Main Subprogram::
16027 @node Consistency-Checking Modes,Binder Error Message Control,,Switches for gnatbind
16028 @anchor{gnat_ugn/building_executable_programs_with_gnat consistency-checking-modes}@anchor{114}@anchor{gnat_ugn/building_executable_programs_with_gnat id35}@anchor{115}
16029 @subsubsection Consistency-Checking Modes
16032 As described earlier, by default @code{gnatbind} checks
16033 that object files are consistent with one another and are consistent
16034 with any source files it can locate. The following switches control binder
16039 @geindex -s (gnatbind)
16047 Require source files to be present. In this mode, the binder must be
16048 able to locate all source files that are referenced, in order to check
16049 their consistency. In normal mode, if a source file cannot be located it
16050 is simply ignored. If you specify this switch, a missing source
16053 @geindex -Wx (gnatbind)
16055 @item @code{-Wx@emph{e}}
16057 Override default wide character encoding for standard Text_IO files.
16058 Normally the default wide character encoding method used for standard
16059 [Wide_[Wide_]]Text_IO files is taken from the encoding specified for
16060 the main source input (see description of switch
16061 @code{-gnatWx} for the compiler). The
16062 use of this switch for the binder (which has the same set of
16063 possible arguments) overrides this default as specified.
16065 @geindex -x (gnatbind)
16069 Exclude source files. In this mode, the binder only checks that ALI
16070 files are consistent with one another. Source files are not accessed.
16071 The binder runs faster in this mode, and there is still a guarantee that
16072 the resulting program is self-consistent.
16073 If a source file has been edited since it was last compiled, and you
16074 specify this switch, the binder will not detect that the object
16075 file is out of date with respect to the source file. Note that this is the
16076 mode that is automatically used by @code{gnatmake} because in this
16077 case the checking against sources has already been performed by
16078 @code{gnatmake} in the course of compilation (i.e., before binding).
16081 @node Binder Error Message Control,Elaboration Control,Consistency-Checking Modes,Switches for gnatbind
16082 @anchor{gnat_ugn/building_executable_programs_with_gnat binder-error-message-control}@anchor{116}@anchor{gnat_ugn/building_executable_programs_with_gnat id36}@anchor{117}
16083 @subsubsection Binder Error Message Control
16086 The following switches provide control over the generation of error
16087 messages from the binder:
16091 @geindex -v (gnatbind)
16099 Verbose mode. In the normal mode, brief error messages are generated to
16100 @code{stderr}. If this switch is present, a header is written
16101 to @code{stdout} and any error messages are directed to @code{stdout}.
16102 All that is written to @code{stderr} is a brief summary message.
16104 @geindex -b (gnatbind)
16108 Generate brief error messages to @code{stderr} even if verbose mode is
16109 specified. This is relevant only when used with the
16112 @geindex -m (gnatbind)
16114 @item @code{-m@emph{n}}
16116 Limits the number of error messages to @code{n}, a decimal integer in the
16117 range 1-999. The binder terminates immediately if this limit is reached.
16119 @geindex -M (gnatbind)
16121 @item @code{-M@emph{xxx}}
16123 Renames the generated main program from @code{main} to @code{xxx}.
16124 This is useful in the case of some cross-building environments, where
16125 the actual main program is separate from the one generated
16126 by @code{gnatbind}.
16128 @geindex -ws (gnatbind)
16134 Suppress all warning messages.
16136 @geindex -we (gnatbind)
16140 Treat any warning messages as fatal errors.
16142 @geindex -t (gnatbind)
16144 @geindex Time stamp checks
16147 @geindex Binder consistency checks
16149 @geindex Consistency checks
16154 The binder performs a number of consistency checks including:
16160 Check that time stamps of a given source unit are consistent
16163 Check that checksums of a given source unit are consistent
16166 Check that consistent versions of @code{GNAT} were used for compilation
16169 Check consistency of configuration pragmas as required
16172 Normally failure of such checks, in accordance with the consistency
16173 requirements of the Ada Reference Manual, causes error messages to be
16174 generated which abort the binder and prevent the output of a binder
16175 file and subsequent link to obtain an executable.
16177 The @code{-t} switch converts these error messages
16178 into warnings, so that
16179 binding and linking can continue to completion even in the presence of such
16180 errors. The result may be a failed link (due to missing symbols), or a
16181 non-functional executable which has undefined semantics.
16185 This means that @code{-t} should be used only in unusual situations,
16191 @node Elaboration Control,Output Control,Binder Error Message Control,Switches for gnatbind
16192 @anchor{gnat_ugn/building_executable_programs_with_gnat elaboration-control}@anchor{111}@anchor{gnat_ugn/building_executable_programs_with_gnat id37}@anchor{118}
16193 @subsubsection Elaboration Control
16196 The following switches provide additional control over the elaboration
16197 order. For further details see @ref{f,,Elaboration Order Handling in GNAT}.
16199 @geindex -f (gnatbind)
16204 @item @code{-f@emph{elab-order}}
16206 Force elaboration order.
16208 @code{elab-order} should be the name of a “forced elaboration order file”, that
16209 is, a text file containing library item names, one per line. A name of the
16210 form “some.unit%s” or “some.unit (spec)” denotes the spec of Some.Unit. A
16211 name of the form “some.unit%b” or “some.unit (body)” denotes the body of
16212 Some.Unit. Each pair of lines is taken to mean that there is an elaboration
16213 dependence of the second line on the first. For example, if the file
16223 then the spec of This will be elaborated before the body of This, and the
16224 body of This will be elaborated before the spec of That, and the spec of That
16225 will be elaborated before the body of That. The first and last of these three
16226 dependences are already required by Ada rules, so this file is really just
16227 forcing the body of This to be elaborated before the spec of That.
16229 The given order must be consistent with Ada rules, or else @code{gnatbind} will
16230 give elaboration cycle errors. For example, if you say x (body) should be
16231 elaborated before x (spec), there will be a cycle, because Ada rules require
16232 x (spec) to be elaborated before x (body); you can’t have the spec and body
16233 both elaborated before each other.
16235 If you later add “with That;” to the body of This, there will be a cycle, in
16236 which case you should erase either “this (body)” or “that (spec)” from the
16237 above forced elaboration order file.
16239 Blank lines and Ada-style comments are ignored. Unit names that do not exist
16240 in the program are ignored. Units in the GNAT predefined library are also
16244 @geindex -p (gnatbind)
16251 Pessimistic elaboration order
16253 This switch is only applicable to the pre-20.x legacy elaboration models.
16254 The post-20.x elaboration model uses a more informed approach of ordering
16257 Normally the binder attempts to choose an elaboration order that is likely to
16258 minimize the likelihood of an elaboration order error resulting in raising a
16259 @code{Program_Error} exception. This switch reverses the action of the binder,
16260 and requests that it deliberately choose an order that is likely to maximize
16261 the likelihood of an elaboration error. This is useful in ensuring
16262 portability and avoiding dependence on accidental fortuitous elaboration
16265 Normally it only makes sense to use the @code{-p} switch if dynamic
16266 elaboration checking is used (@code{-gnatE} switch used for compilation).
16267 This is because in the default static elaboration mode, all necessary
16268 @code{Elaborate} and @code{Elaborate_All} pragmas are implicitly inserted.
16269 These implicit pragmas are still respected by the binder in @code{-p}
16270 mode, so a safe elaboration order is assured.
16272 Note that @code{-p} is not intended for production use; it is more for
16273 debugging/experimental use.
16276 @node Output Control,Dynamic Allocation Control,Elaboration Control,Switches for gnatbind
16277 @anchor{gnat_ugn/building_executable_programs_with_gnat id38}@anchor{119}@anchor{gnat_ugn/building_executable_programs_with_gnat output-control}@anchor{11a}
16278 @subsubsection Output Control
16281 The following switches allow additional control over the output
16282 generated by the binder.
16286 @geindex -c (gnatbind)
16294 Check only. Do not generate the binder output file. In this mode the
16295 binder performs all error checks but does not generate an output file.
16297 @geindex -e (gnatbind)
16301 Output complete list of elaboration-order dependencies, showing the
16302 reason for each dependency. This output can be rather extensive but may
16303 be useful in diagnosing problems with elaboration order. The output is
16304 written to @code{stdout}.
16306 @geindex -h (gnatbind)
16310 Output usage information. The output is written to @code{stdout}.
16312 @geindex -K (gnatbind)
16316 Output linker options to @code{stdout}. Includes library search paths,
16317 contents of pragmas Ident and Linker_Options, and libraries added
16318 by @code{gnatbind}.
16320 @geindex -l (gnatbind)
16324 Output chosen elaboration order. The output is written to @code{stdout}.
16326 @geindex -O (gnatbind)
16330 Output full names of all the object files that must be linked to provide
16331 the Ada component of the program. The output is written to @code{stdout}.
16332 This list includes the files explicitly supplied and referenced by the user
16333 as well as implicitly referenced run-time unit files. The latter are
16334 omitted if the corresponding units reside in shared libraries. The
16335 directory names for the run-time units depend on the system configuration.
16337 @geindex -o (gnatbind)
16339 @item @code{-o @emph{file}}
16341 Set name of output file to @code{file} instead of the normal
16342 @code{b~`mainprog}.adb` default. Note that @code{file} denote the Ada
16343 binder generated body filename.
16344 Note that if this option is used, then linking must be done manually.
16345 It is not possible to use gnatlink in this case, since it cannot locate
16348 @geindex -r (gnatbind)
16352 Generate list of @code{pragma Restrictions} that could be applied to
16353 the current unit. This is useful for code audit purposes, and also may
16354 be used to improve code generation in some cases.
16357 @node Dynamic Allocation Control,Binding with Non-Ada Main Programs,Output Control,Switches for gnatbind
16358 @anchor{gnat_ugn/building_executable_programs_with_gnat dynamic-allocation-control}@anchor{112}@anchor{gnat_ugn/building_executable_programs_with_gnat id39}@anchor{11b}
16359 @subsubsection Dynamic Allocation Control
16362 The heap control switches – @code{-H32} and @code{-H64} –
16363 determine whether dynamic allocation uses 32-bit or 64-bit memory.
16364 They only affect compiler-generated allocations via @code{__gnat_malloc};
16365 explicit calls to @code{malloc} and related functions from the C
16366 run-time library are unaffected.
16373 Allocate memory on 32-bit heap
16377 Allocate memory on 64-bit heap. This is the default
16378 unless explicitly overridden by a @code{'Size} clause on the access type.
16381 These switches are only effective on VMS platforms.
16383 @node Binding with Non-Ada Main Programs,Binding Programs with No Main Subprogram,Dynamic Allocation Control,Switches for gnatbind
16384 @anchor{gnat_ugn/building_executable_programs_with_gnat binding-with-non-ada-main-programs}@anchor{a0}@anchor{gnat_ugn/building_executable_programs_with_gnat id40}@anchor{11c}
16385 @subsubsection Binding with Non-Ada Main Programs
16388 The description so far has assumed that the main
16389 program is in Ada, and that the task of the binder is to generate a
16390 corresponding function @code{main} that invokes this Ada main
16391 program. GNAT also supports the building of executable programs where
16392 the main program is not in Ada, but some of the called routines are
16393 written in Ada and compiled using GNAT (@ref{2c,,Mixed Language Programming}).
16394 The following switch is used in this situation:
16398 @geindex -n (gnatbind)
16406 No main program. The main program is not in Ada.
16409 In this case, most of the functions of the binder are still required,
16410 but instead of generating a main program, the binder generates a file
16411 containing the following callable routines:
16420 @item @code{adainit}
16422 You must call this routine to initialize the Ada part of the program by
16423 calling the necessary elaboration routines. A call to @code{adainit} is
16424 required before the first call to an Ada subprogram.
16426 Note that it is assumed that the basic execution environment must be setup
16427 to be appropriate for Ada execution at the point where the first Ada
16428 subprogram is called. In particular, if the Ada code will do any
16429 floating-point operations, then the FPU must be setup in an appropriate
16430 manner. For the case of the x86, for example, full precision mode is
16431 required. The procedure GNAT.Float_Control.Reset may be used to ensure
16432 that the FPU is in the right state.
16440 @item @code{adafinal}
16442 You must call this routine to perform any library-level finalization
16443 required by the Ada subprograms. A call to @code{adafinal} is required
16444 after the last call to an Ada subprogram, and before the program
16449 @geindex -n (gnatbind)
16452 @geindex multiple input files
16454 If the @code{-n} switch
16455 is given, more than one ALI file may appear on
16456 the command line for @code{gnatbind}. The normal @code{closure}
16457 calculation is performed for each of the specified units. Calculating
16458 the closure means finding out the set of units involved by tracing
16459 @emph{with} references. The reason it is necessary to be able to
16460 specify more than one ALI file is that a given program may invoke two or
16461 more quite separate groups of Ada units.
16463 The binder takes the name of its output file from the last specified ALI
16464 file, unless overridden by the use of the @code{-o file}.
16466 @geindex -o (gnatbind)
16468 The output is an Ada unit in source form that can be compiled with GNAT.
16469 This compilation occurs automatically as part of the @code{gnatlink}
16472 Currently the GNAT run-time requires a FPU using 80 bits mode
16473 precision. Under targets where this is not the default it is required to
16474 call GNAT.Float_Control.Reset before using floating point numbers (this
16475 include float computation, float input and output) in the Ada code. A
16476 side effect is that this could be the wrong mode for the foreign code
16477 where floating point computation could be broken after this call.
16479 @node Binding Programs with No Main Subprogram,,Binding with Non-Ada Main Programs,Switches for gnatbind
16480 @anchor{gnat_ugn/building_executable_programs_with_gnat binding-programs-with-no-main-subprogram}@anchor{11d}@anchor{gnat_ugn/building_executable_programs_with_gnat id41}@anchor{11e}
16481 @subsubsection Binding Programs with No Main Subprogram
16484 It is possible to have an Ada program which does not have a main
16485 subprogram. This program will call the elaboration routines of all the
16486 packages, then the finalization routines.
16488 The following switch is used to bind programs organized in this manner:
16492 @geindex -z (gnatbind)
16500 Normally the binder checks that the unit name given on the command line
16501 corresponds to a suitable main subprogram. When this switch is used,
16502 a list of ALI files can be given, and the execution of the program
16503 consists of elaboration of these units in an appropriate order. Note
16504 that the default wide character encoding method for standard Text_IO
16505 files is always set to Brackets if this switch is set (you can use
16507 @code{-Wx} to override this default).
16510 @node Command-Line Access,Search Paths for gnatbind,Switches for gnatbind,Binding with gnatbind
16511 @anchor{gnat_ugn/building_executable_programs_with_gnat command-line-access}@anchor{11f}@anchor{gnat_ugn/building_executable_programs_with_gnat id42}@anchor{120}
16512 @subsection Command-Line Access
16515 The package @code{Ada.Command_Line} provides access to the command-line
16516 arguments and program name. In order for this interface to operate
16517 correctly, the two variables
16528 are declared in one of the GNAT library routines. These variables must
16529 be set from the actual @code{argc} and @code{argv} values passed to the
16530 main program. With no @emph{n} present, @code{gnatbind}
16531 generates the C main program to automatically set these variables.
16532 If the @emph{n} switch is used, there is no automatic way to
16533 set these variables. If they are not set, the procedures in
16534 @code{Ada.Command_Line} will not be available, and any attempt to use
16535 them will raise @code{Constraint_Error}. If command line access is
16536 required, your main program must set @code{gnat_argc} and
16537 @code{gnat_argv} from the @code{argc} and @code{argv} values passed to
16540 @node Search Paths for gnatbind,Examples of gnatbind Usage,Command-Line Access,Binding with gnatbind
16541 @anchor{gnat_ugn/building_executable_programs_with_gnat id43}@anchor{121}@anchor{gnat_ugn/building_executable_programs_with_gnat search-paths-for-gnatbind}@anchor{76}
16542 @subsection Search Paths for @code{gnatbind}
16545 The binder takes the name of an ALI file as its argument and needs to
16546 locate source files as well as other ALI files to verify object consistency.
16548 For source files, it follows exactly the same search rules as @code{gcc}
16549 (see @ref{73,,Search Paths and the Run-Time Library (RTL)}). For ALI files the
16550 directories searched are:
16556 The directory containing the ALI file named in the command line, unless
16557 the switch @code{-I-} is specified.
16560 All directories specified by @code{-I}
16561 switches on the @code{gnatbind}
16562 command line, in the order given.
16564 @geindex ADA_PRJ_OBJECTS_FILE
16567 Each of the directories listed in the text file whose name is given
16569 @geindex ADA_PRJ_OBJECTS_FILE
16570 @geindex environment variable; ADA_PRJ_OBJECTS_FILE
16571 @code{ADA_PRJ_OBJECTS_FILE} environment variable.
16573 @geindex ADA_PRJ_OBJECTS_FILE
16574 @geindex environment variable; ADA_PRJ_OBJECTS_FILE
16575 @code{ADA_PRJ_OBJECTS_FILE} is normally set by gnatmake or by the gnat
16576 driver when project files are used. It should not normally be set
16579 @geindex ADA_OBJECTS_PATH
16582 Each of the directories listed in the value of the
16583 @geindex ADA_OBJECTS_PATH
16584 @geindex environment variable; ADA_OBJECTS_PATH
16585 @code{ADA_OBJECTS_PATH} environment variable.
16586 Construct this value
16589 @geindex environment variable; PATH
16590 @code{PATH} environment variable: a list of directory
16591 names separated by colons (semicolons when working with the NT version
16595 The content of the @code{ada_object_path} file which is part of the GNAT
16596 installation tree and is used to store standard libraries such as the
16597 GNAT Run-Time Library (RTL) unless the switch @code{-nostdlib} is
16598 specified. See @ref{72,,Installing a library}
16601 @geindex -I (gnatbind)
16603 @geindex -aI (gnatbind)
16605 @geindex -aO (gnatbind)
16607 In the binder the switch @code{-I}
16608 is used to specify both source and
16609 library file paths. Use @code{-aI}
16610 instead if you want to specify
16611 source paths only, and @code{-aO}
16612 if you want to specify library paths
16613 only. This means that for the binder
16614 @code{-I@emph{dir}} is equivalent to
16615 @code{-aI@emph{dir}}
16616 @code{-aO`@emph{dir}}.
16617 The binder generates the bind file (a C language source file) in the
16618 current working directory.
16624 @geindex Interfaces
16628 The packages @code{Ada}, @code{System}, and @code{Interfaces} and their
16629 children make up the GNAT Run-Time Library, together with the package
16630 GNAT and its children, which contain a set of useful additional
16631 library functions provided by GNAT. The sources for these units are
16632 needed by the compiler and are kept together in one directory. The ALI
16633 files and object files generated by compiling the RTL are needed by the
16634 binder and the linker and are kept together in one directory, typically
16635 different from the directory containing the sources. In a normal
16636 installation, you need not specify these directory names when compiling
16637 or binding. Either the environment variables or the built-in defaults
16638 cause these files to be found.
16640 Besides simplifying access to the RTL, a major use of search paths is
16641 in compiling sources from multiple directories. This can make
16642 development environments much more flexible.
16644 @node Examples of gnatbind Usage,,Search Paths for gnatbind,Binding with gnatbind
16645 @anchor{gnat_ugn/building_executable_programs_with_gnat examples-of-gnatbind-usage}@anchor{122}@anchor{gnat_ugn/building_executable_programs_with_gnat id44}@anchor{123}
16646 @subsection Examples of @code{gnatbind} Usage
16649 Here are some examples of @code{gnatbind} invovations:
16657 The main program @code{Hello} (source program in @code{hello.adb}) is
16658 bound using the standard switch settings. The generated main program is
16659 @code{b~hello.adb}. This is the normal, default use of the binder.
16662 gnatbind hello -o mainprog.adb
16665 The main program @code{Hello} (source program in @code{hello.adb}) is
16666 bound using the standard switch settings. The generated main program is
16667 @code{mainprog.adb} with the associated spec in
16668 @code{mainprog.ads}. Note that you must specify the body here not the
16669 spec. Note that if this option is used, then linking must be done manually,
16670 since gnatlink will not be able to find the generated file.
16673 @node Linking with gnatlink,Using the GNU make Utility,Binding with gnatbind,Building Executable Programs with GNAT
16674 @anchor{gnat_ugn/building_executable_programs_with_gnat id45}@anchor{124}@anchor{gnat_ugn/building_executable_programs_with_gnat linking-with-gnatlink}@anchor{c9}
16675 @section Linking with @code{gnatlink}
16680 This chapter discusses @code{gnatlink}, a tool that links
16681 an Ada program and builds an executable file. This utility
16682 invokes the system linker (via the @code{gcc} command)
16683 with a correct list of object files and library references.
16684 @code{gnatlink} automatically determines the list of files and
16685 references for the Ada part of a program. It uses the binder file
16686 generated by the @code{gnatbind} to determine this list.
16689 * Running gnatlink::
16690 * Switches for gnatlink::
16694 @node Running gnatlink,Switches for gnatlink,,Linking with gnatlink
16695 @anchor{gnat_ugn/building_executable_programs_with_gnat id46}@anchor{125}@anchor{gnat_ugn/building_executable_programs_with_gnat running-gnatlink}@anchor{126}
16696 @subsection Running @code{gnatlink}
16699 The form of the @code{gnatlink} command is
16702 $ gnatlink [ switches ] mainprog [.ali]
16703 [ non-Ada objects ] [ linker options ]
16706 The arguments of @code{gnatlink} (switches, main @code{ALI} file,
16708 or linker options) may be in any order, provided that no non-Ada object may
16709 be mistaken for a main @code{ALI} file.
16710 Any file name @code{F} without the @code{.ali}
16711 extension will be taken as the main @code{ALI} file if a file exists
16712 whose name is the concatenation of @code{F} and @code{.ali}.
16714 @code{mainprog.ali} references the ALI file of the main program.
16715 The @code{.ali} extension of this file can be omitted. From this
16716 reference, @code{gnatlink} locates the corresponding binder file
16717 @code{b~mainprog.adb} and, using the information in this file along
16718 with the list of non-Ada objects and linker options, constructs a
16719 linker command file to create the executable.
16721 The arguments other than the @code{gnatlink} switches and the main
16722 @code{ALI} file are passed to the linker uninterpreted.
16723 They typically include the names of
16724 object files for units written in other languages than Ada and any library
16725 references required to resolve references in any of these foreign language
16726 units, or in @code{Import} pragmas in any Ada units.
16728 @code{linker options} is an optional list of linker specific
16730 The default linker called by gnatlink is @code{gcc} which in
16731 turn calls the appropriate system linker.
16733 One useful option for the linker is @code{-s}: it reduces the size of the
16734 executable by removing all symbol table and relocation information from the
16737 Standard options for the linker such as @code{-lmy_lib} or
16738 @code{-Ldir} can be added as is.
16739 For options that are not recognized by
16740 @code{gcc} as linker options, use the @code{gcc} switches
16741 @code{-Xlinker} or @code{-Wl,}.
16743 Refer to the GCC documentation for
16746 Here is an example showing how to generate a linker map:
16749 $ gnatlink my_prog -Wl,-Map,MAPFILE
16752 Using @code{linker options} it is possible to set the program stack and
16754 See @ref{127,,Setting Stack Size from gnatlink} and
16755 @ref{128,,Setting Heap Size from gnatlink}.
16757 @code{gnatlink} determines the list of objects required by the Ada
16758 program and prepends them to the list of objects passed to the linker.
16759 @code{gnatlink} also gathers any arguments set by the use of
16760 @code{pragma Linker_Options} and adds them to the list of arguments
16761 presented to the linker.
16763 @node Switches for gnatlink,,Running gnatlink,Linking with gnatlink
16764 @anchor{gnat_ugn/building_executable_programs_with_gnat id47}@anchor{129}@anchor{gnat_ugn/building_executable_programs_with_gnat switches-for-gnatlink}@anchor{12a}
16765 @subsection Switches for @code{gnatlink}
16768 The following switches are available with the @code{gnatlink} utility:
16770 @geindex --version (gnatlink)
16775 @item @code{--version}
16777 Display Copyright and version, then exit disregarding all other options.
16780 @geindex --help (gnatlink)
16785 @item @code{--help}
16787 If @code{--version} was not used, display usage, then exit disregarding
16791 @geindex Command line length
16793 @geindex -f (gnatlink)
16800 On some targets, the command line length is limited, and @code{gnatlink}
16801 will generate a separate file for the linker if the list of object files
16803 The @code{-f} switch forces this file
16804 to be generated even if
16805 the limit is not exceeded. This is useful in some cases to deal with
16806 special situations where the command line length is exceeded.
16809 @geindex Debugging information
16812 @geindex -g (gnatlink)
16819 The option to include debugging information causes the Ada bind file (in
16820 other words, @code{b~mainprog.adb}) to be compiled with @code{-g}.
16821 In addition, the binder does not delete the @code{b~mainprog.adb},
16822 @code{b~mainprog.o} and @code{b~mainprog.ali} files.
16823 Without @code{-g}, the binder removes these files by default.
16826 @geindex -n (gnatlink)
16833 Do not compile the file generated by the binder. This may be used when
16834 a link is rerun with different options, but there is no need to recompile
16838 @geindex -v (gnatlink)
16845 Verbose mode. Causes additional information to be output, including a full
16846 list of the included object files.
16847 This switch option is most useful when you want
16848 to see what set of object files are being used in the link step.
16851 @geindex -v -v (gnatlink)
16858 Very verbose mode. Requests that the compiler operate in verbose mode when
16859 it compiles the binder file, and that the system linker run in verbose mode.
16862 @geindex -o (gnatlink)
16867 @item @code{-o @emph{exec-name}}
16869 @code{exec-name} specifies an alternate name for the generated
16870 executable program. If this switch is omitted, the executable has the same
16871 name as the main unit. For example, @code{gnatlink try.ali} creates
16872 an executable called @code{try}.
16875 @geindex -B (gnatlink)
16880 @item @code{-B@emph{dir}}
16882 Load compiler executables (for example, @code{gnat1}, the Ada compiler)
16883 from @code{dir} instead of the default location. Only use this switch
16884 when multiple versions of the GNAT compiler are available.
16885 See the @code{Directory Options} section in @cite{The_GNU_Compiler_Collection}
16886 for further details. You would normally use the @code{-b} or
16887 @code{-V} switch instead.
16890 @geindex -M (gnatlink)
16897 When linking an executable, create a map file. The name of the map file
16898 has the same name as the executable with extension “.map”.
16901 @geindex -M= (gnatlink)
16906 @item @code{-M=@emph{mapfile}}
16908 When linking an executable, create a map file. The name of the map file is
16912 @geindex --GCC=compiler_name (gnatlink)
16917 @item @code{--GCC=@emph{compiler_name}}
16919 Program used for compiling the binder file. The default is
16920 @code{gcc}. You need to use quotes around @code{compiler_name} if
16921 @code{compiler_name} contains spaces or other separator characters.
16922 As an example @code{--GCC="foo -x -y"} will instruct @code{gnatlink} to
16923 use @code{foo -x -y} as your compiler. Note that switch @code{-c} is always
16924 inserted after your command name. Thus in the above example the compiler
16925 command that will be used by @code{gnatlink} will be @code{foo -c -x -y}.
16926 A limitation of this syntax is that the name and path name of the executable
16927 itself must not include any embedded spaces. If the compiler executable is
16928 different from the default one (gcc or <prefix>-gcc), then the back-end
16929 switches in the ALI file are not used to compile the binder generated source.
16930 For example, this is the case with @code{--GCC="foo -x -y"}. But the back end
16931 switches will be used for @code{--GCC="gcc -gnatv"}. If several
16932 @code{--GCC=compiler_name} are used, only the last @code{compiler_name}
16933 is taken into account. However, all the additional switches are also taken
16934 into account. Thus,
16935 @code{--GCC="foo -x -y" --GCC="bar -z -t"} is equivalent to
16936 @code{--GCC="bar -x -y -z -t"}.
16939 @geindex --LINK= (gnatlink)
16944 @item @code{--LINK=@emph{name}}
16946 @code{name} is the name of the linker to be invoked. This is especially
16947 useful in mixed language programs since languages such as C++ require
16948 their own linker to be used. When this switch is omitted, the default
16949 name for the linker is @code{gcc}. When this switch is used, the
16950 specified linker is called instead of @code{gcc} with exactly the same
16951 parameters that would have been passed to @code{gcc} so if the desired
16952 linker requires different parameters it is necessary to use a wrapper
16953 script that massages the parameters before invoking the real linker. It
16954 may be useful to control the exact invocation by using the verbose
16958 @node Using the GNU make Utility,,Linking with gnatlink,Building Executable Programs with GNAT
16959 @anchor{gnat_ugn/building_executable_programs_with_gnat id48}@anchor{12b}@anchor{gnat_ugn/building_executable_programs_with_gnat using-the-gnu-make-utility}@anchor{70}
16960 @section Using the GNU @code{make} Utility
16963 @geindex make (GNU)
16966 This chapter offers some examples of makefiles that solve specific
16967 problems. It does not explain how to write a makefile, nor does it try to replace the
16968 @code{gnatmake} utility (@ref{c6,,Building with gnatmake}).
16970 All the examples in this section are specific to the GNU version of
16971 make. Although @code{make} is a standard utility, and the basic language
16972 is the same, these examples use some advanced features found only in
16976 * Using gnatmake in a Makefile::
16977 * Automatically Creating a List of Directories::
16978 * Generating the Command Line Switches::
16979 * Overcoming Command Line Length Limits::
16983 @node Using gnatmake in a Makefile,Automatically Creating a List of Directories,,Using the GNU make Utility
16984 @anchor{gnat_ugn/building_executable_programs_with_gnat id49}@anchor{12c}@anchor{gnat_ugn/building_executable_programs_with_gnat using-gnatmake-in-a-makefile}@anchor{12d}
16985 @subsection Using gnatmake in a Makefile
16988 @c index makefile (GNU make)
16990 Complex project organizations can be handled in a very powerful way by
16991 using GNU make combined with gnatmake. For instance, here is a Makefile
16992 which allows you to build each subsystem of a big project into a separate
16993 shared library. Such a makefile allows you to significantly reduce the link
16994 time of very big applications while maintaining full coherence at
16995 each step of the build process.
16997 The list of dependencies are handled automatically by
16998 @code{gnatmake}. The Makefile is simply used to call gnatmake in each of
16999 the appropriate directories.
17001 Note that you should also read the example on how to automatically
17002 create the list of directories
17003 (@ref{12e,,Automatically Creating a List of Directories})
17004 which might help you in case your project has a lot of subdirectories.
17007 ## This Makefile is intended to be used with the following directory
17009 ## - The sources are split into a series of csc (computer software components)
17010 ## Each of these csc is put in its own directory.
17011 ## Their name are referenced by the directory names.
17012 ## They will be compiled into shared library (although this would also work
17013 ## with static libraries
17014 ## - The main program (and possibly other packages that do not belong to any
17015 ## csc is put in the top level directory (where the Makefile is).
17016 ## toplevel_dir __ first_csc (sources) __ lib (will contain the library)
17017 ## \\_ second_csc (sources) __ lib (will contain the library)
17019 ## Although this Makefile is build for shared library, it is easy to modify
17020 ## to build partial link objects instead (modify the lines with -shared and
17023 ## With this makefile, you can change any file in the system or add any new
17024 ## file, and everything will be recompiled correctly (only the relevant shared
17025 ## objects will be recompiled, and the main program will be re-linked).
17027 # The list of computer software component for your project. This might be
17028 # generated automatically.
17031 # Name of the main program (no extension)
17034 # If we need to build objects with -fPIC, uncomment the following line
17037 # The following variable should give the directory containing libgnat.so
17038 # You can get this directory through 'gnatls -v'. This is usually the last
17039 # directory in the Object_Path.
17042 # The directories for the libraries
17043 # (This macro expands the list of CSC to the list of shared libraries, you
17044 # could simply use the expanded form:
17045 # LIB_DIR=aa/lib/libaa.so bb/lib/libbb.so cc/lib/libcc.so
17046 LIB_DIR=$@{foreach dir,$@{CSC_LIST@},$@{dir@}/lib/lib$@{dir@}.so@}
17048 $@{MAIN@}: objects $@{LIB_DIR@}
17049 gnatbind $@{MAIN@} $@{CSC_LIST:%=-aO%/lib@} -shared
17050 gnatlink $@{MAIN@} $@{CSC_LIST:%=-l%@}
17053 # recompile the sources
17054 gnatmake -c -i $@{MAIN@}.adb $@{NEED_FPIC@} $@{CSC_LIST:%=-I%@}
17056 # Note: In a future version of GNAT, the following commands will be simplified
17057 # by a new tool, gnatmlib
17059 mkdir -p $@{dir $@@ @}
17060 cd $@{dir $@@ @} && gcc -shared -o $@{notdir $@@ @} ../*.o -L$@{GLIB@} -lgnat
17061 cd $@{dir $@@ @} && cp -f ../*.ali .
17063 # The dependencies for the modules
17064 # Note that we have to force the expansion of *.o, since in some cases
17065 # make won't be able to do it itself.
17066 aa/lib/libaa.so: $@{wildcard aa/*.o@}
17067 bb/lib/libbb.so: $@{wildcard bb/*.o@}
17068 cc/lib/libcc.so: $@{wildcard cc/*.o@}
17070 # Make sure all of the shared libraries are in the path before starting the
17073 LD_LIBRARY_PATH=`pwd`/aa/lib:`pwd`/bb/lib:`pwd`/cc/lib ./$@{MAIN@}
17076 $@{RM@} -rf $@{CSC_LIST:%=%/lib@}
17077 $@{RM@} $@{CSC_LIST:%=%/*.ali@}
17078 $@{RM@} $@{CSC_LIST:%=%/*.o@}
17079 $@{RM@} *.o *.ali $@{MAIN@}
17082 @node Automatically Creating a List of Directories,Generating the Command Line Switches,Using gnatmake in a Makefile,Using the GNU make Utility
17083 @anchor{gnat_ugn/building_executable_programs_with_gnat automatically-creating-a-list-of-directories}@anchor{12e}@anchor{gnat_ugn/building_executable_programs_with_gnat id50}@anchor{12f}
17084 @subsection Automatically Creating a List of Directories
17087 In most makefiles, you will have to specify a list of directories, and
17088 store it in a variable. For small projects, it is often easier to
17089 specify each of them by hand, since you then have full control over what
17090 is the proper order for these directories, which ones should be
17093 However, in larger projects, which might involve hundreds of
17094 subdirectories, it might be more convenient to generate this list
17097 The example below presents two methods. The first one, although less
17098 general, gives you more control over the list. It involves wildcard
17099 characters, that are automatically expanded by @code{make}. Its
17100 shortcoming is that you need to explicitly specify some of the
17101 organization of your project, such as for instance the directory tree
17102 depth, whether some directories are found in a separate tree, etc.
17104 The second method is the most general one. It requires an external
17105 program, called @code{find}, which is standard on all Unix systems. All
17106 the directories found under a given root directory will be added to the
17110 # The examples below are based on the following directory hierarchy:
17111 # All the directories can contain any number of files
17112 # ROOT_DIRECTORY -> a -> aa -> aaa
17115 # -> b -> ba -> baa
17118 # This Makefile creates a variable called DIRS, that can be reused any time
17119 # you need this list (see the other examples in this section)
17121 # The root of your project's directory hierarchy
17125 # First method: specify explicitly the list of directories
17126 # This allows you to specify any subset of all the directories you need.
17129 DIRS := a/aa/ a/ab/ b/ba/
17132 # Second method: use wildcards
17133 # Note that the argument(s) to wildcard below should end with a '/'.
17134 # Since wildcards also return file names, we have to filter them out
17135 # to avoid duplicate directory names.
17136 # We thus use make's `@w{`}dir`@w{`} and `@w{`}sort`@w{`} functions.
17137 # It sets DIRs to the following value (note that the directories aaa and baa
17138 # are not given, unless you change the arguments to wildcard).
17139 # DIRS= ./a/a/ ./b/ ./a/aa/ ./a/ab/ ./a/ac/ ./b/ba/ ./b/bb/ ./b/bc/
17142 DIRS := $@{sort $@{dir $@{wildcard $@{ROOT_DIRECTORY@}/*/
17143 $@{ROOT_DIRECTORY@}/*/*/@}@}@}
17146 # Third method: use an external program
17147 # This command is much faster if run on local disks, avoiding NFS slowdowns.
17148 # This is the most complete command: it sets DIRs to the following value:
17149 # DIRS= ./a ./a/aa ./a/aa/aaa ./a/ab ./a/ac ./b ./b/ba ./b/ba/baa ./b/bb ./b/bc
17152 DIRS := $@{shell find $@{ROOT_DIRECTORY@} -type d -print@}
17155 @node Generating the Command Line Switches,Overcoming Command Line Length Limits,Automatically Creating a List of Directories,Using the GNU make Utility
17156 @anchor{gnat_ugn/building_executable_programs_with_gnat generating-the-command-line-switches}@anchor{130}@anchor{gnat_ugn/building_executable_programs_with_gnat id51}@anchor{131}
17157 @subsection Generating the Command Line Switches
17160 Once you have created the list of directories as explained in the
17161 previous section (@ref{12e,,Automatically Creating a List of Directories}),
17162 you can easily generate the command line arguments to pass to gnatmake.
17164 For the sake of completeness, this example assumes that the source path
17165 is not the same as the object path, and that you have two separate lists
17169 # see "Automatically creating a list of directories" to create
17174 GNATMAKE_SWITCHES := $@{patsubst %,-aI%,$@{SOURCE_DIRS@}@}
17175 GNATMAKE_SWITCHES += $@{patsubst %,-aO%,$@{OBJECT_DIRS@}@}
17178 gnatmake $@{GNATMAKE_SWITCHES@} main_unit
17181 @node Overcoming Command Line Length Limits,,Generating the Command Line Switches,Using the GNU make Utility
17182 @anchor{gnat_ugn/building_executable_programs_with_gnat id52}@anchor{132}@anchor{gnat_ugn/building_executable_programs_with_gnat overcoming-command-line-length-limits}@anchor{133}
17183 @subsection Overcoming Command Line Length Limits
17186 One problem that might be encountered on big projects is that many
17187 operating systems limit the length of the command line. It is thus hard to give
17188 gnatmake the list of source and object directories.
17190 This example shows how you can set up environment variables, which will
17191 make @code{gnatmake} behave exactly as if the directories had been
17192 specified on the command line, but have a much higher length limit (or
17193 even none on most systems).
17195 It assumes that you have created a list of directories in your Makefile,
17196 using one of the methods presented in
17197 @ref{12e,,Automatically Creating a List of Directories}.
17198 For the sake of completeness, we assume that the object
17199 path (where the ALI files are found) is different from the sources patch.
17201 Note a small trick in the Makefile below: for efficiency reasons, we
17202 create two temporary variables (SOURCE_LIST and OBJECT_LIST), that are
17203 expanded immediately by @code{make}. This way we overcome the standard
17204 make behavior which is to expand the variables only when they are
17207 On Windows, if you are using the standard Windows command shell, you must
17208 replace colons with semicolons in the assignments to these variables.
17211 # In this example, we create both ADA_INCLUDE_PATH and ADA_OBJECTS_PATH.
17212 # This is the same thing as putting the -I arguments on the command line.
17213 # (the equivalent of using -aI on the command line would be to define
17214 # only ADA_INCLUDE_PATH, the equivalent of -aO is ADA_OBJECTS_PATH).
17215 # You can of course have different values for these variables.
17217 # Note also that we need to keep the previous values of these variables, since
17218 # they might have been set before running 'make' to specify where the GNAT
17219 # library is installed.
17221 # see "Automatically creating a list of directories" to create these
17227 space:=$@{empty@} $@{empty@}
17228 SOURCE_LIST := $@{subst $@{space@},:,$@{SOURCE_DIRS@}@}
17229 OBJECT_LIST := $@{subst $@{space@},:,$@{OBJECT_DIRS@}@}
17230 ADA_INCLUDE_PATH += $@{SOURCE_LIST@}
17231 ADA_OBJECTS_PATH += $@{OBJECT_LIST@}
17232 export ADA_INCLUDE_PATH
17233 export ADA_OBJECTS_PATH
17239 @node GNAT Utility Programs,GNAT and Program Execution,Building Executable Programs with GNAT,Top
17240 @anchor{gnat_ugn/gnat_utility_programs doc}@anchor{134}@anchor{gnat_ugn/gnat_utility_programs gnat-utility-programs}@anchor{b}@anchor{gnat_ugn/gnat_utility_programs id1}@anchor{135}
17241 @chapter GNAT Utility Programs
17244 This chapter describes a number of utility programs:
17251 @ref{136,,The File Cleanup Utility gnatclean}
17254 @ref{137,,The GNAT Library Browser gnatls}
17257 Other GNAT utilities are described elsewhere in this manual:
17263 @ref{42,,Handling Arbitrary File Naming Conventions with gnatname}
17266 @ref{4c,,File Name Krunching with gnatkr}
17269 @ref{1d,,Renaming Files with gnatchop}
17272 @ref{8f,,Preprocessing with gnatprep}
17276 * The File Cleanup Utility gnatclean::
17277 * The GNAT Library Browser gnatls::
17281 @node The File Cleanup Utility gnatclean,The GNAT Library Browser gnatls,,GNAT Utility Programs
17282 @anchor{gnat_ugn/gnat_utility_programs id2}@anchor{138}@anchor{gnat_ugn/gnat_utility_programs the-file-cleanup-utility-gnatclean}@anchor{136}
17283 @section The File Cleanup Utility @code{gnatclean}
17286 @geindex File cleanup tool
17290 @code{gnatclean} is a tool that allows the deletion of files produced by the
17291 compiler, binder and linker, including ALI files, object files, tree files,
17292 expanded source files, library files, interface copy source files, binder
17293 generated files and executable files.
17296 * Running gnatclean::
17297 * Switches for gnatclean::
17301 @node Running gnatclean,Switches for gnatclean,,The File Cleanup Utility gnatclean
17302 @anchor{gnat_ugn/gnat_utility_programs id3}@anchor{139}@anchor{gnat_ugn/gnat_utility_programs running-gnatclean}@anchor{13a}
17303 @subsection Running @code{gnatclean}
17306 The @code{gnatclean} command has the form:
17311 $ gnatclean switches names
17315 where @code{names} is a list of source file names. Suffixes @code{.ads} and
17316 @code{adb} may be omitted. If a project file is specified using switch
17317 @code{-P}, then @code{names} may be completely omitted.
17319 In normal mode, @code{gnatclean} delete the files produced by the compiler and,
17320 if switch @code{-c} is not specified, by the binder and
17321 the linker. In informative-only mode, specified by switch
17322 @code{-n}, the list of files that would have been deleted in
17323 normal mode is listed, but no file is actually deleted.
17325 @node Switches for gnatclean,,Running gnatclean,The File Cleanup Utility gnatclean
17326 @anchor{gnat_ugn/gnat_utility_programs id4}@anchor{13b}@anchor{gnat_ugn/gnat_utility_programs switches-for-gnatclean}@anchor{13c}
17327 @subsection Switches for @code{gnatclean}
17330 @code{gnatclean} recognizes the following switches:
17332 @geindex --version (gnatclean)
17337 @item @code{--version}
17339 Display copyright and version, then exit disregarding all other options.
17342 @geindex --help (gnatclean)
17347 @item @code{--help}
17349 If @code{--version} was not used, display usage, then exit disregarding
17352 @item @code{--subdirs=@emph{subdir}}
17354 Actual object directory of each project file is the subdirectory subdir of the
17355 object directory specified or defaulted in the project file.
17357 @item @code{--unchecked-shared-lib-imports}
17359 By default, shared library projects are not allowed to import static library
17360 projects. When this switch is used on the command line, this restriction is
17364 @geindex -c (gnatclean)
17371 Only attempt to delete the files produced by the compiler, not those produced
17372 by the binder or the linker. The files that are not to be deleted are library
17373 files, interface copy files, binder generated files and executable files.
17376 @geindex -D (gnatclean)
17381 @item @code{-D @emph{dir}}
17383 Indicate that ALI and object files should normally be found in directory @code{dir}.
17386 @geindex -F (gnatclean)
17393 When using project files, if some errors or warnings are detected during
17394 parsing and verbose mode is not in effect (no use of switch
17395 -v), then error lines start with the full path name of the project
17396 file, rather than its simple file name.
17399 @geindex -h (gnatclean)
17406 Output a message explaining the usage of @code{gnatclean}.
17409 @geindex -n (gnatclean)
17416 Informative-only mode. Do not delete any files. Output the list of the files
17417 that would have been deleted if this switch was not specified.
17420 @geindex -P (gnatclean)
17425 @item @code{-P@emph{project}}
17427 Use project file @code{project}. Only one such switch can be used.
17428 When cleaning a project file, the files produced by the compilation of the
17429 immediate sources or inherited sources of the project files are to be
17430 deleted. This is not depending on the presence or not of executable names
17431 on the command line.
17434 @geindex -q (gnatclean)
17441 Quiet output. If there are no errors, do not output anything, except in
17442 verbose mode (switch -v) or in informative-only mode
17446 @geindex -r (gnatclean)
17453 When a project file is specified (using switch -P),
17454 clean all imported and extended project files, recursively. If this switch
17455 is not specified, only the files related to the main project file are to be
17456 deleted. This switch has no effect if no project file is specified.
17459 @geindex -v (gnatclean)
17469 @geindex -vP (gnatclean)
17474 @item @code{-vP@emph{x}}
17476 Indicates the verbosity of the parsing of GNAT project files.
17477 @ref{cf,,Switches Related to Project Files}.
17480 @geindex -X (gnatclean)
17485 @item @code{-X@emph{name}=@emph{value}}
17487 Indicates that external variable @code{name} has the value @code{value}.
17488 The Project Manager will use this value for occurrences of
17489 @code{external(name)} when parsing the project file.
17490 See @ref{cf,,Switches Related to Project Files}.
17493 @geindex -aO (gnatclean)
17498 @item @code{-aO@emph{dir}}
17500 When searching for ALI and object files, look in directory @code{dir}.
17503 @geindex -I (gnatclean)
17508 @item @code{-I@emph{dir}}
17510 Equivalent to @code{-aO@emph{dir}}.
17513 @geindex -I- (gnatclean)
17515 @geindex Source files
17516 @geindex suppressing search
17523 Do not look for ALI or object files in the directory
17524 where @code{gnatclean} was invoked.
17527 @node The GNAT Library Browser gnatls,,The File Cleanup Utility gnatclean,GNAT Utility Programs
17528 @anchor{gnat_ugn/gnat_utility_programs id5}@anchor{13d}@anchor{gnat_ugn/gnat_utility_programs the-gnat-library-browser-gnatls}@anchor{137}
17529 @section The GNAT Library Browser @code{gnatls}
17532 @geindex Library browser
17536 @code{gnatls} is a tool that outputs information about compiled
17537 units. It gives the relationship between objects, unit names and source
17538 files. It can also be used to check the source dependencies of a unit
17539 as well as various characteristics.
17543 * Switches for gnatls::
17544 * Example of gnatls Usage::
17548 @node Running gnatls,Switches for gnatls,,The GNAT Library Browser gnatls
17549 @anchor{gnat_ugn/gnat_utility_programs id6}@anchor{13e}@anchor{gnat_ugn/gnat_utility_programs running-gnatls}@anchor{13f}
17550 @subsection Running @code{gnatls}
17553 The @code{gnatls} command has the form
17558 $ gnatls switches object_or_ali_file
17562 The main argument is the list of object or @code{ali} files
17563 (see @ref{28,,The Ada Library Information Files})
17564 for which information is requested.
17566 In normal mode, without additional option, @code{gnatls} produces a
17567 four-column listing. Each line represents information for a specific
17568 object. The first column gives the full path of the object, the second
17569 column gives the name of the principal unit in this object, the third
17570 column gives the status of the source and the fourth column gives the
17571 full path of the source representing this unit.
17572 Here is a simple example of use:
17578 ./demo1.o demo1 DIF demo1.adb
17579 ./demo2.o demo2 OK demo2.adb
17580 ./hello.o h1 OK hello.adb
17581 ./instr-child.o instr.child MOK instr-child.adb
17582 ./instr.o instr OK instr.adb
17583 ./tef.o tef DIF tef.adb
17584 ./text_io_example.o text_io_example OK text_io_example.adb
17585 ./tgef.o tgef DIF tgef.adb
17589 The first line can be interpreted as follows: the main unit which is
17591 object file @code{demo1.o} is demo1, whose main source is in
17592 @code{demo1.adb}. Furthermore, the version of the source used for the
17593 compilation of demo1 has been modified (DIF). Each source file has a status
17594 qualifier which can be:
17599 @item @emph{OK (unchanged)}
17601 The version of the source file used for the compilation of the
17602 specified unit corresponds exactly to the actual source file.
17604 @item @emph{MOK (slightly modified)}
17606 The version of the source file used for the compilation of the
17607 specified unit differs from the actual source file but not enough to
17608 require recompilation. If you use gnatmake with the option
17609 @code{-m} (minimal recompilation), a file marked
17610 MOK will not be recompiled.
17612 @item @emph{DIF (modified)}
17614 No version of the source found on the path corresponds to the source
17615 used to build this object.
17617 @item @emph{??? (file not found)}
17619 No source file was found for this unit.
17621 @item @emph{HID (hidden, unchanged version not first on PATH)}
17623 The version of the source that corresponds exactly to the source used
17624 for compilation has been found on the path but it is hidden by another
17625 version of the same source that has been modified.
17628 @node Switches for gnatls,Example of gnatls Usage,Running gnatls,The GNAT Library Browser gnatls
17629 @anchor{gnat_ugn/gnat_utility_programs id7}@anchor{140}@anchor{gnat_ugn/gnat_utility_programs switches-for-gnatls}@anchor{141}
17630 @subsection Switches for @code{gnatls}
17633 @code{gnatls} recognizes the following switches:
17635 @geindex --version (gnatls)
17640 @item @code{--version}
17642 Display copyright and version, then exit disregarding all other options.
17645 @geindex --help (gnatls)
17650 @item @code{--help}
17652 If @code{--version} was not used, display usage, then exit disregarding
17656 @geindex -a (gnatls)
17663 Consider all units, including those of the predefined Ada library.
17664 Especially useful with @code{-d}.
17667 @geindex -d (gnatls)
17674 List sources from which specified units depend on.
17677 @geindex -h (gnatls)
17684 Output the list of options.
17687 @geindex -o (gnatls)
17694 Only output information about object files.
17697 @geindex -s (gnatls)
17704 Only output information about source files.
17707 @geindex -u (gnatls)
17714 Only output information about compilation units.
17717 @geindex -files (gnatls)
17722 @item @code{-files=@emph{file}}
17724 Take as arguments the files listed in text file @code{file}.
17725 Text file @code{file} may contain empty lines that are ignored.
17726 Each nonempty line should contain the name of an existing file.
17727 Several such switches may be specified simultaneously.
17730 @geindex -aO (gnatls)
17732 @geindex -aI (gnatls)
17734 @geindex -I (gnatls)
17736 @geindex -I- (gnatls)
17741 @item @code{-aO@emph{dir}}, @code{-aI@emph{dir}}, @code{-I@emph{dir}}, @code{-I-}, @code{-nostdinc}
17743 Source path manipulation. Same meaning as the equivalent @code{gnatmake}
17744 flags (@ref{ce,,Switches for gnatmake}).
17747 @geindex -aP (gnatls)
17752 @item @code{-aP@emph{dir}}
17754 Add @code{dir} at the beginning of the project search dir.
17757 @geindex --RTS (gnatls)
17762 @item @code{--RTS=@emph{rts-path}}
17764 Specifies the default location of the runtime library. Same meaning as the
17765 equivalent @code{gnatmake} flag (@ref{ce,,Switches for gnatmake}).
17768 @geindex -v (gnatls)
17775 Verbose mode. Output the complete source, object and project paths. Do not use
17776 the default column layout but instead use long format giving as much as
17777 information possible on each requested units, including special
17778 characteristics such as:
17784 @emph{Preelaborable}: The unit is preelaborable in the Ada sense.
17787 @emph{No_Elab_Code}: No elaboration code has been produced by the compiler for this unit.
17790 @emph{Pure}: The unit is pure in the Ada sense.
17793 @emph{Elaborate_Body}: The unit contains a pragma Elaborate_Body.
17796 @emph{Remote_Types}: The unit contains a pragma Remote_Types.
17799 @emph{Shared_Passive}: The unit contains a pragma Shared_Passive.
17802 @emph{Predefined}: This unit is part of the predefined environment and cannot be modified
17806 @emph{Remote_Call_Interface}: The unit contains a pragma Remote_Call_Interface.
17810 @node Example of gnatls Usage,,Switches for gnatls,The GNAT Library Browser gnatls
17811 @anchor{gnat_ugn/gnat_utility_programs example-of-gnatls-usage}@anchor{142}@anchor{gnat_ugn/gnat_utility_programs id8}@anchor{143}
17812 @subsection Example of @code{gnatls} Usage
17815 Example of using the verbose switch. Note how the source and
17816 object paths are affected by the -I switch.
17821 $ gnatls -v -I.. demo1.o
17823 GNATLS 5.03w (20041123-34)
17824 Copyright 1997-2004 Free Software Foundation, Inc.
17826 Source Search Path:
17827 <Current_Directory>
17829 /home/comar/local/adainclude/
17831 Object Search Path:
17832 <Current_Directory>
17834 /home/comar/local/lib/gcc-lib/x86-linux/3.4.3/adalib/
17836 Project Search Path:
17837 <Current_Directory>
17838 /home/comar/local/lib/gnat/
17843 Kind => subprogram body
17844 Flags => No_Elab_Code
17845 Source => demo1.adb modified
17849 The following is an example of use of the dependency list.
17850 Note the use of the -s switch
17851 which gives a straight list of source files. This can be useful for
17852 building specialized scripts.
17857 $ gnatls -d demo2.o
17858 ./demo2.o demo2 OK demo2.adb
17864 $ gnatls -d -s -a demo1.o
17866 /home/comar/local/adainclude/ada.ads
17867 /home/comar/local/adainclude/a-finali.ads
17868 /home/comar/local/adainclude/a-filico.ads
17869 /home/comar/local/adainclude/a-stream.ads
17870 /home/comar/local/adainclude/a-tags.ads
17873 /home/comar/local/adainclude/gnat.ads
17874 /home/comar/local/adainclude/g-io.ads
17876 /home/comar/local/adainclude/system.ads
17877 /home/comar/local/adainclude/s-exctab.ads
17878 /home/comar/local/adainclude/s-finimp.ads
17879 /home/comar/local/adainclude/s-finroo.ads
17880 /home/comar/local/adainclude/s-secsta.ads
17881 /home/comar/local/adainclude/s-stalib.ads
17882 /home/comar/local/adainclude/s-stoele.ads
17883 /home/comar/local/adainclude/s-stratt.ads
17884 /home/comar/local/adainclude/s-tasoli.ads
17885 /home/comar/local/adainclude/s-unstyp.ads
17886 /home/comar/local/adainclude/unchconv.ads
17897 @c -- Example: A |withing| unit has a |with| clause, it |withs| a |withed| unit
17899 @node GNAT and Program Execution,Platform-Specific Information,GNAT Utility Programs,Top
17900 @anchor{gnat_ugn/gnat_and_program_execution doc}@anchor{144}@anchor{gnat_ugn/gnat_and_program_execution gnat-and-program-execution}@anchor{c}@anchor{gnat_ugn/gnat_and_program_execution id1}@anchor{145}
17901 @chapter GNAT and Program Execution
17904 This chapter covers several topics:
17910 @ref{146,,Running and Debugging Ada Programs}
17913 @ref{147,,Profiling}
17916 @ref{148,,Improving Performance}
17919 @ref{149,,Overflow Check Handling in GNAT}
17922 @ref{14a,,Performing Dimensionality Analysis in GNAT}
17925 @ref{14b,,Stack Related Facilities}
17928 @ref{14c,,Memory Management Issues}
17932 * Running and Debugging Ada Programs::
17934 * Improving Performance::
17935 * Overflow Check Handling in GNAT::
17936 * Performing Dimensionality Analysis in GNAT::
17937 * Stack Related Facilities::
17938 * Memory Management Issues::
17942 @node Running and Debugging Ada Programs,Profiling,,GNAT and Program Execution
17943 @anchor{gnat_ugn/gnat_and_program_execution id2}@anchor{146}@anchor{gnat_ugn/gnat_and_program_execution running-and-debugging-ada-programs}@anchor{14d}
17944 @section Running and Debugging Ada Programs
17949 This section discusses how to debug Ada programs.
17951 An incorrect Ada program may be handled in three ways by the GNAT compiler:
17957 The illegality may be a violation of the static semantics of Ada. In
17958 that case GNAT diagnoses the constructs in the program that are illegal.
17959 It is then a straightforward matter for the user to modify those parts of
17963 The illegality may be a violation of the dynamic semantics of Ada. In
17964 that case the program compiles and executes, but may generate incorrect
17965 results, or may terminate abnormally with some exception.
17968 When presented with a program that contains convoluted errors, GNAT
17969 itself may terminate abnormally without providing full diagnostics on
17970 the incorrect user program.
17978 * The GNAT Debugger GDB::
17980 * Introduction to GDB Commands::
17981 * Using Ada Expressions::
17982 * Calling User-Defined Subprograms::
17983 * Using the next Command in a Function::
17984 * Stopping When Ada Exceptions Are Raised::
17986 * Debugging Generic Units::
17987 * Remote Debugging with gdbserver::
17988 * GNAT Abnormal Termination or Failure to Terminate::
17989 * Naming Conventions for GNAT Source Files::
17990 * Getting Internal Debugging Information::
17991 * Stack Traceback::
17992 * Pretty-Printers for the GNAT runtime::
17996 @node The GNAT Debugger GDB,Running GDB,,Running and Debugging Ada Programs
17997 @anchor{gnat_ugn/gnat_and_program_execution id3}@anchor{14e}@anchor{gnat_ugn/gnat_and_program_execution the-gnat-debugger-gdb}@anchor{14f}
17998 @subsection The GNAT Debugger GDB
18001 @code{GDB} is a general purpose, platform-independent debugger that
18002 can be used to debug mixed-language programs compiled with @code{gcc},
18003 and in particular is capable of debugging Ada programs compiled with
18004 GNAT. The latest versions of @code{GDB} are Ada-aware and can handle
18005 complex Ada data structures.
18007 See @cite{Debugging with GDB},
18008 for full details on the usage of @code{GDB}, including a section on
18009 its usage on programs. This manual should be consulted for full
18010 details. The section that follows is a brief introduction to the
18011 philosophy and use of @code{GDB}.
18013 When GNAT programs are compiled, the compiler optionally writes debugging
18014 information into the generated object file, including information on
18015 line numbers, and on declared types and variables. This information is
18016 separate from the generated code. It makes the object files considerably
18017 larger, but it does not add to the size of the actual executable that
18018 will be loaded into memory, and has no impact on run-time performance. The
18019 generation of debug information is triggered by the use of the
18020 @code{-g} switch in the @code{gcc} or @code{gnatmake} command
18021 used to carry out the compilations. It is important to emphasize that
18022 the use of these options does not change the generated code.
18024 The debugging information is written in standard system formats that
18025 are used by many tools, including debuggers and profilers. The format
18026 of the information is typically designed to describe C types and
18027 semantics, but GNAT implements a translation scheme which allows full
18028 details about Ada types and variables to be encoded into these
18029 standard C formats. Details of this encoding scheme may be found in
18030 the file exp_dbug.ads in the GNAT source distribution. However, the
18031 details of this encoding are, in general, of no interest to a user,
18032 since @code{GDB} automatically performs the necessary decoding.
18034 When a program is bound and linked, the debugging information is
18035 collected from the object files, and stored in the executable image of
18036 the program. Again, this process significantly increases the size of
18037 the generated executable file, but it does not increase the size of
18038 the executable program itself. Furthermore, if this program is run in
18039 the normal manner, it runs exactly as if the debug information were
18040 not present, and takes no more actual memory.
18042 However, if the program is run under control of @code{GDB}, the
18043 debugger is activated. The image of the program is loaded, at which
18044 point it is ready to run. If a run command is given, then the program
18045 will run exactly as it would have if @code{GDB} were not present. This
18046 is a crucial part of the @code{GDB} design philosophy. @code{GDB} is
18047 entirely non-intrusive until a breakpoint is encountered. If no
18048 breakpoint is ever hit, the program will run exactly as it would if no
18049 debugger were present. When a breakpoint is hit, @code{GDB} accesses
18050 the debugging information and can respond to user commands to inspect
18051 variables, and more generally to report on the state of execution.
18053 @node Running GDB,Introduction to GDB Commands,The GNAT Debugger GDB,Running and Debugging Ada Programs
18054 @anchor{gnat_ugn/gnat_and_program_execution id4}@anchor{150}@anchor{gnat_ugn/gnat_and_program_execution running-gdb}@anchor{151}
18055 @subsection Running GDB
18058 This section describes how to initiate the debugger.
18060 The debugger can be launched from a @code{GNAT Studio} menu or
18061 directly from the command line. The description below covers the latter use.
18062 All the commands shown can be used in the @code{GNAT Studio} debug console window,
18063 but there are usually more GUI-based ways to achieve the same effect.
18065 The command to run @code{GDB} is
18074 where @code{program} is the name of the executable file. This
18075 activates the debugger and results in a prompt for debugger commands.
18076 The simplest command is simply @code{run}, which causes the program to run
18077 exactly as if the debugger were not present. The following section
18078 describes some of the additional commands that can be given to @code{GDB}.
18080 @node Introduction to GDB Commands,Using Ada Expressions,Running GDB,Running and Debugging Ada Programs
18081 @anchor{gnat_ugn/gnat_and_program_execution id5}@anchor{152}@anchor{gnat_ugn/gnat_and_program_execution introduction-to-gdb-commands}@anchor{153}
18082 @subsection Introduction to GDB Commands
18085 @code{GDB} contains a large repertoire of commands.
18086 See @cite{Debugging with GDB} for extensive documentation on the use
18087 of these commands, together with examples of their use. Furthermore,
18088 the command @emph{help} invoked from within GDB activates a simple help
18089 facility which summarizes the available commands and their options.
18090 In this section we summarize a few of the most commonly
18091 used commands to give an idea of what @code{GDB} is about. You should create
18092 a simple program with debugging information and experiment with the use of
18093 these @code{GDB} commands on the program as you read through the
18103 @item @code{set args @emph{arguments}}
18105 The @emph{arguments} list above is a list of arguments to be passed to
18106 the program on a subsequent run command, just as though the arguments
18107 had been entered on a normal invocation of the program. The @code{set args}
18108 command is not needed if the program does not require arguments.
18117 The @code{run} command causes execution of the program to start from
18118 the beginning. If the program is already running, that is to say if
18119 you are currently positioned at a breakpoint, then a prompt will ask
18120 for confirmation that you want to abandon the current execution and
18128 @item @code{breakpoint @emph{location}}
18130 The breakpoint command sets a breakpoint, that is to say a point at which
18131 execution will halt and @code{GDB} will await further
18132 commands. @emph{location} is
18133 either a line number within a file, given in the format @code{file:linenumber},
18134 or it is the name of a subprogram. If you request that a breakpoint be set on
18135 a subprogram that is overloaded, a prompt will ask you to specify on which of
18136 those subprograms you want to breakpoint. You can also
18137 specify that all of them should be breakpointed. If the program is run
18138 and execution encounters the breakpoint, then the program
18139 stops and @code{GDB} signals that the breakpoint was encountered by
18140 printing the line of code before which the program is halted.
18147 @item @code{catch exception @emph{name}}
18149 This command causes the program execution to stop whenever exception
18150 @code{name} is raised. If @code{name} is omitted, then the execution is
18151 suspended when any exception is raised.
18158 @item @code{print @emph{expression}}
18160 This will print the value of the given expression. Most simple
18161 Ada expression formats are properly handled by @code{GDB}, so the expression
18162 can contain function calls, variables, operators, and attribute references.
18169 @item @code{continue}
18171 Continues execution following a breakpoint, until the next breakpoint or the
18172 termination of the program.
18181 Executes a single line after a breakpoint. If the next statement
18182 is a subprogram call, execution continues into (the first statement of)
18183 the called subprogram.
18192 Executes a single line. If this line is a subprogram call, executes and
18193 returns from the call.
18202 Lists a few lines around the current source location. In practice, it
18203 is usually more convenient to have a separate edit window open with the
18204 relevant source file displayed. Successive applications of this command
18205 print subsequent lines. The command can be given an argument which is a
18206 line number, in which case it displays a few lines around the specified one.
18213 @item @code{backtrace}
18215 Displays a backtrace of the call chain. This command is typically
18216 used after a breakpoint has occurred, to examine the sequence of calls that
18217 leads to the current breakpoint. The display includes one line for each
18218 activation record (frame) corresponding to an active subprogram.
18227 At a breakpoint, @code{GDB} can display the values of variables local
18228 to the current frame. The command @code{up} can be used to
18229 examine the contents of other active frames, by moving the focus up
18230 the stack, that is to say from callee to caller, one frame at a time.
18239 Moves the focus of @code{GDB} down from the frame currently being
18240 examined to the frame of its callee (the reverse of the previous command),
18247 @item @code{frame @emph{n}}
18249 Inspect the frame with the given number. The value 0 denotes the frame
18250 of the current breakpoint, that is to say the top of the call stack.
18259 Kills the child process in which the program is running under GDB.
18260 This may be useful for several purposes:
18266 It allows you to recompile and relink your program, since on many systems
18267 you cannot regenerate an executable file while it is running in a process.
18270 You can run your program outside the debugger, on systems that do not
18271 permit executing a program outside GDB while breakpoints are set
18275 It allows you to debug a core dump rather than a running process.
18280 The above list is a very short introduction to the commands that
18281 @code{GDB} provides. Important additional capabilities, including conditional
18282 breakpoints, the ability to execute command sequences on a breakpoint,
18283 the ability to debug at the machine instruction level and many other
18284 features are described in detail in @cite{Debugging with GDB}.
18285 Note that most commands can be abbreviated
18286 (for example, c for continue, bt for backtrace).
18288 @node Using Ada Expressions,Calling User-Defined Subprograms,Introduction to GDB Commands,Running and Debugging Ada Programs
18289 @anchor{gnat_ugn/gnat_and_program_execution id6}@anchor{154}@anchor{gnat_ugn/gnat_and_program_execution using-ada-expressions}@anchor{155}
18290 @subsection Using Ada Expressions
18293 @geindex Ada expressions (in gdb)
18295 @code{GDB} supports a fairly large subset of Ada expression syntax, with some
18296 extensions. The philosophy behind the design of this subset is
18304 That @code{GDB} should provide basic literals and access to operations for
18305 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
18306 leaving more sophisticated computations to subprograms written into the
18307 program (which therefore may be called from @code{GDB}).
18310 That type safety and strict adherence to Ada language restrictions
18311 are not particularly relevant in a debugging context.
18314 That brevity is important to the @code{GDB} user.
18318 Thus, for brevity, the debugger acts as if there were
18319 implicit @code{with} and @code{use} clauses in effect for all user-written
18320 packages, thus making it unnecessary to fully qualify most names with
18321 their packages, regardless of context. Where this causes ambiguity,
18322 @code{GDB} asks the user’s intent.
18324 For details on the supported Ada syntax, see @cite{Debugging with GDB}.
18326 @node Calling User-Defined Subprograms,Using the next Command in a Function,Using Ada Expressions,Running and Debugging Ada Programs
18327 @anchor{gnat_ugn/gnat_and_program_execution calling-user-defined-subprograms}@anchor{156}@anchor{gnat_ugn/gnat_and_program_execution id7}@anchor{157}
18328 @subsection Calling User-Defined Subprograms
18331 An important capability of @code{GDB} is the ability to call user-defined
18332 subprograms while debugging. This is achieved simply by entering
18333 a subprogram call statement in the form:
18338 call subprogram-name (parameters)
18342 The keyword @code{call} can be omitted in the normal case where the
18343 @code{subprogram-name} does not coincide with any of the predefined
18344 @code{GDB} commands.
18346 The effect is to invoke the given subprogram, passing it the
18347 list of parameters that is supplied. The parameters can be expressions and
18348 can include variables from the program being debugged. The
18349 subprogram must be defined
18350 at the library level within your program, and @code{GDB} will call the
18351 subprogram within the environment of your program execution (which
18352 means that the subprogram is free to access or even modify variables
18353 within your program).
18355 The most important use of this facility is in allowing the inclusion of
18356 debugging routines that are tailored to particular data structures
18357 in your program. Such debugging routines can be written to provide a suitably
18358 high-level description of an abstract type, rather than a low-level dump
18359 of its physical layout. After all, the standard
18360 @code{GDB print} command only knows the physical layout of your
18361 types, not their abstract meaning. Debugging routines can provide information
18362 at the desired semantic level and are thus enormously useful.
18364 For example, when debugging GNAT itself, it is crucial to have access to
18365 the contents of the tree nodes used to represent the program internally.
18366 But tree nodes are represented simply by an integer value (which in turn
18367 is an index into a table of nodes).
18368 Using the @code{print} command on a tree node would simply print this integer
18369 value, which is not very useful. But the PN routine (defined in file
18370 treepr.adb in the GNAT sources) takes a tree node as input, and displays
18371 a useful high level representation of the tree node, which includes the
18372 syntactic category of the node, its position in the source, the integers
18373 that denote descendant nodes and parent node, as well as varied
18374 semantic information. To study this example in more detail, you might want to
18375 look at the body of the PN procedure in the stated file.
18377 Another useful application of this capability is to deal with situations of
18378 complex data which are not handled suitably by GDB. For example, if you specify
18379 Convention Fortran for a multi-dimensional array, GDB does not know that
18380 the ordering of array elements has been switched and will not properly
18381 address the array elements. In such a case, instead of trying to print the
18382 elements directly from GDB, you can write a callable procedure that prints
18383 the elements in the desired format.
18385 @node Using the next Command in a Function,Stopping When Ada Exceptions Are Raised,Calling User-Defined Subprograms,Running and Debugging Ada Programs
18386 @anchor{gnat_ugn/gnat_and_program_execution id8}@anchor{158}@anchor{gnat_ugn/gnat_and_program_execution using-the-next-command-in-a-function}@anchor{159}
18387 @subsection Using the @emph{next} Command in a Function
18390 When you use the @code{next} command in a function, the current source
18391 location will advance to the next statement as usual. A special case
18392 arises in the case of a @code{return} statement.
18394 Part of the code for a return statement is the ‘epilogue’ of the function.
18395 This is the code that returns to the caller. There is only one copy of
18396 this epilogue code, and it is typically associated with the last return
18397 statement in the function if there is more than one return. In some
18398 implementations, this epilogue is associated with the first statement
18401 The result is that if you use the @code{next} command from a return
18402 statement that is not the last return statement of the function you
18403 may see a strange apparent jump to the last return statement or to
18404 the start of the function. You should simply ignore this odd jump.
18405 The value returned is always that from the first return statement
18406 that was stepped through.
18408 @node Stopping When Ada Exceptions Are Raised,Ada Tasks,Using the next Command in a Function,Running and Debugging Ada Programs
18409 @anchor{gnat_ugn/gnat_and_program_execution id9}@anchor{15a}@anchor{gnat_ugn/gnat_and_program_execution stopping-when-ada-exceptions-are-raised}@anchor{15b}
18410 @subsection Stopping When Ada Exceptions Are Raised
18413 @geindex Exceptions (in gdb)
18415 You can set catchpoints that stop the program execution when your program
18416 raises selected exceptions.
18425 @item @code{catch exception}
18427 Set a catchpoint that stops execution whenever (any task in the) program
18428 raises any exception.
18435 @item @code{catch exception @emph{name}}
18437 Set a catchpoint that stops execution whenever (any task in the) program
18438 raises the exception @emph{name}.
18445 @item @code{catch exception unhandled}
18447 Set a catchpoint that stops executing whenever (any task in the) program
18448 raises an exception for which there is no handler.
18455 @item @code{info exceptions}, @code{info exceptions @emph{regexp}}
18457 The @code{info exceptions} command permits the user to examine all defined
18458 exceptions within Ada programs. With a regular expression, @emph{regexp}, as
18459 argument, prints out only those exceptions whose name matches @emph{regexp}.
18463 @geindex Tasks (in gdb)
18465 @node Ada Tasks,Debugging Generic Units,Stopping When Ada Exceptions Are Raised,Running and Debugging Ada Programs
18466 @anchor{gnat_ugn/gnat_and_program_execution ada-tasks}@anchor{15c}@anchor{gnat_ugn/gnat_and_program_execution id10}@anchor{15d}
18467 @subsection Ada Tasks
18470 @code{GDB} allows the following task-related commands:
18479 @item @code{info tasks}
18481 This command shows a list of current Ada tasks, as in the following example:
18485 ID TID P-ID Thread Pri State Name
18486 1 8088000 0 807e000 15 Child Activation Wait main_task
18487 2 80a4000 1 80ae000 15 Accept/Select Wait b
18488 3 809a800 1 80a4800 15 Child Activation Wait a
18489 * 4 80ae800 3 80b8000 15 Running c
18492 In this listing, the asterisk before the first task indicates it to be the
18493 currently running task. The first column lists the task ID that is used
18494 to refer to tasks in the following commands.
18498 @geindex Breakpoints and tasks
18504 @code{break`@w{`}*linespec* `@w{`}task} @emph{taskid}, @code{break} @emph{linespec} @code{task} @emph{taskid} @code{if} …
18508 These commands are like the @code{break ... thread ...}.
18509 @emph{linespec} specifies source lines.
18511 Use the qualifier @code{task @emph{taskid}} with a breakpoint command
18512 to specify that you only want @code{GDB} to stop the program when a
18513 particular Ada task reaches this breakpoint. @emph{taskid} is one of the
18514 numeric task identifiers assigned by @code{GDB}, shown in the first
18515 column of the @code{info tasks} display.
18517 If you do not specify @code{task @emph{taskid}} when you set a
18518 breakpoint, the breakpoint applies to @emph{all} tasks of your
18521 You can use the @code{task} qualifier on conditional breakpoints as
18522 well; in this case, place @code{task @emph{taskid}} before the
18523 breakpoint condition (before the @code{if}).
18527 @geindex Task switching (in gdb)
18533 @code{task @emph{taskno}}
18537 This command allows switching to the task referred by @emph{taskno}. In
18538 particular, this allows browsing of the backtrace of the specified
18539 task. It is advisable to switch back to the original task before
18540 continuing execution otherwise the scheduling of the program may be
18545 For more detailed information on the tasking support,
18546 see @cite{Debugging with GDB}.
18548 @geindex Debugging Generic Units
18552 @node Debugging Generic Units,Remote Debugging with gdbserver,Ada Tasks,Running and Debugging Ada Programs
18553 @anchor{gnat_ugn/gnat_and_program_execution debugging-generic-units}@anchor{15e}@anchor{gnat_ugn/gnat_and_program_execution id11}@anchor{15f}
18554 @subsection Debugging Generic Units
18557 GNAT always uses code expansion for generic instantiation. This means that
18558 each time an instantiation occurs, a complete copy of the original code is
18559 made, with appropriate substitutions of formals by actuals.
18561 It is not possible to refer to the original generic entities in
18562 @code{GDB}, but it is always possible to debug a particular instance of
18563 a generic, by using the appropriate expanded names. For example, if we have
18570 generic package k is
18571 procedure kp (v1 : in out integer);
18575 procedure kp (v1 : in out integer) is
18581 package k1 is new k;
18582 package k2 is new k;
18584 var : integer := 1;
18595 Then to break on a call to procedure kp in the k2 instance, simply
18601 (gdb) break g.k2.kp
18605 When the breakpoint occurs, you can step through the code of the
18606 instance in the normal manner and examine the values of local variables, as for
18609 @geindex Remote Debugging with gdbserver
18611 @node Remote Debugging with gdbserver,GNAT Abnormal Termination or Failure to Terminate,Debugging Generic Units,Running and Debugging Ada Programs
18612 @anchor{gnat_ugn/gnat_and_program_execution id12}@anchor{160}@anchor{gnat_ugn/gnat_and_program_execution remote-debugging-with-gdbserver}@anchor{161}
18613 @subsection Remote Debugging with gdbserver
18616 On platforms where gdbserver is supported, it is possible to use this tool
18617 to debug your application remotely. This can be useful in situations
18618 where the program needs to be run on a target host that is different
18619 from the host used for development, particularly when the target has
18620 a limited amount of resources (either CPU and/or memory).
18622 To do so, start your program using gdbserver on the target machine.
18623 gdbserver then automatically suspends the execution of your program
18624 at its entry point, waiting for a debugger to connect to it. The
18625 following commands starts an application and tells gdbserver to
18626 wait for a connection with the debugger on localhost port 4444.
18631 $ gdbserver localhost:4444 program
18632 Process program created; pid = 5685
18633 Listening on port 4444
18637 Once gdbserver has started listening, we can tell the debugger to establish
18638 a connection with this gdbserver, and then start the same debugging session
18639 as if the program was being debugged on the same host, directly under
18640 the control of GDB.
18646 (gdb) target remote targethost:4444
18647 Remote debugging using targethost:4444
18648 0x00007f29936d0af0 in ?? () from /lib64/ld-linux-x86-64.so.
18650 Breakpoint 1 at 0x401f0c: file foo.adb, line 3.
18654 Breakpoint 1, foo () at foo.adb:4
18659 It is also possible to use gdbserver to attach to an already running
18660 program, in which case the execution of that program is simply suspended
18661 until the connection between the debugger and gdbserver is established.
18663 For more information on how to use gdbserver, see the @emph{Using the gdbserver Program}
18664 section in @cite{Debugging with GDB}.
18665 GNAT provides support for gdbserver on x86-linux, x86-windows and x86_64-linux.
18667 @geindex Abnormal Termination or Failure to Terminate
18669 @node GNAT Abnormal Termination or Failure to Terminate,Naming Conventions for GNAT Source Files,Remote Debugging with gdbserver,Running and Debugging Ada Programs
18670 @anchor{gnat_ugn/gnat_and_program_execution gnat-abnormal-termination-or-failure-to-terminate}@anchor{162}@anchor{gnat_ugn/gnat_and_program_execution id13}@anchor{163}
18671 @subsection GNAT Abnormal Termination or Failure to Terminate
18674 When presented with programs that contain serious errors in syntax
18676 GNAT may on rare occasions experience problems in operation, such
18678 segmentation fault or illegal memory access, raising an internal
18679 exception, terminating abnormally, or failing to terminate at all.
18680 In such cases, you can activate
18681 various features of GNAT that can help you pinpoint the construct in your
18682 program that is the likely source of the problem.
18684 The following strategies are presented in increasing order of
18685 difficulty, corresponding to your experience in using GNAT and your
18686 familiarity with compiler internals.
18692 Run @code{gcc} with the @code{-gnatf}. This first
18693 switch causes all errors on a given line to be reported. In its absence,
18694 only the first error on a line is displayed.
18696 The @code{-gnatdO} switch causes errors to be displayed as soon as they
18697 are encountered, rather than after compilation is terminated. If GNAT
18698 terminates prematurely or goes into an infinite loop, the last error
18699 message displayed may help to pinpoint the culprit.
18702 Run @code{gcc} with the @code{-v} (verbose) switch. In this
18703 mode, @code{gcc} produces ongoing information about the progress of the
18704 compilation and provides the name of each procedure as code is
18705 generated. This switch allows you to find which Ada procedure was being
18706 compiled when it encountered a code generation problem.
18709 @geindex -gnatdc switch
18715 Run @code{gcc} with the @code{-gnatdc} switch. This is a GNAT specific
18716 switch that does for the front-end what @code{-v} does
18717 for the back end. The system prints the name of each unit,
18718 either a compilation unit or nested unit, as it is being analyzed.
18721 Finally, you can start
18722 @code{gdb} directly on the @code{gnat1} executable. @code{gnat1} is the
18723 front-end of GNAT, and can be run independently (normally it is just
18724 called from @code{gcc}). You can use @code{gdb} on @code{gnat1} as you
18725 would on a C program (but @ref{14f,,The GNAT Debugger GDB} for caveats). The
18726 @code{where} command is the first line of attack; the variable
18727 @code{lineno} (seen by @code{print lineno}), used by the second phase of
18728 @code{gnat1} and by the @code{gcc} backend, indicates the source line at
18729 which the execution stopped, and @code{input_file name} indicates the name of
18733 @node Naming Conventions for GNAT Source Files,Getting Internal Debugging Information,GNAT Abnormal Termination or Failure to Terminate,Running and Debugging Ada Programs
18734 @anchor{gnat_ugn/gnat_and_program_execution id14}@anchor{164}@anchor{gnat_ugn/gnat_and_program_execution naming-conventions-for-gnat-source-files}@anchor{165}
18735 @subsection Naming Conventions for GNAT Source Files
18738 In order to examine the workings of the GNAT system, the following
18739 brief description of its organization may be helpful:
18745 Files with prefix @code{sc} contain the lexical scanner.
18748 All files prefixed with @code{par} are components of the parser. The
18749 numbers correspond to chapters of the Ada Reference Manual. For example,
18750 parsing of select statements can be found in @code{par-ch9.adb}.
18753 All files prefixed with @code{sem} perform semantic analysis. The
18754 numbers correspond to chapters of the Ada standard. For example, all
18755 issues involving context clauses can be found in @code{sem_ch10.adb}. In
18756 addition, some features of the language require sufficient special processing
18757 to justify their own semantic files: sem_aggr for aggregates, sem_disp for
18758 dynamic dispatching, etc.
18761 All files prefixed with @code{exp} perform normalization and
18762 expansion of the intermediate representation (abstract syntax tree, or AST).
18763 these files use the same numbering scheme as the parser and semantics files.
18764 For example, the construction of record initialization procedures is done in
18765 @code{exp_ch3.adb}.
18768 The files prefixed with @code{bind} implement the binder, which
18769 verifies the consistency of the compilation, determines an order of
18770 elaboration, and generates the bind file.
18773 The files @code{atree.ads} and @code{atree.adb} detail the low-level
18774 data structures used by the front-end.
18777 The files @code{sinfo.ads} and @code{sinfo.adb} detail the structure of
18778 the abstract syntax tree as produced by the parser.
18781 The files @code{einfo.ads} and @code{einfo.adb} detail the attributes of
18782 all entities, computed during semantic analysis.
18785 Library management issues are dealt with in files with prefix
18788 @geindex Annex A (in Ada Reference Manual)
18791 Ada files with the prefix @code{a-} are children of @code{Ada}, as
18792 defined in Annex A.
18794 @geindex Annex B (in Ada reference Manual)
18797 Files with prefix @code{i-} are children of @code{Interfaces}, as
18798 defined in Annex B.
18800 @geindex System (package in Ada Reference Manual)
18803 Files with prefix @code{s-} are children of @code{System}. This includes
18804 both language-defined children and GNAT run-time routines.
18806 @geindex GNAT (package)
18809 Files with prefix @code{g-} are children of @code{GNAT}. These are useful
18810 general-purpose packages, fully documented in their specs. All
18811 the other @code{.c} files are modifications of common @code{gcc} files.
18814 @node Getting Internal Debugging Information,Stack Traceback,Naming Conventions for GNAT Source Files,Running and Debugging Ada Programs
18815 @anchor{gnat_ugn/gnat_and_program_execution getting-internal-debugging-information}@anchor{166}@anchor{gnat_ugn/gnat_and_program_execution id15}@anchor{167}
18816 @subsection Getting Internal Debugging Information
18819 Most compilers have internal debugging switches and modes. GNAT
18820 does also, except GNAT internal debugging switches and modes are not
18821 secret. A summary and full description of all the compiler and binder
18822 debug flags are in the file @code{debug.adb}. You must obtain the
18823 sources of the compiler to see the full detailed effects of these flags.
18825 The switches that print the source of the program (reconstructed from
18826 the internal tree) are of general interest for user programs, as are the
18828 the full internal tree, and the entity table (the symbol table
18829 information). The reconstructed source provides a readable version of the
18830 program after the front-end has completed analysis and expansion,
18831 and is useful when studying the performance of specific constructs.
18832 For example, constraint checks are indicated, complex aggregates
18833 are replaced with loops and assignments, and tasking primitives
18834 are replaced with run-time calls.
18838 @geindex stack traceback
18840 @geindex stack unwinding
18842 @node Stack Traceback,Pretty-Printers for the GNAT runtime,Getting Internal Debugging Information,Running and Debugging Ada Programs
18843 @anchor{gnat_ugn/gnat_and_program_execution id16}@anchor{168}@anchor{gnat_ugn/gnat_and_program_execution stack-traceback}@anchor{169}
18844 @subsection Stack Traceback
18847 Traceback is a mechanism to display the sequence of subprogram calls that
18848 leads to a specified execution point in a program. Often (but not always)
18849 the execution point is an instruction at which an exception has been raised.
18850 This mechanism is also known as @emph{stack unwinding} because it obtains
18851 its information by scanning the run-time stack and recovering the activation
18852 records of all active subprograms. Stack unwinding is one of the most
18853 important tools for program debugging.
18855 The first entry stored in traceback corresponds to the deepest calling level,
18856 that is to say the subprogram currently executing the instruction
18857 from which we want to obtain the traceback.
18859 Note that there is no runtime performance penalty when stack traceback
18860 is enabled, and no exception is raised during program execution.
18863 @geindex non-symbolic
18866 * Non-Symbolic Traceback::
18867 * Symbolic Traceback::
18871 @node Non-Symbolic Traceback,Symbolic Traceback,,Stack Traceback
18872 @anchor{gnat_ugn/gnat_and_program_execution id17}@anchor{16a}@anchor{gnat_ugn/gnat_and_program_execution non-symbolic-traceback}@anchor{16b}
18873 @subsubsection Non-Symbolic Traceback
18876 Note: this feature is not supported on all platforms. See
18877 @code{GNAT.Traceback} spec in @code{g-traceb.ads}
18878 for a complete list of supported platforms.
18880 @subsubheading Tracebacks From an Unhandled Exception
18883 A runtime non-symbolic traceback is a list of addresses of call instructions.
18884 To enable this feature you must use the @code{-E}
18885 @code{gnatbind} option. With this option a stack traceback is stored as part
18886 of exception information. You can retrieve this information using the
18887 @code{addr2line} tool.
18889 Here is a simple example:
18898 raise Constraint_Error;
18912 $ gnatmake stb -bargs -E
18915 Execution terminated by unhandled exception
18916 Exception name: CONSTRAINT_ERROR
18918 Call stack traceback locations:
18919 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
18923 As we see the traceback lists a sequence of addresses for the unhandled
18924 exception @code{CONSTRAINT_ERROR} raised in procedure P1. It is easy to
18925 guess that this exception come from procedure P1. To translate these
18926 addresses into the source lines where the calls appear, the
18927 @code{addr2line} tool, described below, is invaluable. The use of this tool
18928 requires the program to be compiled with debug information.
18933 $ gnatmake -g stb -bargs -E
18936 Execution terminated by unhandled exception
18937 Exception name: CONSTRAINT_ERROR
18939 Call stack traceback locations:
18940 0x401373 0x40138b 0x40139c 0x401335 0x4011c4 0x4011f1 0x77e892a4
18942 $ addr2line --exe=stb 0x401373 0x40138b 0x40139c 0x401335 0x4011c4
18943 0x4011f1 0x77e892a4
18945 00401373 at d:/stb/stb.adb:5
18946 0040138B at d:/stb/stb.adb:10
18947 0040139C at d:/stb/stb.adb:14
18948 00401335 at d:/stb/b~stb.adb:104
18949 004011C4 at /build/.../crt1.c:200
18950 004011F1 at /build/.../crt1.c:222
18951 77E892A4 in ?? at ??:0
18955 The @code{addr2line} tool has several other useful options:
18960 @multitable {xxxxxxxxxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
18967 to get the function name corresponding to any location
18971 @code{--demangle=gnat}
18975 to use the gnat decoding mode for the function names.
18976 Note that for binutils version 2.9.x the option is
18977 simply @code{--demangle}.
18983 $ addr2line --exe=stb --functions --demangle=gnat 0x401373 0x40138b
18984 0x40139c 0x401335 0x4011c4 0x4011f1
18986 00401373 in stb.p1 at d:/stb/stb.adb:5
18987 0040138B in stb.p2 at d:/stb/stb.adb:10
18988 0040139C in stb at d:/stb/stb.adb:14
18989 00401335 in main at d:/stb/b~stb.adb:104
18990 004011C4 in <__mingw_CRTStartup> at /build/.../crt1.c:200
18991 004011F1 in <mainCRTStartup> at /build/.../crt1.c:222
18995 From this traceback we can see that the exception was raised in
18996 @code{stb.adb} at line 5, which was reached from a procedure call in
18997 @code{stb.adb} at line 10, and so on. The @code{b~std.adb} is the binder file,
18998 which contains the call to the main program.
18999 @ref{10e,,Running gnatbind}. The remaining entries are assorted runtime routines,
19000 and the output will vary from platform to platform.
19002 It is also possible to use @code{GDB} with these traceback addresses to debug
19003 the program. For example, we can break at a given code location, as reported
19004 in the stack traceback:
19013 Furthermore, this feature is not implemented inside Windows DLL. Only
19014 the non-symbolic traceback is reported in this case.
19019 (gdb) break *0x401373
19020 Breakpoint 1 at 0x401373: file stb.adb, line 5.
19024 It is important to note that the stack traceback addresses
19025 do not change when debug information is included. This is particularly useful
19026 because it makes it possible to release software without debug information (to
19027 minimize object size), get a field report that includes a stack traceback
19028 whenever an internal bug occurs, and then be able to retrieve the sequence
19029 of calls with the same program compiled with debug information.
19031 @subsubheading Tracebacks From Exception Occurrences
19034 Non-symbolic tracebacks are obtained by using the @code{-E} binder argument.
19035 The stack traceback is attached to the exception information string, and can
19036 be retrieved in an exception handler within the Ada program, by means of the
19037 Ada facilities defined in @code{Ada.Exceptions}. Here is a simple example:
19043 with Ada.Exceptions;
19048 use Ada.Exceptions;
19056 Text_IO.Put_Line (Exception_Information (E));
19070 This program will output:
19077 Exception name: CONSTRAINT_ERROR
19078 Message: stb.adb:12
19079 Call stack traceback locations:
19080 0x4015e4 0x401633 0x401644 0x401461 0x4011c4 0x4011f1 0x77e892a4
19084 @subsubheading Tracebacks From Anywhere in a Program
19087 It is also possible to retrieve a stack traceback from anywhere in a
19088 program. For this you need to
19089 use the @code{GNAT.Traceback} API. This package includes a procedure called
19090 @code{Call_Chain} that computes a complete stack traceback, as well as useful
19091 display procedures described below. It is not necessary to use the
19092 @code{-E} @code{gnatbind} option in this case, because the stack traceback mechanism
19093 is invoked explicitly.
19095 In the following example we compute a traceback at a specific location in
19096 the program, and we display it using @code{GNAT.Debug_Utilities.Image} to
19097 convert addresses to strings:
19103 with GNAT.Traceback;
19104 with GNAT.Debug_Utilities;
19110 use GNAT.Traceback;
19113 TB : Tracebacks_Array (1 .. 10);
19114 -- We are asking for a maximum of 10 stack frames.
19116 -- Len will receive the actual number of stack frames returned.
19118 Call_Chain (TB, Len);
19120 Text_IO.Put ("In STB.P1 : ");
19122 for K in 1 .. Len loop
19123 Text_IO.Put (Debug_Utilities.Image (TB (K)));
19144 In STB.P1 : 16#0040_F1E4# 16#0040_14F2# 16#0040_170B# 16#0040_171C#
19145 16#0040_1461# 16#0040_11C4# 16#0040_11F1# 16#77E8_92A4#
19149 You can then get further information by invoking the @code{addr2line}
19150 tool as described earlier (note that the hexadecimal addresses
19151 need to be specified in C format, with a leading ‘0x’).
19156 @node Symbolic Traceback,,Non-Symbolic Traceback,Stack Traceback
19157 @anchor{gnat_ugn/gnat_and_program_execution id18}@anchor{16c}@anchor{gnat_ugn/gnat_and_program_execution symbolic-traceback}@anchor{16d}
19158 @subsubsection Symbolic Traceback
19161 A symbolic traceback is a stack traceback in which procedure names are
19162 associated with each code location.
19164 Note that this feature is not supported on all platforms. See
19165 @code{GNAT.Traceback.Symbolic} spec in @code{g-trasym.ads} for a complete
19166 list of currently supported platforms.
19168 Note that the symbolic traceback requires that the program be compiled
19169 with debug information. If it is not compiled with debug information
19170 only the non-symbolic information will be valid.
19172 @subsubheading Tracebacks From Exception Occurrences
19175 Here is an example:
19181 with GNAT.Traceback.Symbolic;
19187 raise Constraint_Error;
19204 Ada.Text_IO.Put_Line (GNAT.Traceback.Symbolic.Symbolic_Traceback (E));
19209 $ gnatmake -g .\stb -bargs -E
19212 0040149F in stb.p1 at stb.adb:8
19213 004014B7 in stb.p2 at stb.adb:13
19214 004014CF in stb.p3 at stb.adb:18
19215 004015DD in ada.stb at stb.adb:22
19216 00401461 in main at b~stb.adb:168
19217 004011C4 in __mingw_CRTStartup at crt1.c:200
19218 004011F1 in mainCRTStartup at crt1.c:222
19219 77E892A4 in ?? at ??:0
19223 In the above example the @code{.\} syntax in the @code{gnatmake} command
19224 is currently required by @code{addr2line} for files that are in
19225 the current working directory.
19226 Moreover, the exact sequence of linker options may vary from platform
19228 The above @code{-largs} section is for Windows platforms. By contrast,
19229 under Unix there is no need for the @code{-largs} section.
19230 Differences across platforms are due to details of linker implementation.
19232 @subsubheading Tracebacks From Anywhere in a Program
19235 It is possible to get a symbolic stack traceback
19236 from anywhere in a program, just as for non-symbolic tracebacks.
19237 The first step is to obtain a non-symbolic
19238 traceback, and then call @code{Symbolic_Traceback} to compute the symbolic
19239 information. Here is an example:
19245 with GNAT.Traceback;
19246 with GNAT.Traceback.Symbolic;
19251 use GNAT.Traceback;
19252 use GNAT.Traceback.Symbolic;
19255 TB : Tracebacks_Array (1 .. 10);
19256 -- We are asking for a maximum of 10 stack frames.
19258 -- Len will receive the actual number of stack frames returned.
19260 Call_Chain (TB, Len);
19261 Text_IO.Put_Line (Symbolic_Traceback (TB (1 .. Len)));
19275 @subsubheading Automatic Symbolic Tracebacks
19278 Symbolic tracebacks may also be enabled by using the -Es switch to gnatbind (as
19279 in @code{gprbuild -g ... -bargs -Es}).
19280 This will cause the Exception_Information to contain a symbolic traceback,
19281 which will also be printed if an unhandled exception terminates the
19284 @node Pretty-Printers for the GNAT runtime,,Stack Traceback,Running and Debugging Ada Programs
19285 @anchor{gnat_ugn/gnat_and_program_execution id19}@anchor{16e}@anchor{gnat_ugn/gnat_and_program_execution pretty-printers-for-the-gnat-runtime}@anchor{16f}
19286 @subsection Pretty-Printers for the GNAT runtime
19289 As discussed in @cite{Calling User-Defined Subprograms}, GDB’s
19290 @code{print} command only knows about the physical layout of program data
19291 structures and therefore normally displays only low-level dumps, which
19292 are often hard to understand.
19294 An example of this is when trying to display the contents of an Ada
19295 standard container, such as @code{Ada.Containers.Ordered_Maps.Map}:
19300 with Ada.Containers.Ordered_Maps;
19303 package Int_To_Nat is
19304 new Ada.Containers.Ordered_Maps (Integer, Natural);
19306 Map : Int_To_Nat.Map;
19308 Map.Insert (1, 10);
19309 Map.Insert (2, 20);
19310 Map.Insert (3, 30);
19312 Map.Clear; -- BREAK HERE
19317 When this program is built with debugging information and run under
19318 GDB up to the @code{Map.Clear} statement, trying to print @code{Map} will
19319 yield information that is only relevant to the developers of our standard
19341 Fortunately, GDB has a feature called pretty-printers@footnote{http://docs.adacore.com/gdb-docs/html/gdb.html#Pretty_002dPrinter-Introduction},
19342 which allows customizing how GDB displays data structures. The GDB
19343 shipped with GNAT embeds such pretty-printers for the most common
19344 containers in the standard library. To enable them, either run the
19345 following command manually under GDB or add it to your @code{.gdbinit} file:
19350 python import gnatdbg; gnatdbg.setup()
19354 Once this is done, GDB’s @code{print} command will automatically use
19355 these pretty-printers when appropriate. Using the previous example:
19361 $1 = pp.int_to_nat.map of length 3 = @{
19369 Pretty-printers are invoked each time GDB tries to display a value,
19370 including when displaying the arguments of a called subprogram (in
19371 GDB’s @code{backtrace} command) or when printing the value returned by a
19372 function (in GDB’s @code{finish} command).
19374 To display a value without involving pretty-printers, @code{print} can be
19375 invoked with its @code{/r} option:
19386 Finer control of pretty-printers is also possible: see GDB's online documentation@footnote{http://docs.adacore.com/gdb-docs/html/gdb.html#Pretty_002dPrinter-Commands}
19387 for more information.
19391 @node Profiling,Improving Performance,Running and Debugging Ada Programs,GNAT and Program Execution
19392 @anchor{gnat_ugn/gnat_and_program_execution id20}@anchor{170}@anchor{gnat_ugn/gnat_and_program_execution profiling}@anchor{147}
19396 This section describes how to use the @code{gprof} profiler tool on Ada programs.
19403 * Profiling an Ada Program with gprof::
19407 @node Profiling an Ada Program with gprof,,,Profiling
19408 @anchor{gnat_ugn/gnat_and_program_execution id21}@anchor{171}@anchor{gnat_ugn/gnat_and_program_execution profiling-an-ada-program-with-gprof}@anchor{172}
19409 @subsection Profiling an Ada Program with gprof
19412 This section is not meant to be an exhaustive documentation of @code{gprof}.
19413 Full documentation for it can be found in the @cite{GNU Profiler User’s Guide}
19414 documentation that is part of this GNAT distribution.
19416 Profiling a program helps determine the parts of a program that are executed
19417 most often, and are therefore the most time-consuming.
19419 @code{gprof} is the standard GNU profiling tool; it has been enhanced to
19420 better handle Ada programs and multitasking.
19421 It is currently supported on the following platforms
19433 In order to profile a program using @code{gprof}, several steps are needed:
19439 Instrument the code, which requires a full recompilation of the project with the
19443 Execute the program under the analysis conditions, i.e. with the desired
19447 Analyze the results using the @code{gprof} tool.
19450 The following sections detail the different steps, and indicate how
19451 to interpret the results.
19454 * Compilation for profiling::
19455 * Program execution::
19457 * Interpretation of profiling results::
19461 @node Compilation for profiling,Program execution,,Profiling an Ada Program with gprof
19462 @anchor{gnat_ugn/gnat_and_program_execution compilation-for-profiling}@anchor{173}@anchor{gnat_ugn/gnat_and_program_execution id22}@anchor{174}
19463 @subsubsection Compilation for profiling
19467 @geindex for profiling
19469 @geindex -pg (gnatlink)
19470 @geindex for profiling
19472 In order to profile a program the first step is to tell the compiler
19473 to generate the necessary profiling information. The compiler switch to be used
19474 is @code{-pg}, which must be added to other compilation switches. This
19475 switch needs to be specified both during compilation and link stages, and can
19476 be specified once when using gnatmake:
19481 $ gnatmake -f -pg -P my_project
19485 Note that only the objects that were compiled with the @code{-pg} switch will
19486 be profiled; if you need to profile your whole project, use the @code{-f}
19487 gnatmake switch to force full recompilation.
19489 @node Program execution,Running gprof,Compilation for profiling,Profiling an Ada Program with gprof
19490 @anchor{gnat_ugn/gnat_and_program_execution id23}@anchor{175}@anchor{gnat_ugn/gnat_and_program_execution program-execution}@anchor{176}
19491 @subsubsection Program execution
19494 Once the program has been compiled for profiling, you can run it as usual.
19496 The only constraint imposed by profiling is that the program must terminate
19497 normally. An interrupted program (via a Ctrl-C, kill, etc.) will not be
19500 Once the program completes execution, a data file called @code{gmon.out} is
19501 generated in the directory where the program was launched from. If this file
19502 already exists, it will be overwritten.
19504 @node Running gprof,Interpretation of profiling results,Program execution,Profiling an Ada Program with gprof
19505 @anchor{gnat_ugn/gnat_and_program_execution id24}@anchor{177}@anchor{gnat_ugn/gnat_and_program_execution running-gprof}@anchor{178}
19506 @subsubsection Running gprof
19509 The @code{gprof} tool is called as follow:
19514 $ gprof my_prog gmon.out
19527 The complete form of the gprof command line is the following:
19532 $ gprof [switches] [executable [data-file]]
19536 @code{gprof} supports numerous switches. The order of these
19537 switch does not matter. The full list of options can be found in
19538 the GNU Profiler User’s Guide documentation that comes with this documentation.
19540 The following is the subset of those switches that is most relevant:
19542 @geindex --demangle (gprof)
19547 @item @code{--demangle[=@emph{style}]}, @code{--no-demangle}
19549 These options control whether symbol names should be demangled when
19550 printing output. The default is to demangle C++ symbols. The
19551 @code{--no-demangle} option may be used to turn off demangling. Different
19552 compilers have different mangling styles. The optional demangling style
19553 argument can be used to choose an appropriate demangling style for your
19554 compiler, in particular Ada symbols generated by GNAT can be demangled using
19555 @code{--demangle=gnat}.
19558 @geindex -e (gprof)
19563 @item @code{-e @emph{function_name}}
19565 The @code{-e @emph{function}} option tells @code{gprof} not to print
19566 information about the function @code{function_name} (and its
19567 children…) in the call graph. The function will still be listed
19568 as a child of any functions that call it, but its index number will be
19569 shown as @code{[not printed]}. More than one @code{-e} option may be
19570 given; only one @code{function_name} may be indicated with each @code{-e}
19574 @geindex -E (gprof)
19579 @item @code{-E @emph{function_name}}
19581 The @code{-E @emph{function}} option works like the @code{-e} option, but
19582 execution time spent in the function (and children who were not called from
19583 anywhere else), will not be used to compute the percentages-of-time for
19584 the call graph. More than one @code{-E} option may be given; only one
19585 @code{function_name} may be indicated with each @code{-E`} option.
19588 @geindex -f (gprof)
19593 @item @code{-f @emph{function_name}}
19595 The @code{-f @emph{function}} option causes @code{gprof} to limit the
19596 call graph to the function @code{function_name} and its children (and
19597 their children…). More than one @code{-f} option may be given;
19598 only one @code{function_name} may be indicated with each @code{-f}
19602 @geindex -F (gprof)
19607 @item @code{-F @emph{function_name}}
19609 The @code{-F @emph{function}} option works like the @code{-f} option, but
19610 only time spent in the function and its children (and their
19611 children…) will be used to determine total-time and
19612 percentages-of-time for the call graph. More than one @code{-F} option
19613 may be given; only one @code{function_name} may be indicated with each
19614 @code{-F} option. The @code{-F} option overrides the @code{-E} option.
19617 @node Interpretation of profiling results,,Running gprof,Profiling an Ada Program with gprof
19618 @anchor{gnat_ugn/gnat_and_program_execution id25}@anchor{179}@anchor{gnat_ugn/gnat_and_program_execution interpretation-of-profiling-results}@anchor{17a}
19619 @subsubsection Interpretation of profiling results
19622 The results of the profiling analysis are represented by two arrays: the
19623 ‘flat profile’ and the ‘call graph’. Full documentation of those outputs
19624 can be found in the GNU Profiler User’s Guide.
19626 The flat profile shows the time spent in each function of the program, and how
19627 many time it has been called. This allows you to locate easily the most
19628 time-consuming functions.
19630 The call graph shows, for each subprogram, the subprograms that call it,
19631 and the subprograms that it calls. It also provides an estimate of the time
19632 spent in each of those callers/called subprograms.
19634 @node Improving Performance,Overflow Check Handling in GNAT,Profiling,GNAT and Program Execution
19635 @anchor{gnat_ugn/gnat_and_program_execution id26}@anchor{148}@anchor{gnat_ugn/gnat_and_program_execution improving-performance}@anchor{17b}
19636 @section Improving Performance
19639 @geindex Improving performance
19641 This section presents several topics related to program performance.
19642 It first describes some of the tradeoffs that need to be considered
19643 and some of the techniques for making your program run faster.
19645 It then documents the unused subprogram/data elimination feature,
19646 which can reduce the size of program executables.
19649 * Performance Considerations::
19650 * Text_IO Suggestions::
19651 * Reducing Size of Executables with Unused Subprogram/Data Elimination::
19655 @node Performance Considerations,Text_IO Suggestions,,Improving Performance
19656 @anchor{gnat_ugn/gnat_and_program_execution id27}@anchor{17c}@anchor{gnat_ugn/gnat_and_program_execution performance-considerations}@anchor{17d}
19657 @subsection Performance Considerations
19660 The GNAT system provides a number of options that allow a trade-off
19667 performance of the generated code
19670 speed of compilation
19673 minimization of dependences and recompilation
19676 the degree of run-time checking.
19679 The defaults (if no options are selected) aim at improving the speed
19680 of compilation and minimizing dependences, at the expense of performance
19681 of the generated code:
19690 no inlining of subprogram calls
19693 all run-time checks enabled except overflow and elaboration checks
19696 These options are suitable for most program development purposes. This
19697 section describes how you can modify these choices, and also provides
19698 some guidelines on debugging optimized code.
19701 * Controlling Run-Time Checks::
19702 * Use of Restrictions::
19703 * Optimization Levels::
19704 * Debugging Optimized Code::
19705 * Inlining of Subprograms::
19706 * Floating Point Operations::
19707 * Vectorization of loops::
19708 * Other Optimization Switches::
19709 * Optimization and Strict Aliasing::
19710 * Aliased Variables and Optimization::
19711 * Atomic Variables and Optimization::
19712 * Passive Task Optimization::
19716 @node Controlling Run-Time Checks,Use of Restrictions,,Performance Considerations
19717 @anchor{gnat_ugn/gnat_and_program_execution controlling-run-time-checks}@anchor{17e}@anchor{gnat_ugn/gnat_and_program_execution id28}@anchor{17f}
19718 @subsubsection Controlling Run-Time Checks
19721 By default, GNAT generates all run-time checks, except stack overflow
19722 checks, and checks for access before elaboration on subprogram
19723 calls. The latter are not required in default mode, because all
19724 necessary checking is done at compile time.
19726 @geindex -gnatp (gcc)
19728 @geindex -gnato (gcc)
19730 The gnat switch, @code{-gnatp} allows this default to be modified. See
19731 @ref{ea,,Run-Time Checks}.
19733 Our experience is that the default is suitable for most development
19736 Elaboration checks are off by default, and also not needed by default, since
19737 GNAT uses a static elaboration analysis approach that avoids the need for
19738 run-time checking. This manual contains a full chapter discussing the issue
19739 of elaboration checks, and if the default is not satisfactory for your use,
19740 you should read this chapter.
19742 For validity checks, the minimal checks required by the Ada Reference
19743 Manual (for case statements and assignments to array elements) are on
19744 by default. These can be suppressed by use of the @code{-gnatVn} switch.
19745 Note that in Ada 83, there were no validity checks, so if the Ada 83 mode
19746 is acceptable (or when comparing GNAT performance with an Ada 83 compiler),
19747 it may be reasonable to routinely use @code{-gnatVn}. Validity checks
19748 are also suppressed entirely if @code{-gnatp} is used.
19750 @geindex Overflow checks
19757 @geindex Unsuppress
19759 @geindex pragma Suppress
19761 @geindex pragma Unsuppress
19763 Note that the setting of the switches controls the default setting of
19764 the checks. They may be modified using either @code{pragma Suppress} (to
19765 remove checks) or @code{pragma Unsuppress} (to add back suppressed
19766 checks) in the program source.
19768 @node Use of Restrictions,Optimization Levels,Controlling Run-Time Checks,Performance Considerations
19769 @anchor{gnat_ugn/gnat_and_program_execution id29}@anchor{180}@anchor{gnat_ugn/gnat_and_program_execution use-of-restrictions}@anchor{181}
19770 @subsubsection Use of Restrictions
19773 The use of pragma Restrictions allows you to control which features are
19774 permitted in your program. Apart from the obvious point that if you avoid
19775 relatively expensive features like finalization (enforceable by the use
19776 of pragma Restrictions (No_Finalization), the use of this pragma does not
19777 affect the generated code in most cases.
19779 One notable exception to this rule is that the possibility of task abort
19780 results in some distributed overhead, particularly if finalization or
19781 exception handlers are used. The reason is that certain sections of code
19782 have to be marked as non-abortable.
19784 If you use neither the @code{abort} statement, nor asynchronous transfer
19785 of control (@code{select ... then abort}), then this distributed overhead
19786 is removed, which may have a general positive effect in improving
19787 overall performance. Especially code involving frequent use of tasking
19788 constructs and controlled types will show much improved performance.
19789 The relevant restrictions pragmas are
19794 pragma Restrictions (No_Abort_Statements);
19795 pragma Restrictions (Max_Asynchronous_Select_Nesting => 0);
19799 It is recommended that these restriction pragmas be used if possible. Note
19800 that this also means that you can write code without worrying about the
19801 possibility of an immediate abort at any point.
19803 @node Optimization Levels,Debugging Optimized Code,Use of Restrictions,Performance Considerations
19804 @anchor{gnat_ugn/gnat_and_program_execution id30}@anchor{182}@anchor{gnat_ugn/gnat_and_program_execution optimization-levels}@anchor{ed}
19805 @subsubsection Optimization Levels
19810 Without any optimization option,
19811 the compiler’s goal is to reduce the cost of
19812 compilation and to make debugging produce the expected results.
19813 Statements are independent: if you stop the program with a breakpoint between
19814 statements, you can then assign a new value to any variable or change
19815 the program counter to any other statement in the subprogram and get exactly
19816 the results you would expect from the source code.
19818 Turning on optimization makes the compiler attempt to improve the
19819 performance and/or code size at the expense of compilation time and
19820 possibly the ability to debug the program.
19822 If you use multiple
19823 -O options, with or without level numbers,
19824 the last such option is the one that is effective.
19826 The default is optimization off. This results in the fastest compile
19827 times, but GNAT makes absolutely no attempt to optimize, and the
19828 generated programs are considerably larger and slower than when
19829 optimization is enabled. You can use the
19830 @code{-O} switch (the permitted forms are @code{-O0}, @code{-O1}
19831 @code{-O2}, @code{-O3}, and @code{-Os})
19832 to @code{gcc} to control the optimization level:
19843 No optimization (the default);
19844 generates unoptimized code but has
19845 the fastest compilation time.
19847 Note that many other compilers do substantial optimization even
19848 if ‘no optimization’ is specified. With gcc, it is very unusual
19849 to use @code{-O0} for production if execution time is of any concern,
19850 since @code{-O0} means (almost) no optimization. This difference
19851 between gcc and other compilers should be kept in mind when
19852 doing performance comparisons.
19861 Moderate optimization;
19862 optimizes reasonably well but does not
19863 degrade compilation time significantly.
19873 generates highly optimized code and has
19874 the slowest compilation time.
19883 Full optimization as in @code{-O2};
19884 also uses more aggressive automatic inlining of subprograms within a unit
19885 (@ref{100,,Inlining of Subprograms}) and attempts to vectorize loops.
19894 Optimize space usage (code and data) of resulting program.
19898 Higher optimization levels perform more global transformations on the
19899 program and apply more expensive analysis algorithms in order to generate
19900 faster and more compact code. The price in compilation time, and the
19901 resulting improvement in execution time,
19902 both depend on the particular application and the hardware environment.
19903 You should experiment to find the best level for your application.
19905 Since the precise set of optimizations done at each level will vary from
19906 release to release (and sometime from target to target), it is best to think
19907 of the optimization settings in general terms.
19908 See the @emph{Options That Control Optimization} section in
19909 @cite{Using the GNU Compiler Collection (GCC)}
19911 the @code{-O} settings and a number of @code{-f} options that
19912 individually enable or disable specific optimizations.
19914 Unlike some other compilation systems, @code{gcc} has
19915 been tested extensively at all optimization levels. There are some bugs
19916 which appear only with optimization turned on, but there have also been
19917 bugs which show up only in @emph{unoptimized} code. Selecting a lower
19918 level of optimization does not improve the reliability of the code
19919 generator, which in practice is highly reliable at all optimization
19922 Note regarding the use of @code{-O3}: The use of this optimization level
19923 ought not to be automatically preferred over that of level @code{-O2},
19924 since it often results in larger executables which may run more slowly.
19925 See further discussion of this point in @ref{100,,Inlining of Subprograms}.
19927 @node Debugging Optimized Code,Inlining of Subprograms,Optimization Levels,Performance Considerations
19928 @anchor{gnat_ugn/gnat_and_program_execution debugging-optimized-code}@anchor{183}@anchor{gnat_ugn/gnat_and_program_execution id31}@anchor{184}
19929 @subsubsection Debugging Optimized Code
19932 @geindex Debugging optimized code
19934 @geindex Optimization and debugging
19936 Although it is possible to do a reasonable amount of debugging at
19937 nonzero optimization levels,
19938 the higher the level the more likely that
19939 source-level constructs will have been eliminated by optimization.
19940 For example, if a loop is strength-reduced, the loop
19941 control variable may be completely eliminated and thus cannot be
19942 displayed in the debugger.
19943 This can only happen at @code{-O2} or @code{-O3}.
19944 Explicit temporary variables that you code might be eliminated at
19945 level @code{-O1} or higher.
19949 The use of the @code{-g} switch,
19950 which is needed for source-level debugging,
19951 affects the size of the program executable on disk,
19952 and indeed the debugging information can be quite large.
19953 However, it has no effect on the generated code (and thus does not
19954 degrade performance)
19956 Since the compiler generates debugging tables for a compilation unit before
19957 it performs optimizations, the optimizing transformations may invalidate some
19958 of the debugging data. You therefore need to anticipate certain
19959 anomalous situations that may arise while debugging optimized code.
19960 These are the most common cases:
19966 @emph{The ‘hopping Program Counter’:} Repeated @code{step} or @code{next}
19968 the PC bouncing back and forth in the code. This may result from any of
19969 the following optimizations:
19975 @emph{Common subexpression elimination:} using a single instance of code for a
19976 quantity that the source computes several times. As a result you
19977 may not be able to stop on what looks like a statement.
19980 @emph{Invariant code motion:} moving an expression that does not change within a
19981 loop, to the beginning of the loop.
19984 @emph{Instruction scheduling:} moving instructions so as to
19985 overlap loads and stores (typically) with other code, or in
19986 general to move computations of values closer to their uses. Often
19987 this causes you to pass an assignment statement without the assignment
19988 happening and then later bounce back to the statement when the
19989 value is actually needed. Placing a breakpoint on a line of code
19990 and then stepping over it may, therefore, not always cause all the
19991 expected side-effects.
19995 @emph{The ‘big leap’:} More commonly known as @emph{cross-jumping}, in which
19996 two identical pieces of code are merged and the program counter suddenly
19997 jumps to a statement that is not supposed to be executed, simply because
19998 it (and the code following) translates to the same thing as the code
19999 that @emph{was} supposed to be executed. This effect is typically seen in
20000 sequences that end in a jump, such as a @code{goto}, a @code{return}, or
20001 a @code{break} in a C @code{switch} statement.
20004 @emph{The ‘roving variable’:} The symptom is an unexpected value in a variable.
20005 There are various reasons for this effect:
20011 In a subprogram prologue, a parameter may not yet have been moved to its
20015 A variable may be dead, and its register re-used. This is
20016 probably the most common cause.
20019 As mentioned above, the assignment of a value to a variable may
20023 A variable may be eliminated entirely by value propagation or
20024 other means. In this case, GCC may incorrectly generate debugging
20025 information for the variable
20028 In general, when an unexpected value appears for a local variable or parameter
20029 you should first ascertain if that value was actually computed by
20030 your program, as opposed to being incorrectly reported by the debugger.
20032 array elements in an object designated by an access value
20033 are generally less of a problem, once you have ascertained that the access
20035 Typically, this means checking variables in the preceding code and in the
20036 calling subprogram to verify that the value observed is explainable from other
20037 values (one must apply the procedure recursively to those
20038 other values); or re-running the code and stopping a little earlier
20039 (perhaps before the call) and stepping to better see how the variable obtained
20040 the value in question; or continuing to step @emph{from} the point of the
20041 strange value to see if code motion had simply moved the variable’s
20045 In light of such anomalies, a recommended technique is to use @code{-O0}
20046 early in the software development cycle, when extensive debugging capabilities
20047 are most needed, and then move to @code{-O1} and later @code{-O2} as
20048 the debugger becomes less critical.
20049 Whether to use the @code{-g} switch in the release version is
20050 a release management issue.
20051 Note that if you use @code{-g} you can then use the @code{strip} program
20052 on the resulting executable,
20053 which removes both debugging information and global symbols.
20055 @node Inlining of Subprograms,Floating Point Operations,Debugging Optimized Code,Performance Considerations
20056 @anchor{gnat_ugn/gnat_and_program_execution id32}@anchor{185}@anchor{gnat_ugn/gnat_and_program_execution inlining-of-subprograms}@anchor{100}
20057 @subsubsection Inlining of Subprograms
20060 A call to a subprogram in the current unit is inlined if all the
20061 following conditions are met:
20067 The optimization level is at least @code{-O1}.
20070 The called subprogram is suitable for inlining: It must be small enough
20071 and not contain something that @code{gcc} cannot support in inlined
20074 @geindex pragma Inline
20079 Any one of the following applies: @code{pragma Inline} is applied to the
20080 subprogram; the subprogram is local to the unit and called once from
20081 within it; the subprogram is small and optimization level @code{-O2} is
20082 specified; optimization level @code{-O3} is specified.
20085 Calls to subprograms in @emph{with}ed units are normally not inlined.
20086 To achieve actual inlining (that is, replacement of the call by the code
20087 in the body of the subprogram), the following conditions must all be true:
20093 The optimization level is at least @code{-O1}.
20096 The called subprogram is suitable for inlining: It must be small enough
20097 and not contain something that @code{gcc} cannot support in inlined
20101 There is a @code{pragma Inline} for the subprogram.
20104 The @code{-gnatn} switch is used on the command line.
20107 Even if all these conditions are met, it may not be possible for
20108 the compiler to inline the call, due to the length of the body,
20109 or features in the body that make it impossible for the compiler
20110 to do the inlining.
20112 Note that specifying the @code{-gnatn} switch causes additional
20113 compilation dependencies. Consider the following:
20135 With the default behavior (no @code{-gnatn} switch specified), the
20136 compilation of the @code{Main} procedure depends only on its own source,
20137 @code{main.adb}, and the spec of the package in file @code{r.ads}. This
20138 means that editing the body of @code{R} does not require recompiling
20141 On the other hand, the call @code{R.Q} is not inlined under these
20142 circumstances. If the @code{-gnatn} switch is present when @code{Main}
20143 is compiled, the call will be inlined if the body of @code{Q} is small
20144 enough, but now @code{Main} depends on the body of @code{R} in
20145 @code{r.adb} as well as on the spec. This means that if this body is edited,
20146 the main program must be recompiled. Note that this extra dependency
20147 occurs whether or not the call is in fact inlined by @code{gcc}.
20149 The use of front end inlining with @code{-gnatN} generates similar
20150 additional dependencies.
20152 @geindex -fno-inline (gcc)
20154 Note: The @code{-fno-inline} switch overrides all other conditions and ensures that
20155 no inlining occurs, unless requested with pragma Inline_Always for @code{gcc}
20156 back-ends. The extra dependences resulting from @code{-gnatn} will still be active,
20157 even if this switch is used to suppress the resulting inlining actions.
20159 @geindex -fno-inline-functions (gcc)
20161 Note: The @code{-fno-inline-functions} switch can be used to prevent
20162 automatic inlining of subprograms if @code{-O3} is used.
20164 @geindex -fno-inline-small-functions (gcc)
20166 Note: The @code{-fno-inline-small-functions} switch can be used to prevent
20167 automatic inlining of small subprograms if @code{-O2} is used.
20169 @geindex -fno-inline-functions-called-once (gcc)
20171 Note: The @code{-fno-inline-functions-called-once} switch
20172 can be used to prevent inlining of subprograms local to the unit
20173 and called once from within it if @code{-O1} is used.
20175 Note regarding the use of @code{-O3}: @code{-gnatn} is made up of two
20176 sub-switches @code{-gnatn1} and @code{-gnatn2} that can be directly
20177 specified in lieu of it, @code{-gnatn} being translated into one of them
20178 based on the optimization level. With @code{-O2} or below, @code{-gnatn}
20179 is equivalent to @code{-gnatn1} which activates pragma @code{Inline} with
20180 moderate inlining across modules. With @code{-O3}, @code{-gnatn} is
20181 equivalent to @code{-gnatn2} which activates pragma @code{Inline} with
20182 full inlining across modules. If you have used pragma @code{Inline} in
20183 appropriate cases, then it is usually much better to use @code{-O2}
20184 and @code{-gnatn} and avoid the use of @code{-O3} which has the additional
20185 effect of inlining subprograms you did not think should be inlined. We have
20186 found that the use of @code{-O3} may slow down the compilation and increase
20187 the code size by performing excessive inlining, leading to increased
20188 instruction cache pressure from the increased code size and thus minor
20189 performance improvements. So the bottom line here is that you should not
20190 automatically assume that @code{-O3} is better than @code{-O2}, and
20191 indeed you should use @code{-O3} only if tests show that it actually
20192 improves performance for your program.
20194 @node Floating Point Operations,Vectorization of loops,Inlining of Subprograms,Performance Considerations
20195 @anchor{gnat_ugn/gnat_and_program_execution floating-point-operations}@anchor{186}@anchor{gnat_ugn/gnat_and_program_execution id33}@anchor{187}
20196 @subsubsection Floating Point Operations
20199 @geindex Floating-Point Operations
20201 On almost all targets, GNAT maps Float and Long_Float to the 32-bit and
20202 64-bit standard IEEE floating-point representations, and operations will
20203 use standard IEEE arithmetic as provided by the processor. On most, but
20204 not all, architectures, the attribute Machine_Overflows is False for these
20205 types, meaning that the semantics of overflow is implementation-defined.
20206 In the case of GNAT, these semantics correspond to the normal IEEE
20207 treatment of infinities and NaN (not a number) values. For example,
20208 1.0 / 0.0 yields plus infinitiy and 0.0 / 0.0 yields a NaN. By
20209 avoiding explicit overflow checks, the performance is greatly improved
20210 on many targets. However, if required, floating-point overflow can be
20211 enabled by the use of the pragma Check_Float_Overflow.
20213 Another consideration that applies specifically to x86 32-bit
20214 architectures is which form of floating-point arithmetic is used.
20215 By default the operations use the old style x86 floating-point,
20216 which implements an 80-bit extended precision form (on these
20217 architectures the type Long_Long_Float corresponds to that form).
20218 In addition, generation of efficient code in this mode means that
20219 the extended precision form will be used for intermediate results.
20220 This may be helpful in improving the final precision of a complex
20221 expression. However it means that the results obtained on the x86
20222 will be different from those on other architectures, and for some
20223 algorithms, the extra intermediate precision can be detrimental.
20225 In addition to this old-style floating-point, all modern x86 chips
20226 implement an alternative floating-point operation model referred
20227 to as SSE2. In this model there is no extended form, and furthermore
20228 execution performance is significantly enhanced. To force GNAT to use
20229 this more modern form, use both of the switches:
20233 -msse2 -mfpmath=sse
20236 A unit compiled with these switches will automatically use the more
20237 efficient SSE2 instruction set for Float and Long_Float operations.
20238 Note that the ABI has the same form for both floating-point models,
20239 so it is permissible to mix units compiled with and without these
20242 @node Vectorization of loops,Other Optimization Switches,Floating Point Operations,Performance Considerations
20243 @anchor{gnat_ugn/gnat_and_program_execution id34}@anchor{188}@anchor{gnat_ugn/gnat_and_program_execution vectorization-of-loops}@anchor{189}
20244 @subsubsection Vectorization of loops
20247 @geindex Optimization Switches
20249 You can take advantage of the auto-vectorizer present in the @code{gcc}
20250 back end to vectorize loops with GNAT. The corresponding command line switch
20251 is @code{-ftree-vectorize} but, as it is enabled by default at @code{-O3}
20252 and other aggressive optimizations helpful for vectorization also are enabled
20253 by default at this level, using @code{-O3} directly is recommended.
20255 You also need to make sure that the target architecture features a supported
20256 SIMD instruction set. For example, for the x86 architecture, you should at
20257 least specify @code{-msse2} to get significant vectorization (but you don’t
20258 need to specify it for x86-64 as it is part of the base 64-bit architecture).
20259 Similarly, for the PowerPC architecture, you should specify @code{-maltivec}.
20261 The preferred loop form for vectorization is the @code{for} iteration scheme.
20262 Loops with a @code{while} iteration scheme can also be vectorized if they are
20263 very simple, but the vectorizer will quickly give up otherwise. With either
20264 iteration scheme, the flow of control must be straight, in particular no
20265 @code{exit} statement may appear in the loop body. The loop may however
20266 contain a single nested loop, if it can be vectorized when considered alone:
20271 A : array (1..4, 1..4) of Long_Float;
20272 S : array (1..4) of Long_Float;
20276 for I in A'Range(1) loop
20277 for J in A'Range(2) loop
20278 S (I) := S (I) + A (I, J);
20285 The vectorizable operations depend on the targeted SIMD instruction set, but
20286 the adding and some of the multiplying operators are generally supported, as
20287 well as the logical operators for modular types. Note that compiling
20288 with @code{-gnatp} might well reveal cases where some checks do thwart
20291 Type conversions may also prevent vectorization if they involve semantics that
20292 are not directly supported by the code generator or the SIMD instruction set.
20293 A typical example is direct conversion from floating-point to integer types.
20294 The solution in this case is to use the following idiom:
20299 Integer (S'Truncation (F))
20303 if @code{S} is the subtype of floating-point object @code{F}.
20305 In most cases, the vectorizable loops are loops that iterate over arrays.
20306 All kinds of array types are supported, i.e. constrained array types with
20312 type Array_Type is array (1 .. 4) of Long_Float;
20316 constrained array types with dynamic bounds:
20321 type Array_Type is array (1 .. Q.N) of Long_Float;
20323 type Array_Type is array (Q.K .. 4) of Long_Float;
20325 type Array_Type is array (Q.K .. Q.N) of Long_Float;
20329 or unconstrained array types:
20334 type Array_Type is array (Positive range <>) of Long_Float;
20338 The quality of the generated code decreases when the dynamic aspect of the
20339 array type increases, the worst code being generated for unconstrained array
20340 types. This is so because, the less information the compiler has about the
20341 bounds of the array, the more fallback code it needs to generate in order to
20342 fix things up at run time.
20344 It is possible to specify that a given loop should be subject to vectorization
20345 preferably to other optimizations by means of pragma @code{Loop_Optimize}:
20350 pragma Loop_Optimize (Vector);
20354 placed immediately within the loop will convey the appropriate hint to the
20355 compiler for this loop.
20357 It is also possible to help the compiler generate better vectorized code
20358 for a given loop by asserting that there are no loop-carried dependencies
20359 in the loop. Consider for example the procedure:
20364 type Arr is array (1 .. 4) of Long_Float;
20366 procedure Add (X, Y : not null access Arr; R : not null access Arr) is
20368 for I in Arr'Range loop
20369 R(I) := X(I) + Y(I);
20375 By default, the compiler cannot unconditionally vectorize the loop because
20376 assigning to a component of the array designated by R in one iteration could
20377 change the value read from the components of the array designated by X or Y
20378 in a later iteration. As a result, the compiler will generate two versions
20379 of the loop in the object code, one vectorized and the other not vectorized,
20380 as well as a test to select the appropriate version at run time. This can
20381 be overcome by another hint:
20386 pragma Loop_Optimize (Ivdep);
20390 placed immediately within the loop will tell the compiler that it can safely
20391 omit the non-vectorized version of the loop as well as the run-time test.
20393 @node Other Optimization Switches,Optimization and Strict Aliasing,Vectorization of loops,Performance Considerations
20394 @anchor{gnat_ugn/gnat_and_program_execution id35}@anchor{18a}@anchor{gnat_ugn/gnat_and_program_execution other-optimization-switches}@anchor{18b}
20395 @subsubsection Other Optimization Switches
20398 @geindex Optimization Switches
20400 Since GNAT uses the @code{gcc} back end, all the specialized
20401 @code{gcc} optimization switches are potentially usable. These switches
20402 have not been extensively tested with GNAT but can generally be expected
20403 to work. Examples of switches in this category are @code{-funroll-loops}
20404 and the various target-specific @code{-m} options (in particular, it has
20405 been observed that @code{-march=xxx} can significantly improve performance
20406 on appropriate machines). For full details of these switches, see
20407 the @emph{Submodel Options} section in the @emph{Hardware Models and Configurations}
20408 chapter of @cite{Using the GNU Compiler Collection (GCC)}.
20410 @node Optimization and Strict Aliasing,Aliased Variables and Optimization,Other Optimization Switches,Performance Considerations
20411 @anchor{gnat_ugn/gnat_and_program_execution id36}@anchor{18c}@anchor{gnat_ugn/gnat_and_program_execution optimization-and-strict-aliasing}@anchor{e4}
20412 @subsubsection Optimization and Strict Aliasing
20417 @geindex Strict Aliasing
20419 @geindex No_Strict_Aliasing
20421 The strong typing capabilities of Ada allow an optimizer to generate
20422 efficient code in situations where other languages would be forced to
20423 make worst case assumptions preventing such optimizations. Consider
20424 the following example:
20430 type Int1 is new Integer;
20431 type Int2 is new Integer;
20432 type Int1A is access Int1;
20433 type Int2A is access Int2;
20440 for J in Data'Range loop
20441 if Data (J) = Int1V.all then
20442 Int2V.all := Int2V.all + 1;
20450 In this example, since the variable @code{Int1V} can only access objects
20451 of type @code{Int1}, and @code{Int2V} can only access objects of type
20452 @code{Int2}, there is no possibility that the assignment to
20453 @code{Int2V.all} affects the value of @code{Int1V.all}. This means that
20454 the compiler optimizer can “know” that the value @code{Int1V.all} is constant
20455 for all iterations of the loop and avoid the extra memory reference
20456 required to dereference it each time through the loop.
20458 This kind of optimization, called strict aliasing analysis, is
20459 triggered by specifying an optimization level of @code{-O2} or
20460 higher or @code{-Os} and allows GNAT to generate more efficient code
20461 when access values are involved.
20463 However, although this optimization is always correct in terms of
20464 the formal semantics of the Ada Reference Manual, difficulties can
20465 arise if features like @code{Unchecked_Conversion} are used to break
20466 the typing system. Consider the following complete program example:
20472 type int1 is new integer;
20473 type int2 is new integer;
20474 type a1 is access int1;
20475 type a2 is access int2;
20480 function to_a2 (Input : a1) return a2;
20483 with Unchecked_Conversion;
20485 function to_a2 (Input : a1) return a2 is
20487 new Unchecked_Conversion (a1, a2);
20489 return to_a2u (Input);
20495 with Text_IO; use Text_IO;
20497 v1 : a1 := new int1;
20498 v2 : a2 := to_a2 (v1);
20502 put_line (int1'image (v1.all));
20507 This program prints out 0 in @code{-O0} or @code{-O1}
20508 mode, but it prints out 1 in @code{-O2} mode. That’s
20509 because in strict aliasing mode, the compiler can and
20510 does assume that the assignment to @code{v2.all} could not
20511 affect the value of @code{v1.all}, since different types
20514 This behavior is not a case of non-conformance with the standard, since
20515 the Ada RM specifies that an unchecked conversion where the resulting
20516 bit pattern is not a correct value of the target type can result in an
20517 abnormal value and attempting to reference an abnormal value makes the
20518 execution of a program erroneous. That’s the case here since the result
20519 does not point to an object of type @code{int2}. This means that the
20520 effect is entirely unpredictable.
20522 However, although that explanation may satisfy a language
20523 lawyer, in practice an applications programmer expects an
20524 unchecked conversion involving pointers to create true
20525 aliases and the behavior of printing 1 seems plain wrong.
20526 In this case, the strict aliasing optimization is unwelcome.
20528 Indeed the compiler recognizes this possibility, and the
20529 unchecked conversion generates a warning:
20534 p2.adb:5:07: warning: possible aliasing problem with type "a2"
20535 p2.adb:5:07: warning: use -fno-strict-aliasing switch for references
20536 p2.adb:5:07: warning: or use "pragma No_Strict_Aliasing (a2);"
20540 Unfortunately the problem is recognized when compiling the body of
20541 package @code{p2}, but the actual “bad” code is generated while
20542 compiling the body of @code{m} and this latter compilation does not see
20543 the suspicious @code{Unchecked_Conversion}.
20545 As implied by the warning message, there are approaches you can use to
20546 avoid the unwanted strict aliasing optimization in a case like this.
20548 One possibility is to simply avoid the use of @code{-O2}, but
20549 that is a bit drastic, since it throws away a number of useful
20550 optimizations that do not involve strict aliasing assumptions.
20552 A less drastic approach is to compile the program using the
20553 option @code{-fno-strict-aliasing}. Actually it is only the
20554 unit containing the dereferencing of the suspicious pointer
20555 that needs to be compiled. So in this case, if we compile
20556 unit @code{m} with this switch, then we get the expected
20557 value of zero printed. Analyzing which units might need
20558 the switch can be painful, so a more reasonable approach
20559 is to compile the entire program with options @code{-O2}
20560 and @code{-fno-strict-aliasing}. If the performance is
20561 satisfactory with this combination of options, then the
20562 advantage is that the entire issue of possible “wrong”
20563 optimization due to strict aliasing is avoided.
20565 To avoid the use of compiler switches, the configuration
20566 pragma @code{No_Strict_Aliasing} with no parameters may be
20567 used to specify that for all access types, the strict
20568 aliasing optimization should be suppressed.
20570 However, these approaches are still overkill, in that they causes
20571 all manipulations of all access values to be deoptimized. A more
20572 refined approach is to concentrate attention on the specific
20573 access type identified as problematic.
20575 First, if a careful analysis of uses of the pointer shows
20576 that there are no possible problematic references, then
20577 the warning can be suppressed by bracketing the
20578 instantiation of @code{Unchecked_Conversion} to turn
20584 pragma Warnings (Off);
20586 new Unchecked_Conversion (a1, a2);
20587 pragma Warnings (On);
20591 Of course that approach is not appropriate for this particular
20592 example, since indeed there is a problematic reference. In this
20593 case we can take one of two other approaches.
20595 The first possibility is to move the instantiation of unchecked
20596 conversion to the unit in which the type is declared. In
20597 this example, we would move the instantiation of
20598 @code{Unchecked_Conversion} from the body of package
20599 @code{p2} to the spec of package @code{p1}. Now the
20600 warning disappears. That’s because any use of the
20601 access type knows there is a suspicious unchecked
20602 conversion, and the strict aliasing optimization
20603 is automatically suppressed for the type.
20605 If it is not practical to move the unchecked conversion to the same unit
20606 in which the destination access type is declared (perhaps because the
20607 source type is not visible in that unit), you may use pragma
20608 @code{No_Strict_Aliasing} for the type. This pragma must occur in the
20609 same declarative sequence as the declaration of the access type:
20614 type a2 is access int2;
20615 pragma No_Strict_Aliasing (a2);
20619 Here again, the compiler now knows that the strict aliasing optimization
20620 should be suppressed for any reference to type @code{a2} and the
20621 expected behavior is obtained.
20623 Finally, note that although the compiler can generate warnings for
20624 simple cases of unchecked conversions, there are tricker and more
20625 indirect ways of creating type incorrect aliases which the compiler
20626 cannot detect. Examples are the use of address overlays and unchecked
20627 conversions involving composite types containing access types as
20628 components. In such cases, no warnings are generated, but there can
20629 still be aliasing problems. One safe coding practice is to forbid the
20630 use of address clauses for type overlaying, and to allow unchecked
20631 conversion only for primitive types. This is not really a significant
20632 restriction since any possible desired effect can be achieved by
20633 unchecked conversion of access values.
20635 The aliasing analysis done in strict aliasing mode can certainly
20636 have significant benefits. We have seen cases of large scale
20637 application code where the time is increased by up to 5% by turning
20638 this optimization off. If you have code that includes significant
20639 usage of unchecked conversion, you might want to just stick with
20640 @code{-O1} and avoid the entire issue. If you get adequate
20641 performance at this level of optimization level, that’s probably
20642 the safest approach. If tests show that you really need higher
20643 levels of optimization, then you can experiment with @code{-O2}
20644 and @code{-O2 -fno-strict-aliasing} to see how much effect this
20645 has on size and speed of the code. If you really need to use
20646 @code{-O2} with strict aliasing in effect, then you should
20647 review any uses of unchecked conversion of access types,
20648 particularly if you are getting the warnings described above.
20650 @node Aliased Variables and Optimization,Atomic Variables and Optimization,Optimization and Strict Aliasing,Performance Considerations
20651 @anchor{gnat_ugn/gnat_and_program_execution aliased-variables-and-optimization}@anchor{18d}@anchor{gnat_ugn/gnat_and_program_execution id37}@anchor{18e}
20652 @subsubsection Aliased Variables and Optimization
20657 There are scenarios in which programs may
20658 use low level techniques to modify variables
20659 that otherwise might be considered to be unassigned. For example,
20660 a variable can be passed to a procedure by reference, which takes
20661 the address of the parameter and uses the address to modify the
20662 variable’s value, even though it is passed as an IN parameter.
20663 Consider the following example:
20669 Max_Length : constant Natural := 16;
20670 type Char_Ptr is access all Character;
20672 procedure Get_String(Buffer: Char_Ptr; Size : Integer);
20673 pragma Import (C, Get_String, "get_string");
20675 Name : aliased String (1 .. Max_Length) := (others => ' ');
20678 function Addr (S : String) return Char_Ptr is
20679 function To_Char_Ptr is
20680 new Ada.Unchecked_Conversion (System.Address, Char_Ptr);
20682 return To_Char_Ptr (S (S'First)'Address);
20686 Temp := Addr (Name);
20687 Get_String (Temp, Max_Length);
20692 where Get_String is a C function that uses the address in Temp to
20693 modify the variable @code{Name}. This code is dubious, and arguably
20694 erroneous, and the compiler would be entitled to assume that
20695 @code{Name} is never modified, and generate code accordingly.
20697 However, in practice, this would cause some existing code that
20698 seems to work with no optimization to start failing at high
20699 levels of optimzization.
20701 What the compiler does for such cases is to assume that marking
20702 a variable as aliased indicates that some “funny business” may
20703 be going on. The optimizer recognizes the aliased keyword and
20704 inhibits optimizations that assume the value cannot be assigned.
20705 This means that the above example will in fact “work” reliably,
20706 that is, it will produce the expected results.
20708 @node Atomic Variables and Optimization,Passive Task Optimization,Aliased Variables and Optimization,Performance Considerations
20709 @anchor{gnat_ugn/gnat_and_program_execution atomic-variables-and-optimization}@anchor{18f}@anchor{gnat_ugn/gnat_and_program_execution id38}@anchor{190}
20710 @subsubsection Atomic Variables and Optimization
20715 There are two considerations with regard to performance when
20716 atomic variables are used.
20718 First, the RM only guarantees that access to atomic variables
20719 be atomic, it has nothing to say about how this is achieved,
20720 though there is a strong implication that this should not be
20721 achieved by explicit locking code. Indeed GNAT will never
20722 generate any locking code for atomic variable access (it will
20723 simply reject any attempt to make a variable or type atomic
20724 if the atomic access cannot be achieved without such locking code).
20726 That being said, it is important to understand that you cannot
20727 assume that the entire variable will always be accessed. Consider
20734 A,B,C,D : Character;
20737 for R'Alignment use 4;
20740 pragma Atomic (RV);
20747 You cannot assume that the reference to @code{RV.B}
20748 will read the entire 32-bit
20749 variable with a single load instruction. It is perfectly legitimate if
20750 the hardware allows it to do a byte read of just the B field. This read
20751 is still atomic, which is all the RM requires. GNAT can and does take
20752 advantage of this, depending on the architecture and optimization level.
20753 Any assumption to the contrary is non-portable and risky. Even if you
20754 examine the assembly language and see a full 32-bit load, this might
20755 change in a future version of the compiler.
20757 If your application requires that all accesses to @code{RV} in this
20758 example be full 32-bit loads, you need to make a copy for the access
20765 RV_Copy : constant R := RV;
20772 Now the reference to RV must read the whole variable.
20773 Actually one can imagine some compiler which figures
20774 out that the whole copy is not required (because only
20775 the B field is actually accessed), but GNAT
20776 certainly won’t do that, and we don’t know of any
20777 compiler that would not handle this right, and the
20778 above code will in practice work portably across
20779 all architectures (that permit the Atomic declaration).
20781 The second issue with atomic variables has to do with
20782 the possible requirement of generating synchronization
20783 code. For more details on this, consult the sections on
20784 the pragmas Enable/Disable_Atomic_Synchronization in the
20785 GNAT Reference Manual. If performance is critical, and
20786 such synchronization code is not required, it may be
20787 useful to disable it.
20789 @node Passive Task Optimization,,Atomic Variables and Optimization,Performance Considerations
20790 @anchor{gnat_ugn/gnat_and_program_execution id39}@anchor{191}@anchor{gnat_ugn/gnat_and_program_execution passive-task-optimization}@anchor{192}
20791 @subsubsection Passive Task Optimization
20794 @geindex Passive Task
20796 A passive task is one which is sufficiently simple that
20797 in theory a compiler could recognize it an implement it
20798 efficiently without creating a new thread. The original design
20799 of Ada 83 had in mind this kind of passive task optimization, but
20800 only a few Ada 83 compilers attempted it. The problem was that
20801 it was difficult to determine the exact conditions under which
20802 the optimization was possible. The result is a very fragile
20803 optimization where a very minor change in the program can
20804 suddenly silently make a task non-optimizable.
20806 With the revisiting of this issue in Ada 95, there was general
20807 agreement that this approach was fundamentally flawed, and the
20808 notion of protected types was introduced. When using protected
20809 types, the restrictions are well defined, and you KNOW that the
20810 operations will be optimized, and furthermore this optimized
20811 performance is fully portable.
20813 Although it would theoretically be possible for GNAT to attempt to
20814 do this optimization, but it really doesn’t make sense in the
20815 context of Ada 95, and none of the Ada 95 compilers implement
20816 this optimization as far as we know. In particular GNAT never
20817 attempts to perform this optimization.
20819 In any new Ada 95 code that is written, you should always
20820 use protected types in place of tasks that might be able to
20821 be optimized in this manner.
20822 Of course this does not help if you have legacy Ada 83 code
20823 that depends on this optimization, but it is unusual to encounter
20824 a case where the performance gains from this optimization
20827 Your program should work correctly without this optimization. If
20828 you have performance problems, then the most practical
20829 approach is to figure out exactly where these performance problems
20830 arise, and update those particular tasks to be protected types. Note
20831 that typically clients of the tasks who call entries, will not have
20832 to be modified, only the task definition itself.
20834 @node Text_IO Suggestions,Reducing Size of Executables with Unused Subprogram/Data Elimination,Performance Considerations,Improving Performance
20835 @anchor{gnat_ugn/gnat_and_program_execution id40}@anchor{193}@anchor{gnat_ugn/gnat_and_program_execution text-io-suggestions}@anchor{194}
20836 @subsection @code{Text_IO} Suggestions
20839 @geindex Text_IO and performance
20841 The @code{Ada.Text_IO} package has fairly high overheads due in part to
20842 the requirement of maintaining page and line counts. If performance
20843 is critical, a recommendation is to use @code{Stream_IO} instead of
20844 @code{Text_IO} for volume output, since this package has less overhead.
20846 If @code{Text_IO} must be used, note that by default output to the standard
20847 output and standard error files is unbuffered (this provides better
20848 behavior when output statements are used for debugging, or if the
20849 progress of a program is observed by tracking the output, e.g. by
20850 using the Unix @emph{tail -f} command to watch redirected output.
20852 If you are generating large volumes of output with @code{Text_IO} and
20853 performance is an important factor, use a designated file instead
20854 of the standard output file, or change the standard output file to
20855 be buffered using @code{Interfaces.C_Streams.setvbuf}.
20857 @node Reducing Size of Executables with Unused Subprogram/Data Elimination,,Text_IO Suggestions,Improving Performance
20858 @anchor{gnat_ugn/gnat_and_program_execution id41}@anchor{195}@anchor{gnat_ugn/gnat_and_program_execution reducing-size-of-executables-with-unused-subprogram-data-elimination}@anchor{196}
20859 @subsection Reducing Size of Executables with Unused Subprogram/Data Elimination
20862 @geindex Uunused subprogram/data elimination
20864 This section describes how you can eliminate unused subprograms and data from
20865 your executable just by setting options at compilation time.
20868 * About unused subprogram/data elimination::
20869 * Compilation options::
20870 * Example of unused subprogram/data elimination::
20874 @node About unused subprogram/data elimination,Compilation options,,Reducing Size of Executables with Unused Subprogram/Data Elimination
20875 @anchor{gnat_ugn/gnat_and_program_execution about-unused-subprogram-data-elimination}@anchor{197}@anchor{gnat_ugn/gnat_and_program_execution id42}@anchor{198}
20876 @subsubsection About unused subprogram/data elimination
20879 By default, an executable contains all code and data of its composing objects
20880 (directly linked or coming from statically linked libraries), even data or code
20881 never used by this executable.
20883 This feature will allow you to eliminate such unused code from your
20884 executable, making it smaller (in disk and in memory).
20886 This functionality is available on all Linux platforms except for the IA-64
20887 architecture and on all cross platforms using the ELF binary file format.
20888 In both cases GNU binutils version 2.16 or later are required to enable it.
20890 @node Compilation options,Example of unused subprogram/data elimination,About unused subprogram/data elimination,Reducing Size of Executables with Unused Subprogram/Data Elimination
20891 @anchor{gnat_ugn/gnat_and_program_execution compilation-options}@anchor{199}@anchor{gnat_ugn/gnat_and_program_execution id43}@anchor{19a}
20892 @subsubsection Compilation options
20895 The operation of eliminating the unused code and data from the final executable
20896 is directly performed by the linker.
20898 @geindex -ffunction-sections (gcc)
20900 @geindex -fdata-sections (gcc)
20902 In order to do this, it has to work with objects compiled with the
20904 @code{-ffunction-sections} @code{-fdata-sections}.
20906 These options are usable with C and Ada files.
20907 They will place respectively each
20908 function or data in a separate section in the resulting object file.
20910 Once the objects and static libraries are created with these options, the
20911 linker can perform the dead code elimination. You can do this by setting
20912 the @code{-Wl,--gc-sections} option to gcc command or in the
20913 @code{-largs} section of @code{gnatmake}. This will perform a
20914 garbage collection of code and data never referenced.
20916 If the linker performs a partial link (@code{-r} linker option), then you
20917 will need to provide the entry point using the @code{-e} / @code{--entry}
20920 Note that objects compiled without the @code{-ffunction-sections} and
20921 @code{-fdata-sections} options can still be linked with the executable.
20922 However, no dead code elimination will be performed on those objects (they will
20925 The GNAT static library is now compiled with -ffunction-sections and
20926 -fdata-sections on some platforms. This allows you to eliminate the unused code
20927 and data of the GNAT library from your executable.
20929 @node Example of unused subprogram/data elimination,,Compilation options,Reducing Size of Executables with Unused Subprogram/Data Elimination
20930 @anchor{gnat_ugn/gnat_and_program_execution example-of-unused-subprogram-data-elimination}@anchor{19b}@anchor{gnat_ugn/gnat_and_program_execution id44}@anchor{19c}
20931 @subsubsection Example of unused subprogram/data elimination
20934 Here is a simple example:
20947 Used_Data : Integer;
20948 Unused_Data : Integer;
20950 procedure Used (Data : Integer);
20951 procedure Unused (Data : Integer);
20954 package body Aux is
20955 procedure Used (Data : Integer) is
20960 procedure Unused (Data : Integer) is
20962 Unused_Data := Data;
20968 @code{Unused} and @code{Unused_Data} are never referenced in this code
20969 excerpt, and hence they may be safely removed from the final executable.
20976 $ nm test | grep used
20977 020015f0 T aux__unused
20978 02005d88 B aux__unused_data
20979 020015cc T aux__used
20980 02005d84 B aux__used_data
20982 $ gnatmake test -cargs -fdata-sections -ffunction-sections \\
20983 -largs -Wl,--gc-sections
20985 $ nm test | grep used
20986 02005350 T aux__used
20987 0201ffe0 B aux__used_data
20991 It can be observed that the procedure @code{Unused} and the object
20992 @code{Unused_Data} are removed by the linker when using the
20993 appropriate options.
20995 @geindex Overflow checks
20997 @geindex Checks (overflow)
20999 @node Overflow Check Handling in GNAT,Performing Dimensionality Analysis in GNAT,Improving Performance,GNAT and Program Execution
21000 @anchor{gnat_ugn/gnat_and_program_execution id45}@anchor{149}@anchor{gnat_ugn/gnat_and_program_execution overflow-check-handling-in-gnat}@anchor{19d}
21001 @section Overflow Check Handling in GNAT
21004 This section explains how to control the handling of overflow checks.
21008 * Management of Overflows in GNAT::
21009 * Specifying the Desired Mode::
21010 * Default Settings::
21011 * Implementation Notes::
21015 @node Background,Management of Overflows in GNAT,,Overflow Check Handling in GNAT
21016 @anchor{gnat_ugn/gnat_and_program_execution background}@anchor{19e}@anchor{gnat_ugn/gnat_and_program_execution id46}@anchor{19f}
21017 @subsection Background
21020 Overflow checks are checks that the compiler may make to ensure
21021 that intermediate results are not out of range. For example:
21032 If @code{A} has the value @code{Integer'Last}, then the addition may cause
21033 overflow since the result is out of range of the type @code{Integer}.
21034 In this case @code{Constraint_Error} will be raised if checks are
21037 A trickier situation arises in examples like the following:
21048 where @code{A} is @code{Integer'Last} and @code{C} is @code{-1}.
21049 Now the final result of the expression on the right hand side is
21050 @code{Integer'Last} which is in range, but the question arises whether the
21051 intermediate addition of @code{(A + 1)} raises an overflow error.
21053 The (perhaps surprising) answer is that the Ada language
21054 definition does not answer this question. Instead it leaves
21055 it up to the implementation to do one of two things if overflow
21056 checks are enabled.
21062 raise an exception (@code{Constraint_Error}), or
21065 yield the correct mathematical result which is then used in
21066 subsequent operations.
21069 If the compiler chooses the first approach, then the assignment of this
21070 example will indeed raise @code{Constraint_Error} if overflow checking is
21071 enabled, or result in erroneous execution if overflow checks are suppressed.
21073 But if the compiler
21074 chooses the second approach, then it can perform both additions yielding
21075 the correct mathematical result, which is in range, so no exception
21076 will be raised, and the right result is obtained, regardless of whether
21077 overflow checks are suppressed.
21079 Note that in the first example an
21080 exception will be raised in either case, since if the compiler
21081 gives the correct mathematical result for the addition, it will
21082 be out of range of the target type of the assignment, and thus
21083 fails the range check.
21085 This lack of specified behavior in the handling of overflow for
21086 intermediate results is a source of non-portability, and can thus
21087 be problematic when programs are ported. Most typically this arises
21088 in a situation where the original compiler did not raise an exception,
21089 and then the application is moved to a compiler where the check is
21090 performed on the intermediate result and an unexpected exception is
21093 Furthermore, when using Ada 2012’s preconditions and other
21094 assertion forms, another issue arises. Consider:
21099 procedure P (A, B : Integer) with
21100 Pre => A + B <= Integer'Last;
21104 One often wants to regard arithmetic in a context like this from
21105 a mathematical point of view. So for example, if the two actual parameters
21106 for a call to @code{P} are both @code{Integer'Last}, then
21107 the precondition should be regarded as False. If we are executing
21108 in a mode with run-time checks enabled for preconditions, then we would
21109 like this precondition to fail, rather than raising an exception
21110 because of the intermediate overflow.
21112 However, the language definition leaves the specification of
21113 whether the above condition fails (raising @code{Assert_Error}) or
21114 causes an intermediate overflow (raising @code{Constraint_Error})
21115 up to the implementation.
21117 The situation is worse in a case such as the following:
21122 procedure Q (A, B, C : Integer) with
21123 Pre => A + B + C <= Integer'Last;
21132 Q (A => Integer'Last, B => 1, C => -1);
21136 From a mathematical point of view the precondition
21137 is True, but at run time we may (but are not guaranteed to) get an
21138 exception raised because of the intermediate overflow (and we really
21139 would prefer this precondition to be considered True at run time).
21141 @node Management of Overflows in GNAT,Specifying the Desired Mode,Background,Overflow Check Handling in GNAT
21142 @anchor{gnat_ugn/gnat_and_program_execution id47}@anchor{1a0}@anchor{gnat_ugn/gnat_and_program_execution management-of-overflows-in-gnat}@anchor{1a1}
21143 @subsection Management of Overflows in GNAT
21146 To deal with the portability issue, and with the problem of
21147 mathematical versus run-time interpretation of the expressions in
21148 assertions, GNAT provides comprehensive control over the handling
21149 of intermediate overflow. GNAT can operate in three modes, and
21150 furthemore, permits separate selection of operating modes for
21151 the expressions within assertions (here the term ‘assertions’
21152 is used in the technical sense, which includes preconditions and so forth)
21153 and for expressions appearing outside assertions.
21155 The three modes are:
21161 @emph{Use base type for intermediate operations} (@code{STRICT})
21163 In this mode, all intermediate results for predefined arithmetic
21164 operators are computed using the base type, and the result must
21165 be in range of the base type. If this is not the
21166 case then either an exception is raised (if overflow checks are
21167 enabled) or the execution is erroneous (if overflow checks are suppressed).
21168 This is the normal default mode.
21171 @emph{Most intermediate overflows avoided} (@code{MINIMIZED})
21173 In this mode, the compiler attempts to avoid intermediate overflows by
21174 using a larger integer type, typically @code{Long_Long_Integer},
21175 as the type in which arithmetic is
21176 performed for predefined arithmetic operators. This may be slightly more
21178 run time (compared to suppressing intermediate overflow checks), though
21179 the cost is negligible on modern 64-bit machines. For the examples given
21180 earlier, no intermediate overflows would have resulted in exceptions,
21181 since the intermediate results are all in the range of
21182 @code{Long_Long_Integer} (typically 64-bits on nearly all implementations
21183 of GNAT). In addition, if checks are enabled, this reduces the number of
21184 checks that must be made, so this choice may actually result in an
21185 improvement in space and time behavior.
21187 However, there are cases where @code{Long_Long_Integer} is not large
21188 enough, consider the following example:
21193 procedure R (A, B, C, D : Integer) with
21194 Pre => (A**2 * B**2) / (C**2 * D**2) <= 10;
21198 where @code{A} = @code{B} = @code{C} = @code{D} = @code{Integer'Last}.
21199 Now the intermediate results are
21200 out of the range of @code{Long_Long_Integer} even though the final result
21201 is in range and the precondition is True (from a mathematical point
21202 of view). In such a case, operating in this mode, an overflow occurs
21203 for the intermediate computation (which is why this mode
21204 says @emph{most} intermediate overflows are avoided). In this case,
21205 an exception is raised if overflow checks are enabled, and the
21206 execution is erroneous if overflow checks are suppressed.
21209 @emph{All intermediate overflows avoided} (@code{ELIMINATED})
21211 In this mode, the compiler avoids all intermediate overflows
21212 by using arbitrary precision arithmetic as required. In this
21213 mode, the above example with @code{A**2 * B**2} would
21214 not cause intermediate overflow, because the intermediate result
21215 would be evaluated using sufficient precision, and the result
21216 of evaluating the precondition would be True.
21218 This mode has the advantage of avoiding any intermediate
21219 overflows, but at the expense of significant run-time overhead,
21220 including the use of a library (included automatically in this
21221 mode) for multiple-precision arithmetic.
21223 This mode provides cleaner semantics for assertions, since now
21224 the run-time behavior emulates true arithmetic behavior for the
21225 predefined arithmetic operators, meaning that there is never a
21226 conflict between the mathematical view of the assertion, and its
21229 Note that in this mode, the behavior is unaffected by whether or
21230 not overflow checks are suppressed, since overflow does not occur.
21231 It is possible for gigantic intermediate expressions to raise
21232 @code{Storage_Error} as a result of attempting to compute the
21233 results of such expressions (e.g. @code{Integer'Last ** Integer'Last})
21234 but overflow is impossible.
21237 Note that these modes apply only to the evaluation of predefined
21238 arithmetic, membership, and comparison operators for signed integer
21241 For fixed-point arithmetic, checks can be suppressed. But if checks
21243 then fixed-point values are always checked for overflow against the
21244 base type for intermediate expressions (that is such checks always
21245 operate in the equivalent of @code{STRICT} mode).
21247 For floating-point, on nearly all architectures, @code{Machine_Overflows}
21248 is False, and IEEE infinities are generated, so overflow exceptions
21249 are never raised. If you want to avoid infinities, and check that
21250 final results of expressions are in range, then you can declare a
21251 constrained floating-point type, and range checks will be carried
21252 out in the normal manner (with infinite values always failing all
21255 @node Specifying the Desired Mode,Default Settings,Management of Overflows in GNAT,Overflow Check Handling in GNAT
21256 @anchor{gnat_ugn/gnat_and_program_execution id48}@anchor{1a2}@anchor{gnat_ugn/gnat_and_program_execution specifying-the-desired-mode}@anchor{e9}
21257 @subsection Specifying the Desired Mode
21260 @geindex pragma Overflow_Mode
21262 The desired mode of for handling intermediate overflow can be specified using
21263 either the @code{Overflow_Mode} pragma or an equivalent compiler switch.
21264 The pragma has the form
21269 pragma Overflow_Mode ([General =>] MODE [, [Assertions =>] MODE]);
21273 where @code{MODE} is one of
21279 @code{STRICT}: intermediate overflows checked (using base type)
21282 @code{MINIMIZED}: minimize intermediate overflows
21285 @code{ELIMINATED}: eliminate intermediate overflows
21288 The case is ignored, so @code{MINIMIZED}, @code{Minimized} and
21289 @code{minimized} all have the same effect.
21291 If only the @code{General} parameter is present, then the given @code{MODE} applies
21292 to expressions both within and outside assertions. If both arguments
21293 are present, then @code{General} applies to expressions outside assertions,
21294 and @code{Assertions} applies to expressions within assertions. For example:
21299 pragma Overflow_Mode
21300 (General => Minimized, Assertions => Eliminated);
21304 specifies that general expressions outside assertions be evaluated
21305 in ‘minimize intermediate overflows’ mode, and expressions within
21306 assertions be evaluated in ‘eliminate intermediate overflows’ mode.
21307 This is often a reasonable choice, avoiding excessive overhead
21308 outside assertions, but assuring a high degree of portability
21309 when importing code from another compiler, while incurring
21310 the extra overhead for assertion expressions to ensure that
21311 the behavior at run time matches the expected mathematical
21314 The @code{Overflow_Mode} pragma has the same scoping and placement
21315 rules as pragma @code{Suppress}, so it can occur either as a
21316 configuration pragma, specifying a default for the whole
21317 program, or in a declarative scope, where it applies to the
21318 remaining declarations and statements in that scope.
21320 Note that pragma @code{Overflow_Mode} does not affect whether
21321 overflow checks are enabled or suppressed. It only controls the
21322 method used to compute intermediate values. To control whether
21323 overflow checking is enabled or suppressed, use pragma @code{Suppress}
21324 or @code{Unsuppress} in the usual manner.
21326 @geindex -gnato? (gcc)
21328 @geindex -gnato?? (gcc)
21330 Additionally, a compiler switch @code{-gnato?} or @code{-gnato??}
21331 can be used to control the checking mode default (which can be subsequently
21332 overridden using pragmas).
21334 Here @code{?} is one of the digits @code{1} through @code{3}:
21339 @multitable {xxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
21346 use base type for intermediate operations (@code{STRICT})
21354 minimize intermediate overflows (@code{MINIMIZED})
21362 eliminate intermediate overflows (@code{ELIMINATED})
21368 As with the pragma, if only one digit appears then it applies to all
21369 cases; if two digits are given, then the first applies outside
21370 assertions, and the second within assertions. Thus the equivalent
21371 of the example pragma above would be
21374 If no digits follow the @code{-gnato}, then it is equivalent to
21376 causing all intermediate operations to be computed using the base
21377 type (@code{STRICT} mode).
21379 @node Default Settings,Implementation Notes,Specifying the Desired Mode,Overflow Check Handling in GNAT
21380 @anchor{gnat_ugn/gnat_and_program_execution default-settings}@anchor{1a3}@anchor{gnat_ugn/gnat_and_program_execution id49}@anchor{1a4}
21381 @subsection Default Settings
21384 The default mode for overflow checks is
21393 which causes all computations both inside and outside assertions to use
21396 This retains compatibility with previous versions of
21397 GNAT which suppressed overflow checks by default and always
21398 used the base type for computation of intermediate results.
21400 @c Sphinx allows no emphasis within :index: role. As a workaround we
21401 @c point the index to "switch" and use emphasis for "-gnato".
21404 @geindex -gnato (gcc)
21405 switch @code{-gnato} (with no digits following)
21415 which causes overflow checking of all intermediate overflows
21416 both inside and outside assertions against the base type.
21418 The pragma @code{Suppress (Overflow_Check)} disables overflow
21419 checking, but it has no effect on the method used for computing
21420 intermediate results.
21422 The pragma @code{Unsuppress (Overflow_Check)} enables overflow
21423 checking, but it has no effect on the method used for computing
21424 intermediate results.
21426 @node Implementation Notes,,Default Settings,Overflow Check Handling in GNAT
21427 @anchor{gnat_ugn/gnat_and_program_execution id50}@anchor{1a5}@anchor{gnat_ugn/gnat_and_program_execution implementation-notes}@anchor{1a6}
21428 @subsection Implementation Notes
21431 In practice on typical 64-bit machines, the @code{MINIMIZED} mode is
21432 reasonably efficient, and can be generally used. It also helps
21433 to ensure compatibility with code imported from some other
21436 Setting all intermediate overflows checking (@code{CHECKED} mode)
21437 makes sense if you want to
21438 make sure that your code is compatible with any other possible
21439 Ada implementation. This may be useful in ensuring portability
21440 for code that is to be exported to some other compiler than GNAT.
21442 The Ada standard allows the reassociation of expressions at
21443 the same precedence level if no parentheses are present. For
21444 example, @code{A+B+C} parses as though it were @code{(A+B)+C}, but
21445 the compiler can reintepret this as @code{A+(B+C)}, possibly
21446 introducing or eliminating an overflow exception. The GNAT
21447 compiler never takes advantage of this freedom, and the
21448 expression @code{A+B+C} will be evaluated as @code{(A+B)+C}.
21449 If you need the other order, you can write the parentheses
21450 explicitly @code{A+(B+C)} and GNAT will respect this order.
21452 The use of @code{ELIMINATED} mode will cause the compiler to
21453 automatically include an appropriate arbitrary precision
21454 integer arithmetic package. The compiler will make calls
21455 to this package, though only in cases where it cannot be
21456 sure that @code{Long_Long_Integer} is sufficient to guard against
21457 intermediate overflows. This package does not use dynamic
21458 allocation, but it does use the secondary stack, so an
21459 appropriate secondary stack package must be present (this
21460 is always true for standard full Ada, but may require
21461 specific steps for restricted run times such as ZFP).
21463 Although @code{ELIMINATED} mode causes expressions to use arbitrary
21464 precision arithmetic, avoiding overflow, the final result
21465 must be in an appropriate range. This is true even if the
21466 final result is of type @code{[Long_[Long_]]Integer'Base}, which
21467 still has the same bounds as its associated constrained
21470 Currently, the @code{ELIMINATED} mode is only available on target
21471 platforms for which @code{Long_Long_Integer} is 64-bits (nearly all GNAT
21474 @node Performing Dimensionality Analysis in GNAT,Stack Related Facilities,Overflow Check Handling in GNAT,GNAT and Program Execution
21475 @anchor{gnat_ugn/gnat_and_program_execution id51}@anchor{14a}@anchor{gnat_ugn/gnat_and_program_execution performing-dimensionality-analysis-in-gnat}@anchor{1a7}
21476 @section Performing Dimensionality Analysis in GNAT
21479 @geindex Dimensionality analysis
21481 The GNAT compiler supports dimensionality checking. The user can
21482 specify physical units for objects, and the compiler will verify that uses
21483 of these objects are compatible with their dimensions, in a fashion that is
21484 familiar to engineering practice. The dimensions of algebraic expressions
21485 (including powers with static exponents) are computed from their constituents.
21487 @geindex Dimension_System aspect
21489 @geindex Dimension aspect
21491 This feature depends on Ada 2012 aspect specifications, and is available from
21492 version 7.0.1 of GNAT onwards.
21493 The GNAT-specific aspect @code{Dimension_System}
21494 allows you to define a system of units; the aspect @code{Dimension}
21495 then allows the user to declare dimensioned quantities within a given system.
21496 (These aspects are described in the @emph{Implementation Defined Aspects}
21497 chapter of the @emph{GNAT Reference Manual}).
21499 The major advantage of this model is that it does not require the declaration of
21500 multiple operators for all possible combinations of types: it is only necessary
21501 to use the proper subtypes in object declarations.
21503 @geindex System.Dim.Mks package (GNAT library)
21505 @geindex MKS_Type type
21507 The simplest way to impose dimensionality checking on a computation is to make
21508 use of one of the instantiations of the package @code{System.Dim.Generic_Mks}, which
21509 are part of the GNAT library. This generic package defines a floating-point
21510 type @code{MKS_Type}, for which a sequence of dimension names are specified,
21511 together with their conventional abbreviations. The following should be read
21512 together with the full specification of the package, in file
21513 @code{s-digemk.ads}.
21517 @geindex s-digemk.ads file
21520 type Mks_Type is new Float_Type
21522 Dimension_System => (
21523 (Unit_Name => Meter, Unit_Symbol => 'm', Dim_Symbol => 'L'),
21524 (Unit_Name => Kilogram, Unit_Symbol => "kg", Dim_Symbol => 'M'),
21525 (Unit_Name => Second, Unit_Symbol => 's', Dim_Symbol => 'T'),
21526 (Unit_Name => Ampere, Unit_Symbol => 'A', Dim_Symbol => 'I'),
21527 (Unit_Name => Kelvin, Unit_Symbol => 'K', Dim_Symbol => "Theta"),
21528 (Unit_Name => Mole, Unit_Symbol => "mol", Dim_Symbol => 'N'),
21529 (Unit_Name => Candela, Unit_Symbol => "cd", Dim_Symbol => 'J'));
21533 The package then defines a series of subtypes that correspond to these
21534 conventional units. For example:
21539 subtype Length is Mks_Type
21541 Dimension => (Symbol => 'm', Meter => 1, others => 0);
21545 and similarly for @code{Mass}, @code{Time}, @code{Electric_Current},
21546 @code{Thermodynamic_Temperature}, @code{Amount_Of_Substance}, and
21547 @code{Luminous_Intensity} (the standard set of units of the SI system).
21549 The package also defines conventional names for values of each unit, for
21555 m : constant Length := 1.0;
21556 kg : constant Mass := 1.0;
21557 s : constant Time := 1.0;
21558 A : constant Electric_Current := 1.0;
21562 as well as useful multiples of these units:
21567 cm : constant Length := 1.0E-02;
21568 g : constant Mass := 1.0E-03;
21569 min : constant Time := 60.0;
21570 day : constant Time := 60.0 * 24.0 * min;
21575 There are three instantiations of @code{System.Dim.Generic_Mks} defined in the
21582 @code{System.Dim.Float_Mks} based on @code{Float} defined in @code{s-diflmk.ads}.
21585 @code{System.Dim.Long_Mks} based on @code{Long_Float} defined in @code{s-dilomk.ads}.
21588 @code{System.Dim.Mks} based on @code{Long_Long_Float} defined in @code{s-dimmks.ads}.
21591 Using one of these packages, you can then define a derived unit by providing
21592 the aspect that specifies its dimensions within the MKS system, as well as the
21593 string to be used for output of a value of that unit:
21598 subtype Acceleration is Mks_Type
21599 with Dimension => ("m/sec^2",
21606 Here is a complete example of use:
21611 with System.Dim.MKS; use System.Dim.Mks;
21612 with System.Dim.Mks_IO; use System.Dim.Mks_IO;
21613 with Text_IO; use Text_IO;
21614 procedure Free_Fall is
21615 subtype Acceleration is Mks_Type
21616 with Dimension => ("m/sec^2", 1, 0, -2, others => 0);
21617 G : constant acceleration := 9.81 * m / (s ** 2);
21618 T : Time := 10.0*s;
21622 Put ("Gravitational constant: ");
21623 Put (G, Aft => 2, Exp => 0); Put_Line ("");
21624 Distance := 0.5 * G * T ** 2;
21625 Put ("distance travelled in 10 seconds of free fall ");
21626 Put (Distance, Aft => 2, Exp => 0);
21632 Execution of this program yields:
21637 Gravitational constant: 9.81 m/sec^2
21638 distance travelled in 10 seconds of free fall 490.50 m
21642 However, incorrect assignments such as:
21648 Distance := 5.0 * kg;
21652 are rejected with the following diagnoses:
21658 >>> dimensions mismatch in assignment
21659 >>> left-hand side has dimension [L]
21660 >>> right-hand side is dimensionless
21662 Distance := 5.0 * kg:
21663 >>> dimensions mismatch in assignment
21664 >>> left-hand side has dimension [L]
21665 >>> right-hand side has dimension [M]
21669 The dimensions of an expression are properly displayed, even if there is
21670 no explicit subtype for it. If we add to the program:
21675 Put ("Final velocity: ");
21676 Put (G * T, Aft =>2, Exp =>0);
21681 then the output includes:
21686 Final velocity: 98.10 m.s**(-1)
21689 @geindex Dimensionable type
21691 @geindex Dimensioned subtype
21694 The type @code{Mks_Type} is said to be a @emph{dimensionable type} since it has a
21695 @code{Dimension_System} aspect, and the subtypes @code{Length}, @code{Mass}, etc.,
21696 are said to be @emph{dimensioned subtypes} since each one has a @code{Dimension}
21701 @geindex Dimension Vector (for a dimensioned subtype)
21703 @geindex Dimension aspect
21705 @geindex Dimension_System aspect
21708 The @code{Dimension} aspect of a dimensioned subtype @code{S} defines a mapping
21709 from the base type’s Unit_Names to integer (or, more generally, rational)
21710 values. This mapping is the @emph{dimension vector} (also referred to as the
21711 @emph{dimensionality}) for that subtype, denoted by @code{DV(S)}, and thus for each
21712 object of that subtype. Intuitively, the value specified for each
21713 @code{Unit_Name} is the exponent associated with that unit; a zero value
21714 means that the unit is not used. For example:
21720 Acc : Acceleration;
21728 Here @code{DV(Acc)} = @code{DV(Acceleration)} =
21729 @code{(Meter=>1, Kilogram=>0, Second=>-2, Ampere=>0, Kelvin=>0, Mole=>0, Candela=>0)}.
21730 Symbolically, we can express this as @code{Meter / Second**2}.
21732 The dimension vector of an arithmetic expression is synthesized from the
21733 dimension vectors of its components, with compile-time dimensionality checks
21734 that help prevent mismatches such as using an @code{Acceleration} where a
21735 @code{Length} is required.
21737 The dimension vector of the result of an arithmetic expression @emph{expr}, or
21738 @code{DV(@emph{expr})}, is defined as follows, assuming conventional
21739 mathematical definitions for the vector operations that are used:
21745 If @emph{expr} is of the type @emph{universal_real}, or is not of a dimensioned subtype,
21746 then @emph{expr} is dimensionless; @code{DV(@emph{expr})} is the empty vector.
21749 @code{DV(@emph{op expr})}, where @emph{op} is a unary operator, is @code{DV(@emph{expr})}
21752 @code{DV(@emph{expr1 op expr2})} where @emph{op} is “+” or “-” is @code{DV(@emph{expr1})}
21753 provided that @code{DV(@emph{expr1})} = @code{DV(@emph{expr2})}.
21754 If this condition is not met then the construct is illegal.
21757 @code{DV(@emph{expr1} * @emph{expr2})} is @code{DV(@emph{expr1})} + @code{DV(@emph{expr2})},
21758 and @code{DV(@emph{expr1} / @emph{expr2})} = @code{DV(@emph{expr1})} - @code{DV(@emph{expr2})}.
21759 In this context if one of the @emph{expr}s is dimensionless then its empty
21760 dimension vector is treated as @code{(others => 0)}.
21763 @code{DV(@emph{expr} ** @emph{power})} is @emph{power} * @code{DV(@emph{expr})},
21764 provided that @emph{power} is a static rational value. If this condition is not
21765 met then the construct is illegal.
21768 Note that, by the above rules, it is illegal to use binary “+” or “-” to
21769 combine a dimensioned and dimensionless value. Thus an expression such as
21770 @code{acc-10.0} is illegal, where @code{acc} is an object of subtype
21771 @code{Acceleration}.
21773 The dimensionality checks for relationals use the same rules as
21774 for “+” and “-“, except when comparing to a literal; thus
21792 and is thus illegal, but
21801 is accepted with a warning. Analogously a conditional expression requires the
21802 same dimension vector for each branch (with no exception for literals).
21804 The dimension vector of a type conversion @code{T(@emph{expr})} is defined
21805 as follows, based on the nature of @code{T}:
21811 If @code{T} is a dimensioned subtype then @code{DV(T(@emph{expr}))} is @code{DV(T)}
21812 provided that either @emph{expr} is dimensionless or
21813 @code{DV(T)} = @code{DV(@emph{expr})}. The conversion is illegal
21814 if @emph{expr} is dimensioned and @code{DV(@emph{expr})} /= @code{DV(T)}.
21815 Note that vector equality does not require that the corresponding
21816 Unit_Names be the same.
21818 As a consequence of the above rule, it is possible to convert between
21819 different dimension systems that follow the same international system
21820 of units, with the seven physical components given in the standard order
21821 (length, mass, time, etc.). Thus a length in meters can be converted to
21822 a length in inches (with a suitable conversion factor) but cannot be
21823 converted, for example, to a mass in pounds.
21826 If @code{T} is the base type for @emph{expr} (and the dimensionless root type of
21827 the dimension system), then @code{DV(T(@emph{expr}))} is @code{DV(expr)}.
21828 Thus, if @emph{expr} is of a dimensioned subtype of @code{T}, the conversion may
21829 be regarded as a “view conversion” that preserves dimensionality.
21831 This rule makes it possible to write generic code that can be instantiated
21832 with compatible dimensioned subtypes. The generic unit will contain
21833 conversions that will consequently be present in instantiations, but
21834 conversions to the base type will preserve dimensionality and make it
21835 possible to write generic code that is correct with respect to
21839 Otherwise (i.e., @code{T} is neither a dimensioned subtype nor a dimensionable
21840 base type), @code{DV(T(@emph{expr}))} is the empty vector. Thus a dimensioned
21841 value can be explicitly converted to a non-dimensioned subtype, which
21842 of course then escapes dimensionality analysis.
21845 The dimension vector for a type qualification @code{T'(@emph{expr})} is the same
21846 as for the type conversion @code{T(@emph{expr})}.
21848 An assignment statement
21857 requires @code{DV(Source)} = @code{DV(Target)}, and analogously for parameter
21858 passing (the dimension vector for the actual parameter must be equal to the
21859 dimension vector for the formal parameter).
21861 @node Stack Related Facilities,Memory Management Issues,Performing Dimensionality Analysis in GNAT,GNAT and Program Execution
21862 @anchor{gnat_ugn/gnat_and_program_execution id52}@anchor{14b}@anchor{gnat_ugn/gnat_and_program_execution stack-related-facilities}@anchor{1a8}
21863 @section Stack Related Facilities
21866 This section describes some useful tools associated with stack
21867 checking and analysis. In
21868 particular, it deals with dynamic and static stack usage measurements.
21871 * Stack Overflow Checking::
21872 * Static Stack Usage Analysis::
21873 * Dynamic Stack Usage Analysis::
21877 @node Stack Overflow Checking,Static Stack Usage Analysis,,Stack Related Facilities
21878 @anchor{gnat_ugn/gnat_and_program_execution id53}@anchor{1a9}@anchor{gnat_ugn/gnat_and_program_execution stack-overflow-checking}@anchor{e5}
21879 @subsection Stack Overflow Checking
21882 @geindex Stack Overflow Checking
21884 @geindex -fstack-check (gcc)
21886 For most operating systems, @code{gcc} does not perform stack overflow
21887 checking by default. This means that if the main environment task or
21888 some other task exceeds the available stack space, then unpredictable
21889 behavior will occur. Most native systems offer some level of protection by
21890 adding a guard page at the end of each task stack. This mechanism is usually
21891 not enough for dealing properly with stack overflow situations because
21892 a large local variable could “jump” above the guard page.
21893 Furthermore, when the
21894 guard page is hit, there may not be any space left on the stack for executing
21895 the exception propagation code. Enabling stack checking avoids
21898 To activate stack checking, compile all units with the @code{gcc} option
21899 @code{-fstack-check}. For example:
21904 $ gcc -c -fstack-check package1.adb
21908 Units compiled with this option will generate extra instructions to check
21909 that any use of the stack (for procedure calls or for declaring local
21910 variables in declare blocks) does not exceed the available stack space.
21911 If the space is exceeded, then a @code{Storage_Error} exception is raised.
21913 For declared tasks, the default stack size is defined by the GNAT runtime,
21914 whose size may be modified at bind time through the @code{-d} bind switch
21915 (@ref{110,,Switches for gnatbind}). Task specific stack sizes may be set using the
21916 @code{Storage_Size} pragma.
21918 For the environment task, the stack size is determined by the operating system.
21919 Consequently, to modify the size of the environment task please refer to your
21920 operating system documentation.
21922 @node Static Stack Usage Analysis,Dynamic Stack Usage Analysis,Stack Overflow Checking,Stack Related Facilities
21923 @anchor{gnat_ugn/gnat_and_program_execution id54}@anchor{1aa}@anchor{gnat_ugn/gnat_and_program_execution static-stack-usage-analysis}@anchor{e6}
21924 @subsection Static Stack Usage Analysis
21927 @geindex Static Stack Usage Analysis
21929 @geindex -fstack-usage
21931 A unit compiled with @code{-fstack-usage} will generate an extra file
21933 the maximum amount of stack used, on a per-function basis.
21934 The file has the same
21935 basename as the target object file with a @code{.su} extension.
21936 Each line of this file is made up of three fields:
21942 The name of the function.
21948 One or more qualifiers: @code{static}, @code{dynamic}, @code{bounded}.
21951 The second field corresponds to the size of the known part of the function
21954 The qualifier @code{static} means that the function frame size
21956 It usually means that all local variables have a static size.
21957 In this case, the second field is a reliable measure of the function stack
21960 The qualifier @code{dynamic} means that the function frame size is not static.
21961 It happens mainly when some local variables have a dynamic size. When this
21962 qualifier appears alone, the second field is not a reliable measure
21963 of the function stack analysis. When it is qualified with @code{bounded}, it
21964 means that the second field is a reliable maximum of the function stack
21967 A unit compiled with @code{-Wstack-usage} will issue a warning for each
21968 subprogram whose stack usage might be larger than the specified amount of
21969 bytes. The wording is in keeping with the qualifier documented above.
21971 @node Dynamic Stack Usage Analysis,,Static Stack Usage Analysis,Stack Related Facilities
21972 @anchor{gnat_ugn/gnat_and_program_execution dynamic-stack-usage-analysis}@anchor{113}@anchor{gnat_ugn/gnat_and_program_execution id55}@anchor{1ab}
21973 @subsection Dynamic Stack Usage Analysis
21976 It is possible to measure the maximum amount of stack used by a task, by
21977 adding a switch to @code{gnatbind}, as:
21982 $ gnatbind -u0 file
21986 With this option, at each task termination, its stack usage is output on
21988 Note that this switch is not compatible with tools like
21989 Valgrind and DrMemory; they will report errors.
21991 It is not always convenient to output the stack usage when the program
21992 is still running. Hence, it is possible to delay this output until program
21993 termination. for a given number of tasks specified as the argument of the
21994 @code{-u} option. For instance:
21999 $ gnatbind -u100 file
22003 will buffer the stack usage information of the first 100 tasks to terminate and
22004 output this info at program termination. Results are displayed in four
22010 Index | Task Name | Stack Size | Stack Usage
22020 @emph{Index} is a number associated with each task.
22023 @emph{Task Name} is the name of the task analyzed.
22026 @emph{Stack Size} is the maximum size for the stack.
22029 @emph{Stack Usage} is the measure done by the stack analyzer.
22030 In order to prevent overflow, the stack
22031 is not entirely analyzed, and it’s not possible to know exactly how
22032 much has actually been used.
22035 By default the environment task stack, the stack that contains the main unit,
22036 is not processed. To enable processing of the environment task stack, the
22037 environment variable GNAT_STACK_LIMIT needs to be set to the maximum size of
22038 the environment task stack. This amount is given in kilobytes. For example:
22043 $ set GNAT_STACK_LIMIT 1600
22047 would specify to the analyzer that the environment task stack has a limit
22048 of 1.6 megabytes. Any stack usage beyond this will be ignored by the analysis.
22050 The package @code{GNAT.Task_Stack_Usage} provides facilities to get
22051 stack-usage reports at run time. See its body for the details.
22053 @node Memory Management Issues,,Stack Related Facilities,GNAT and Program Execution
22054 @anchor{gnat_ugn/gnat_and_program_execution id56}@anchor{14c}@anchor{gnat_ugn/gnat_and_program_execution memory-management-issues}@anchor{1ac}
22055 @section Memory Management Issues
22058 This section describes some useful memory pools provided in the GNAT library
22059 and in particular the GNAT Debug Pool facility, which can be used to detect
22060 incorrect uses of access values (including ‘dangling references’).
22064 * Some Useful Memory Pools::
22065 * The GNAT Debug Pool Facility::
22069 @node Some Useful Memory Pools,The GNAT Debug Pool Facility,,Memory Management Issues
22070 @anchor{gnat_ugn/gnat_and_program_execution id57}@anchor{1ad}@anchor{gnat_ugn/gnat_and_program_execution some-useful-memory-pools}@anchor{1ae}
22071 @subsection Some Useful Memory Pools
22074 @geindex Memory Pool
22079 The @code{System.Pool_Global} package offers the Unbounded_No_Reclaim_Pool
22080 storage pool. Allocations use the standard system call @code{malloc} while
22081 deallocations use the standard system call @code{free}. No reclamation is
22082 performed when the pool goes out of scope. For performance reasons, the
22083 standard default Ada allocators/deallocators do not use any explicit storage
22084 pools but if they did, they could use this storage pool without any change in
22085 behavior. That is why this storage pool is used when the user
22086 manages to make the default implicit allocator explicit as in this example:
22091 type T1 is access Something;
22092 -- no Storage pool is defined for T2
22094 type T2 is access Something_Else;
22095 for T2'Storage_Pool use T1'Storage_Pool;
22096 -- the above is equivalent to
22097 for T2'Storage_Pool use System.Pool_Global.Global_Pool_Object;
22101 The @code{System.Pool_Local} package offers the @code{Unbounded_Reclaim_Pool} storage
22102 pool. The allocation strategy is similar to @code{Pool_Local}
22103 except that the all
22104 storage allocated with this pool is reclaimed when the pool object goes out of
22105 scope. This pool provides a explicit mechanism similar to the implicit one
22106 provided by several Ada 83 compilers for allocations performed through a local
22107 access type and whose purpose was to reclaim memory when exiting the
22108 scope of a given local access. As an example, the following program does not
22109 leak memory even though it does not perform explicit deallocation:
22114 with System.Pool_Local;
22115 procedure Pooloc1 is
22116 procedure Internal is
22117 type A is access Integer;
22118 X : System.Pool_Local.Unbounded_Reclaim_Pool;
22119 for A'Storage_Pool use X;
22122 for I in 1 .. 50 loop
22127 for I in 1 .. 100 loop
22134 The @code{System.Pool_Size} package implements the @code{Stack_Bounded_Pool} used when
22135 @code{Storage_Size} is specified for an access type.
22136 The whole storage for the pool is
22137 allocated at once, usually on the stack at the point where the access type is
22138 elaborated. It is automatically reclaimed when exiting the scope where the
22139 access type is defined. This package is not intended to be used directly by the
22140 user and it is implicitly used for each such declaration:
22145 type T1 is access Something;
22146 for T1'Storage_Size use 10_000;
22150 @node The GNAT Debug Pool Facility,,Some Useful Memory Pools,Memory Management Issues
22151 @anchor{gnat_ugn/gnat_and_program_execution id58}@anchor{1af}@anchor{gnat_ugn/gnat_and_program_execution the-gnat-debug-pool-facility}@anchor{1b0}
22152 @subsection The GNAT Debug Pool Facility
22155 @geindex Debug Pool
22159 @geindex memory corruption
22161 The use of unchecked deallocation and unchecked conversion can easily
22162 lead to incorrect memory references. The problems generated by such
22163 references are usually difficult to tackle because the symptoms can be
22164 very remote from the origin of the problem. In such cases, it is
22165 very helpful to detect the problem as early as possible. This is the
22166 purpose of the Storage Pool provided by @code{GNAT.Debug_Pools}.
22168 In order to use the GNAT specific debugging pool, the user must
22169 associate a debug pool object with each of the access types that may be
22170 related to suspected memory problems. See Ada Reference Manual 13.11.
22175 type Ptr is access Some_Type;
22176 Pool : GNAT.Debug_Pools.Debug_Pool;
22177 for Ptr'Storage_Pool use Pool;
22181 @code{GNAT.Debug_Pools} is derived from a GNAT-specific kind of
22182 pool: the @code{Checked_Pool}. Such pools, like standard Ada storage pools,
22183 allow the user to redefine allocation and deallocation strategies. They
22184 also provide a checkpoint for each dereference, through the use of
22185 the primitive operation @code{Dereference} which is implicitly called at
22186 each dereference of an access value.
22188 Once an access type has been associated with a debug pool, operations on
22189 values of the type may raise four distinct exceptions,
22190 which correspond to four potential kinds of memory corruption:
22196 @code{GNAT.Debug_Pools.Accessing_Not_Allocated_Storage}
22199 @code{GNAT.Debug_Pools.Accessing_Deallocated_Storage}
22202 @code{GNAT.Debug_Pools.Freeing_Not_Allocated_Storage}
22205 @code{GNAT.Debug_Pools.Freeing_Deallocated_Storage}
22208 For types associated with a Debug_Pool, dynamic allocation is performed using
22209 the standard GNAT allocation routine. References to all allocated chunks of
22210 memory are kept in an internal dictionary. Several deallocation strategies are
22211 provided, whereupon the user can choose to release the memory to the system,
22212 keep it allocated for further invalid access checks, or fill it with an easily
22213 recognizable pattern for debug sessions. The memory pattern is the old IBM
22214 hexadecimal convention: @code{16#DEADBEEF#}.
22216 See the documentation in the file g-debpoo.ads for more information on the
22217 various strategies.
22219 Upon each dereference, a check is made that the access value denotes a
22220 properly allocated memory location. Here is a complete example of use of
22221 @code{Debug_Pools}, that includes typical instances of memory corruption:
22226 with Gnat.Io; use Gnat.Io;
22227 with Unchecked_Deallocation;
22228 with Unchecked_Conversion;
22229 with GNAT.Debug_Pools;
22230 with System.Storage_Elements;
22231 with Ada.Exceptions; use Ada.Exceptions;
22232 procedure Debug_Pool_Test is
22234 type T is access Integer;
22235 type U is access all T;
22237 P : GNAT.Debug_Pools.Debug_Pool;
22238 for T'Storage_Pool use P;
22240 procedure Free is new Unchecked_Deallocation (Integer, T);
22241 function UC is new Unchecked_Conversion (U, T);
22244 procedure Info is new GNAT.Debug_Pools.Print_Info(Put_Line);
22254 Put_Line (Integer'Image(B.all));
22256 when E : others => Put_Line ("raised: " & Exception_Name (E));
22261 when E : others => Put_Line ("raised: " & Exception_Name (E));
22265 Put_Line (Integer'Image(B.all));
22267 when E : others => Put_Line ("raised: " & Exception_Name (E));
22272 when E : others => Put_Line ("raised: " & Exception_Name (E));
22275 end Debug_Pool_Test;
22279 The debug pool mechanism provides the following precise diagnostics on the
22280 execution of this erroneous program:
22286 Total allocated bytes : 0
22287 Total deallocated bytes : 0
22288 Current Water Mark: 0
22292 Total allocated bytes : 8
22293 Total deallocated bytes : 0
22294 Current Water Mark: 8
22297 raised: GNAT.DEBUG_POOLS.ACCESSING_DEALLOCATED_STORAGE
22298 raised: GNAT.DEBUG_POOLS.FREEING_DEALLOCATED_STORAGE
22299 raised: GNAT.DEBUG_POOLS.ACCESSING_NOT_ALLOCATED_STORAGE
22300 raised: GNAT.DEBUG_POOLS.FREEING_NOT_ALLOCATED_STORAGE
22302 Total allocated bytes : 8
22303 Total deallocated bytes : 4
22304 Current Water Mark: 4
22310 @c -- Non-breaking space in running text
22311 @c -- E.g. Ada |nbsp| 95
22313 @node Platform-Specific Information,Example of Binder Output File,GNAT and Program Execution,Top
22314 @anchor{gnat_ugn/platform_specific_information doc}@anchor{1b1}@anchor{gnat_ugn/platform_specific_information id1}@anchor{1b2}@anchor{gnat_ugn/platform_specific_information platform-specific-information}@anchor{d}
22315 @chapter Platform-Specific Information
22318 This appendix contains information relating to the implementation
22319 of run-time libraries on various platforms and also covers
22320 topics related to the GNAT implementation on Windows and Mac OS.
22323 * Run-Time Libraries::
22324 * Specifying a Run-Time Library::
22325 * GNU/Linux Topics::
22326 * Microsoft Windows Topics::
22331 @node Run-Time Libraries,Specifying a Run-Time Library,,Platform-Specific Information
22332 @anchor{gnat_ugn/platform_specific_information id2}@anchor{1b3}@anchor{gnat_ugn/platform_specific_information run-time-libraries}@anchor{1b4}
22333 @section Run-Time Libraries
22336 @geindex Tasking and threads libraries
22338 @geindex Threads libraries and tasking
22340 @geindex Run-time libraries (platform-specific information)
22342 The GNAT run-time implementation may vary with respect to both the
22343 underlying threads library and the exception-handling scheme.
22344 For threads support, the default run-time will bind to the thread
22345 package of the underlying operating system.
22347 For exception handling, either or both of two models are supplied:
22351 @geindex Zero-Cost Exceptions
22353 @geindex ZCX (Zero-Cost Exceptions)
22360 @strong{Zero-Cost Exceptions} (“ZCX”),
22361 which uses binder-generated tables that
22362 are interrogated at run time to locate a handler.
22364 @geindex setjmp/longjmp Exception Model
22366 @geindex SJLJ (setjmp/longjmp Exception Model)
22369 @strong{setjmp / longjmp} (‘SJLJ’),
22370 which uses dynamically-set data to establish
22371 the set of handlers
22374 Most programs should experience a substantial speed improvement by
22375 being compiled with a ZCX run-time.
22376 This is especially true for
22377 tasking applications or applications with many exception handlers.
22378 Note however that the ZCX run-time does not support asynchronous abort
22379 of tasks (@code{abort} and @code{select-then-abort} constructs) and will instead
22380 implement abort by polling points in the runtime. You can also add additional
22381 polling points explicitly if needed in your application via @code{pragma
22384 This section summarizes which combinations of threads and exception support
22385 are supplied on various GNAT platforms.
22388 * Summary of Run-Time Configurations::
22392 @node Summary of Run-Time Configurations,,,Run-Time Libraries
22393 @anchor{gnat_ugn/platform_specific_information id3}@anchor{1b5}@anchor{gnat_ugn/platform_specific_information summary-of-run-time-configurations}@anchor{1b6}
22394 @subsection Summary of Run-Time Configurations
22398 @multitable {xxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxx} {xxxxxxxxxxxxxx}
22455 native Win32 threads
22467 native Win32 threads
22492 @node Specifying a Run-Time Library,GNU/Linux Topics,Run-Time Libraries,Platform-Specific Information
22493 @anchor{gnat_ugn/platform_specific_information id4}@anchor{1b7}@anchor{gnat_ugn/platform_specific_information specifying-a-run-time-library}@anchor{1b8}
22494 @section Specifying a Run-Time Library
22497 The @code{adainclude} subdirectory containing the sources of the GNAT
22498 run-time library, and the @code{adalib} subdirectory containing the
22499 @code{ALI} files and the static and/or shared GNAT library, are located
22500 in the gcc target-dependent area:
22505 target=$prefix/lib/gcc/gcc-*dumpmachine*/gcc-*dumpversion*/
22509 As indicated above, on some platforms several run-time libraries are supplied.
22510 These libraries are installed in the target dependent area and
22511 contain a complete source and binary subdirectory. The detailed description
22512 below explains the differences between the different libraries in terms of
22513 their thread support.
22515 The default run-time library (when GNAT is installed) is @emph{rts-native}.
22516 This default run-time is selected by the means of soft links.
22517 For example on x86-linux:
22520 @c -- $(target-dir)
22522 @c -- +--- adainclude----------+
22524 @c -- +--- adalib-----------+ |
22526 @c -- +--- rts-native | |
22528 @c -- | +--- adainclude <---+
22530 @c -- | +--- adalib <----+
22532 @c -- +--- rts-sjlj
22534 @c -- +--- adainclude
22542 _______/ / \ \_________________
22545 ADAINCLUDE ADALIB rts-native rts-sjlj
22550 +-------------> adainclude adalib adainclude adalib
22553 +---------------------+
22555 Run-Time Library Directory Structure
22556 (Upper-case names and dotted/dashed arrows represent soft links)
22559 If the @emph{rts-sjlj} library is to be selected on a permanent basis,
22560 these soft links can be modified with the following commands:
22566 $ rm -f adainclude adalib
22567 $ ln -s rts-sjlj/adainclude adainclude
22568 $ ln -s rts-sjlj/adalib adalib
22572 Alternatively, you can specify @code{rts-sjlj/adainclude} in the file
22573 @code{$target/ada_source_path} and @code{rts-sjlj/adalib} in
22574 @code{$target/ada_object_path}.
22576 @geindex --RTS option
22578 Selecting another run-time library temporarily can be
22579 achieved by using the @code{--RTS} switch, e.g., @code{--RTS=sjlj}
22580 @anchor{gnat_ugn/platform_specific_information choosing-the-scheduling-policy}@anchor{1b9}
22581 @geindex SCHED_FIFO scheduling policy
22583 @geindex SCHED_RR scheduling policy
22585 @geindex SCHED_OTHER scheduling policy
22588 * Choosing the Scheduling Policy::
22592 @node Choosing the Scheduling Policy,,,Specifying a Run-Time Library
22593 @anchor{gnat_ugn/platform_specific_information id5}@anchor{1ba}
22594 @subsection Choosing the Scheduling Policy
22597 When using a POSIX threads implementation, you have a choice of several
22598 scheduling policies: @code{SCHED_FIFO}, @code{SCHED_RR} and @code{SCHED_OTHER}.
22600 Typically, the default is @code{SCHED_OTHER}, while using @code{SCHED_FIFO}
22601 or @code{SCHED_RR} requires special (e.g., root) privileges.
22603 @geindex pragma Time_Slice
22605 @geindex -T0 option
22607 @geindex pragma Task_Dispatching_Policy
22609 By default, GNAT uses the @code{SCHED_OTHER} policy. To specify
22611 you can use one of the following:
22617 @code{pragma Time_Slice (0.0)}
22620 the corresponding binder option @code{-T0}
22623 @code{pragma Task_Dispatching_Policy (FIFO_Within_Priorities)}
22626 To specify @code{SCHED_RR},
22627 you should use @code{pragma Time_Slice} with a
22628 value greater than 0.0, or else use the corresponding @code{-T}
22631 To make sure a program is running as root, you can put something like
22632 this in a library package body in your application:
22637 function geteuid return Integer;
22638 pragma Import (C, geteuid, "geteuid");
22639 Ignore : constant Boolean :=
22640 (if geteuid = 0 then True else raise Program_Error with "must be root");
22644 It gets the effective user id, and if it’s not 0 (i.e. root), it raises
22651 @node GNU/Linux Topics,Microsoft Windows Topics,Specifying a Run-Time Library,Platform-Specific Information
22652 @anchor{gnat_ugn/platform_specific_information gnu-linux-topics}@anchor{1bb}@anchor{gnat_ugn/platform_specific_information id6}@anchor{1bc}
22653 @section GNU/Linux Topics
22656 This section describes topics that are specific to GNU/Linux platforms.
22659 * Required Packages on GNU/Linux::
22663 @node Required Packages on GNU/Linux,,,GNU/Linux Topics
22664 @anchor{gnat_ugn/platform_specific_information id7}@anchor{1bd}@anchor{gnat_ugn/platform_specific_information required-packages-on-gnu-linux}@anchor{1be}
22665 @subsection Required Packages on GNU/Linux
22668 GNAT requires the C library developer’s package to be installed.
22669 The name of of that package depends on your GNU/Linux distribution:
22675 RedHat, SUSE: @code{glibc-devel};
22678 Debian, Ubuntu: @code{libc6-dev} (normally installed by default).
22681 If using the 32-bit version of GNAT on a 64-bit version of GNU/Linux,
22682 you’ll need the 32-bit version of the following packages:
22688 RedHat, SUSE: @code{glibc.i686}, @code{glibc-devel.i686}, @code{ncurses-libs.i686}
22691 Debian, Ubuntu: @code{libc6:i386}, @code{libc6-dev:i386}, @code{lib32ncursesw5}
22694 Other GNU/Linux distributions might be choosing a different name
22695 for those packages.
22699 @node Microsoft Windows Topics,Mac OS Topics,GNU/Linux Topics,Platform-Specific Information
22700 @anchor{gnat_ugn/platform_specific_information id8}@anchor{1bf}@anchor{gnat_ugn/platform_specific_information microsoft-windows-topics}@anchor{1c0}
22701 @section Microsoft Windows Topics
22704 This section describes topics that are specific to the Microsoft Windows
22709 * Using GNAT on Windows::
22710 * Using a network installation of GNAT::
22711 * CONSOLE and WINDOWS subsystems::
22712 * Temporary Files::
22713 * Disabling Command Line Argument Expansion::
22714 * Windows Socket Timeouts::
22715 * Mixed-Language Programming on Windows::
22716 * Windows Specific Add-Ons::
22720 @node Using GNAT on Windows,Using a network installation of GNAT,,Microsoft Windows Topics
22721 @anchor{gnat_ugn/platform_specific_information id9}@anchor{1c1}@anchor{gnat_ugn/platform_specific_information using-gnat-on-windows}@anchor{1c2}
22722 @subsection Using GNAT on Windows
22725 One of the strengths of the GNAT technology is that its tool set
22726 (@code{gcc}, @code{gnatbind}, @code{gnatlink}, @code{gnatmake}, the
22727 @code{gdb} debugger, etc.) is used in the same way regardless of the
22730 On Windows this tool set is complemented by a number of Microsoft-specific
22731 tools that have been provided to facilitate interoperability with Windows
22732 when this is required. With these tools:
22738 You can build applications using the @code{CONSOLE} or @code{WINDOWS}
22742 You can use any Dynamically Linked Library (DLL) in your Ada code (both
22743 relocatable and non-relocatable DLLs are supported).
22746 You can build Ada DLLs for use in other applications. These applications
22747 can be written in a language other than Ada (e.g., C, C++, etc). Again both
22748 relocatable and non-relocatable Ada DLLs are supported.
22751 You can include Windows resources in your Ada application.
22754 You can use or create COM/DCOM objects.
22757 Immediately below are listed all known general GNAT-for-Windows restrictions.
22758 Other restrictions about specific features like Windows Resources and DLLs
22759 are listed in separate sections below.
22765 It is not possible to use @code{GetLastError} and @code{SetLastError}
22766 when tasking, protected records, or exceptions are used. In these
22767 cases, in order to implement Ada semantics, the GNAT run-time system
22768 calls certain Win32 routines that set the last error variable to 0 upon
22769 success. It should be possible to use @code{GetLastError} and
22770 @code{SetLastError} when tasking, protected record, and exception
22771 features are not used, but it is not guaranteed to work.
22774 It is not possible to link against Microsoft C++ libraries except for
22775 import libraries. Interfacing must be done by the mean of DLLs.
22778 It is possible to link against Microsoft C libraries. Yet the preferred
22779 solution is to use C/C++ compiler that comes with GNAT, since it
22780 doesn’t require having two different development environments and makes the
22781 inter-language debugging experience smoother.
22784 When the compilation environment is located on FAT32 drives, users may
22785 experience recompilations of the source files that have not changed if
22786 Daylight Saving Time (DST) state has changed since the last time files
22787 were compiled. NTFS drives do not have this problem.
22790 No components of the GNAT toolset use any entries in the Windows
22791 registry. The only entries that can be created are file associations and
22792 PATH settings, provided the user has chosen to create them at installation
22793 time, as well as some minimal book-keeping information needed to correctly
22794 uninstall or integrate different GNAT products.
22797 @node Using a network installation of GNAT,CONSOLE and WINDOWS subsystems,Using GNAT on Windows,Microsoft Windows Topics
22798 @anchor{gnat_ugn/platform_specific_information id10}@anchor{1c3}@anchor{gnat_ugn/platform_specific_information using-a-network-installation-of-gnat}@anchor{1c4}
22799 @subsection Using a network installation of GNAT
22802 Make sure the system on which GNAT is installed is accessible from the
22803 current machine, i.e., the install location is shared over the network.
22804 Shared resources are accessed on Windows by means of UNC paths, which
22805 have the format @code{\\\\server\\sharename\\path}
22807 In order to use such a network installation, simply add the UNC path of the
22808 @code{bin} directory of your GNAT installation in front of your PATH. For
22809 example, if GNAT is installed in @code{\GNAT} directory of a share location
22810 called @code{c-drive} on a machine @code{LOKI}, the following command will
22816 $ path \\loki\c-drive\gnat\bin;%path%`
22820 Be aware that every compilation using the network installation results in the
22821 transfer of large amounts of data across the network and will likely cause
22822 serious performance penalty.
22824 @node CONSOLE and WINDOWS subsystems,Temporary Files,Using a network installation of GNAT,Microsoft Windows Topics
22825 @anchor{gnat_ugn/platform_specific_information console-and-windows-subsystems}@anchor{1c5}@anchor{gnat_ugn/platform_specific_information id11}@anchor{1c6}
22826 @subsection CONSOLE and WINDOWS subsystems
22829 @geindex CONSOLE Subsystem
22831 @geindex WINDOWS Subsystem
22835 There are two main subsystems under Windows. The @code{CONSOLE} subsystem
22836 (which is the default subsystem) will always create a console when
22837 launching the application. This is not something desirable when the
22838 application has a Windows GUI. To get rid of this console the
22839 application must be using the @code{WINDOWS} subsystem. To do so
22840 the @code{-mwindows} linker option must be specified.
22845 $ gnatmake winprog -largs -mwindows
22849 @node Temporary Files,Disabling Command Line Argument Expansion,CONSOLE and WINDOWS subsystems,Microsoft Windows Topics
22850 @anchor{gnat_ugn/platform_specific_information id12}@anchor{1c7}@anchor{gnat_ugn/platform_specific_information temporary-files}@anchor{1c8}
22851 @subsection Temporary Files
22854 @geindex Temporary files
22856 It is possible to control where temporary files gets created by setting
22859 @geindex environment variable; TMP
22860 @code{TMP} environment variable. The file will be created:
22866 Under the directory pointed to by the
22868 @geindex environment variable; TMP
22869 @code{TMP} environment variable if
22870 this directory exists.
22873 Under @code{c:\temp}, if the
22875 @geindex environment variable; TMP
22876 @code{TMP} environment variable is not
22877 set (or not pointing to a directory) and if this directory exists.
22880 Under the current working directory otherwise.
22883 This allows you to determine exactly where the temporary
22884 file will be created. This is particularly useful in networked
22885 environments where you may not have write access to some
22888 @node Disabling Command Line Argument Expansion,Windows Socket Timeouts,Temporary Files,Microsoft Windows Topics
22889 @anchor{gnat_ugn/platform_specific_information disabling-command-line-argument-expansion}@anchor{1c9}
22890 @subsection Disabling Command Line Argument Expansion
22893 @geindex Command Line Argument Expansion
22895 By default, an executable compiled for the Windows platform will do
22896 the following postprocessing on the arguments passed on the command
22903 If the argument contains the characters @code{*} and/or @code{?}, then
22904 file expansion will be attempted. For example, if the current directory
22905 contains @code{a.txt} and @code{b.txt}, then when calling:
22908 $ my_ada_program *.txt
22911 The following arguments will effectively be passed to the main program
22912 (for example when using @code{Ada.Command_Line.Argument}):
22915 Ada.Command_Line.Argument (1) -> "a.txt"
22916 Ada.Command_Line.Argument (2) -> "b.txt"
22920 Filename expansion can be disabled for a given argument by using single
22921 quotes. Thus, calling:
22924 $ my_ada_program '*.txt'
22930 Ada.Command_Line.Argument (1) -> "*.txt"
22934 Note that if the program is launched from a shell such as Cygwin Bash
22935 then quote removal might be performed by the shell.
22937 In some contexts it might be useful to disable this feature (for example if
22938 the program performs its own argument expansion). In order to do this, a C
22939 symbol needs to be defined and set to @code{0}. You can do this by
22940 adding the following code fragment in one of your Ada units:
22943 Do_Argv_Expansion : Integer := 0;
22944 pragma Export (C, Do_Argv_Expansion, "__gnat_do_argv_expansion");
22947 The results of previous examples will be respectively:
22950 Ada.Command_Line.Argument (1) -> "*.txt"
22956 Ada.Command_Line.Argument (1) -> "'*.txt'"
22959 @node Windows Socket Timeouts,Mixed-Language Programming on Windows,Disabling Command Line Argument Expansion,Microsoft Windows Topics
22960 @anchor{gnat_ugn/platform_specific_information windows-socket-timeouts}@anchor{1ca}
22961 @subsection Windows Socket Timeouts
22964 Microsoft Windows desktops older than @code{8.0} and Microsoft Windows Servers
22965 older than @code{2019} set a socket timeout 500 milliseconds longer than the value
22966 set by setsockopt with @code{SO_RCVTIMEO} and @code{SO_SNDTIMEO} options. The GNAT
22967 runtime makes a correction for the difference in the corresponding Windows
22968 versions. For Windows Server starting with version @code{2019}, the user must
22969 provide a manifest file for the GNAT runtime to be able to recognize that
22970 the Windows version does not need the timeout correction. The manifest file
22971 should be located in the same directory as the executable file, and its file
22972 name must match the executable name suffixed by @code{.manifest}. For example,
22973 if the executable name is @code{sock_wto.exe}, then the manifest file name
22974 has to be @code{sock_wto.exe.manifest}. The manifest file must contain at
22975 least the following data:
22978 <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
22979 <assembly xmlns="urn:schemas-microsoft-com:asm.v1" manifestVersion="1.0">
22980 <compatibility xmlns="urn:schemas-microsoft-com:compatibility.v1">
22982 <!-- Windows Vista -->
22983 <supportedOS Id="@{e2011457-1546-43c5-a5fe-008deee3d3f0@}"/>
22985 <supportedOS Id="@{35138b9a-5d96-4fbd-8e2d-a2440225f93a@}"/>
22987 <supportedOS Id="@{4a2f28e3-53b9-4441-ba9c-d69d4a4a6e38@}"/>
22988 <!-- Windows 8.1 -->
22989 <supportedOS Id="@{1f676c76-80e1-4239-95bb-83d0f6d0da78@}"/>
22990 <!-- Windows 10 -->
22991 <supportedOS Id="@{8e0f7a12-bfb3-4fe8-b9a5-48fd50a15a9a@}"/>
22997 Without the manifest file, the socket timeout is going to be overcorrected on
22998 these Windows Server versions and the actual time is going to be 500
22999 milliseconds shorter than what was set with GNAT.Sockets.Set_Socket_Option.
23000 Note that on Microsoft Windows versions where correction is necessary, there
23001 is no way to set a socket timeout shorter than 500 ms. If a socket timeout
23002 shorter than 500 ms is needed on these Windows versions, a call to
23003 Check_Selector should be added before any socket read or write operations.
23005 @node Mixed-Language Programming on Windows,Windows Specific Add-Ons,Windows Socket Timeouts,Microsoft Windows Topics
23006 @anchor{gnat_ugn/platform_specific_information id13}@anchor{1cb}@anchor{gnat_ugn/platform_specific_information mixed-language-programming-on-windows}@anchor{1cc}
23007 @subsection Mixed-Language Programming on Windows
23010 Developing pure Ada applications on Windows is no different than on
23011 other GNAT-supported platforms. However, when developing or porting an
23012 application that contains a mix of Ada and C/C++, the choice of your
23013 Windows C/C++ development environment conditions your overall
23014 interoperability strategy.
23016 If you use @code{gcc} or Microsoft C to compile the non-Ada part of
23017 your application, there are no Windows-specific restrictions that
23018 affect the overall interoperability with your Ada code. If you do want
23019 to use the Microsoft tools for your C++ code, you have two choices:
23025 Encapsulate your C++ code in a DLL to be linked with your Ada
23026 application. In this case, use the Microsoft or whatever environment to
23027 build the DLL and use GNAT to build your executable
23028 (@ref{1cd,,Using DLLs with GNAT}).
23031 Or you can encapsulate your Ada code in a DLL to be linked with the
23032 other part of your application. In this case, use GNAT to build the DLL
23033 (@ref{1ce,,Building DLLs with GNAT Project files}) and use the Microsoft
23034 or whatever environment to build your executable.
23037 In addition to the description about C main in
23038 @ref{2c,,Mixed Language Programming} section, if the C main uses a
23039 stand-alone library it is required on x86-windows to
23040 setup the SEH context. For this the C main must looks like this:
23046 extern void adainit (void);
23047 extern void adafinal (void);
23048 extern void __gnat_initialize(void*);
23049 extern void call_to_ada (void);
23051 int main (int argc, char *argv[])
23055 /* Initialize the SEH context */
23056 __gnat_initialize (&SEH);
23060 /* Then call Ada services in the stand-alone library */
23069 Note that this is not needed on x86_64-windows where the Windows
23070 native SEH support is used.
23073 * Windows Calling Conventions::
23074 * Introduction to Dynamic Link Libraries (DLLs): Introduction to Dynamic Link Libraries DLLs.
23075 * Using DLLs with GNAT::
23076 * Building DLLs with GNAT Project files::
23077 * Building DLLs with GNAT::
23078 * Building DLLs with gnatdll::
23079 * Ada DLLs and Finalization::
23080 * Creating a Spec for Ada DLLs::
23081 * GNAT and Windows Resources::
23082 * Using GNAT DLLs from Microsoft Visual Studio Applications::
23083 * Debugging a DLL::
23084 * Setting Stack Size from gnatlink::
23085 * Setting Heap Size from gnatlink::
23089 @node Windows Calling Conventions,Introduction to Dynamic Link Libraries DLLs,,Mixed-Language Programming on Windows
23090 @anchor{gnat_ugn/platform_specific_information id14}@anchor{1cf}@anchor{gnat_ugn/platform_specific_information windows-calling-conventions}@anchor{1d0}
23091 @subsubsection Windows Calling Conventions
23098 This section pertain only to Win32. On Win64 there is a single native
23099 calling convention. All convention specifiers are ignored on this
23102 When a subprogram @code{F} (caller) calls a subprogram @code{G}
23103 (callee), there are several ways to push @code{G}’s parameters on the
23104 stack and there are several possible scenarios to clean up the stack
23105 upon @code{G}’s return. A calling convention is an agreed upon software
23106 protocol whereby the responsibilities between the caller (@code{F}) and
23107 the callee (@code{G}) are clearly defined. Several calling conventions
23108 are available for Windows:
23114 @code{C} (Microsoft defined)
23117 @code{Stdcall} (Microsoft defined)
23120 @code{Win32} (GNAT specific)
23123 @code{DLL} (GNAT specific)
23127 * C Calling Convention::
23128 * Stdcall Calling Convention::
23129 * Win32 Calling Convention::
23130 * DLL Calling Convention::
23134 @node C Calling Convention,Stdcall Calling Convention,,Windows Calling Conventions
23135 @anchor{gnat_ugn/platform_specific_information c-calling-convention}@anchor{1d1}@anchor{gnat_ugn/platform_specific_information id15}@anchor{1d2}
23136 @subsubsection @code{C} Calling Convention
23139 This is the default calling convention used when interfacing to C/C++
23140 routines compiled with either @code{gcc} or Microsoft Visual C++.
23142 In the @code{C} calling convention subprogram parameters are pushed on the
23143 stack by the caller from right to left. The caller itself is in charge of
23144 cleaning up the stack after the call. In addition, the name of a routine
23145 with @code{C} calling convention is mangled by adding a leading underscore.
23147 The name to use on the Ada side when importing (or exporting) a routine
23148 with @code{C} calling convention is the name of the routine. For
23149 instance the C function:
23154 int get_val (long);
23158 should be imported from Ada as follows:
23163 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
23164 pragma Import (C, Get_Val, External_Name => "get_val");
23168 Note that in this particular case the @code{External_Name} parameter could
23169 have been omitted since, when missing, this parameter is taken to be the
23170 name of the Ada entity in lower case. When the @code{Link_Name} parameter
23171 is missing, as in the above example, this parameter is set to be the
23172 @code{External_Name} with a leading underscore.
23174 When importing a variable defined in C, you should always use the @code{C}
23175 calling convention unless the object containing the variable is part of a
23176 DLL (in which case you should use the @code{Stdcall} calling
23177 convention, @ref{1d3,,Stdcall Calling Convention}).
23179 @node Stdcall Calling Convention,Win32 Calling Convention,C Calling Convention,Windows Calling Conventions
23180 @anchor{gnat_ugn/platform_specific_information id16}@anchor{1d4}@anchor{gnat_ugn/platform_specific_information stdcall-calling-convention}@anchor{1d3}
23181 @subsubsection @code{Stdcall} Calling Convention
23184 This convention, which was the calling convention used for Pascal
23185 programs, is used by Microsoft for all the routines in the Win32 API for
23186 efficiency reasons. It must be used to import any routine for which this
23187 convention was specified.
23189 In the @code{Stdcall} calling convention subprogram parameters are pushed
23190 on the stack by the caller from right to left. The callee (and not the
23191 caller) is in charge of cleaning the stack on routine exit. In addition,
23192 the name of a routine with @code{Stdcall} calling convention is mangled by
23193 adding a leading underscore (as for the @code{C} calling convention) and a
23194 trailing @code{@@@emph{nn}}, where @code{nn} is the overall size (in
23195 bytes) of the parameters passed to the routine.
23197 The name to use on the Ada side when importing a C routine with a
23198 @code{Stdcall} calling convention is the name of the C routine. The leading
23199 underscore and trailing @code{@@@emph{nn}} are added automatically by
23200 the compiler. For instance the Win32 function:
23205 APIENTRY int get_val (long);
23209 should be imported from Ada as follows:
23214 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
23215 pragma Import (Stdcall, Get_Val);
23216 -- On the x86 a long is 4 bytes, so the Link_Name is "_get_val@@4"
23220 As for the @code{C} calling convention, when the @code{External_Name}
23221 parameter is missing, it is taken to be the name of the Ada entity in lower
23222 case. If instead of writing the above import pragma you write:
23227 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
23228 pragma Import (Stdcall, Get_Val, External_Name => "retrieve_val");
23232 then the imported routine is @code{_retrieve_val@@4}. However, if instead
23233 of specifying the @code{External_Name} parameter you specify the
23234 @code{Link_Name} as in the following example:
23239 function Get_Val (V : Interfaces.C.long) return Interfaces.C.int;
23240 pragma Import (Stdcall, Get_Val, Link_Name => "retrieve_val");
23244 then the imported routine is @code{retrieve_val}, that is, there is no
23245 decoration at all. No leading underscore and no Stdcall suffix
23246 @code{@@@emph{nn}}.
23248 This is especially important as in some special cases a DLL’s entry
23249 point name lacks a trailing @code{@@@emph{nn}} while the exported
23250 name generated for a call has it.
23252 It is also possible to import variables defined in a DLL by using an
23253 import pragma for a variable. As an example, if a DLL contains a
23254 variable defined as:
23263 then, to access this variable from Ada you should write:
23268 My_Var : Interfaces.C.int;
23269 pragma Import (Stdcall, My_Var);
23273 Note that to ease building cross-platform bindings this convention
23274 will be handled as a @code{C} calling convention on non-Windows platforms.
23276 @node Win32 Calling Convention,DLL Calling Convention,Stdcall Calling Convention,Windows Calling Conventions
23277 @anchor{gnat_ugn/platform_specific_information id17}@anchor{1d5}@anchor{gnat_ugn/platform_specific_information win32-calling-convention}@anchor{1d6}
23278 @subsubsection @code{Win32} Calling Convention
23281 This convention, which is GNAT-specific is fully equivalent to the
23282 @code{Stdcall} calling convention described above.
23284 @node DLL Calling Convention,,Win32 Calling Convention,Windows Calling Conventions
23285 @anchor{gnat_ugn/platform_specific_information dll-calling-convention}@anchor{1d7}@anchor{gnat_ugn/platform_specific_information id18}@anchor{1d8}
23286 @subsubsection @code{DLL} Calling Convention
23289 This convention, which is GNAT-specific is fully equivalent to the
23290 @code{Stdcall} calling convention described above.
23292 @node Introduction to Dynamic Link Libraries DLLs,Using DLLs with GNAT,Windows Calling Conventions,Mixed-Language Programming on Windows
23293 @anchor{gnat_ugn/platform_specific_information id19}@anchor{1d9}@anchor{gnat_ugn/platform_specific_information introduction-to-dynamic-link-libraries-dlls}@anchor{1da}
23294 @subsubsection Introduction to Dynamic Link Libraries (DLLs)
23299 A Dynamically Linked Library (DLL) is a library that can be shared by
23300 several applications running under Windows. A DLL can contain any number of
23301 routines and variables.
23303 One advantage of DLLs is that you can change and enhance them without
23304 forcing all the applications that depend on them to be relinked or
23305 recompiled. However, you should be aware than all calls to DLL routines are
23306 slower since, as you will understand below, such calls are indirect.
23308 To illustrate the remainder of this section, suppose that an application
23309 wants to use the services of a DLL @code{API.dll}. To use the services
23310 provided by @code{API.dll} you must statically link against the DLL or
23311 an import library which contains a jump table with an entry for each
23312 routine and variable exported by the DLL. In the Microsoft world this
23313 import library is called @code{API.lib}. When using GNAT this import
23314 library is called either @code{libAPI.dll.a}, @code{libapi.dll.a},
23315 @code{libAPI.a} or @code{libapi.a} (names are case insensitive).
23317 After you have linked your application with the DLL or the import library
23318 and you run your application, here is what happens:
23324 Your application is loaded into memory.
23327 The DLL @code{API.dll} is mapped into the address space of your
23328 application. This means that:
23334 The DLL will use the stack of the calling thread.
23337 The DLL will use the virtual address space of the calling process.
23340 The DLL will allocate memory from the virtual address space of the calling
23344 Handles (pointers) can be safely exchanged between routines in the DLL
23345 routines and routines in the application using the DLL.
23349 The entries in the jump table (from the import library @code{libAPI.dll.a}
23350 or @code{API.lib} or automatically created when linking against a DLL)
23351 which is part of your application are initialized with the addresses
23352 of the routines and variables in @code{API.dll}.
23355 If present in @code{API.dll}, routines @code{DllMain} or
23356 @code{DllMainCRTStartup} are invoked. These routines typically contain
23357 the initialization code needed for the well-being of the routines and
23358 variables exported by the DLL.
23361 There is an additional point which is worth mentioning. In the Windows
23362 world there are two kind of DLLs: relocatable and non-relocatable
23363 DLLs. Non-relocatable DLLs can only be loaded at a very specific address
23364 in the target application address space. If the addresses of two
23365 non-relocatable DLLs overlap and these happen to be used by the same
23366 application, a conflict will occur and the application will run
23367 incorrectly. Hence, when possible, it is always preferable to use and
23368 build relocatable DLLs. Both relocatable and non-relocatable DLLs are
23369 supported by GNAT. Note that the @code{-s} linker option (see GNU Linker
23370 User’s Guide) removes the debugging symbols from the DLL but the DLL can
23371 still be relocated.
23373 As a side note, an interesting difference between Microsoft DLLs and
23374 Unix shared libraries, is the fact that on most Unix systems all public
23375 routines are exported by default in a Unix shared library, while under
23376 Windows it is possible (but not required) to list exported routines in
23377 a definition file (see @ref{1db,,The Definition File}).
23379 @node Using DLLs with GNAT,Building DLLs with GNAT Project files,Introduction to Dynamic Link Libraries DLLs,Mixed-Language Programming on Windows
23380 @anchor{gnat_ugn/platform_specific_information id20}@anchor{1dc}@anchor{gnat_ugn/platform_specific_information using-dlls-with-gnat}@anchor{1cd}
23381 @subsubsection Using DLLs with GNAT
23384 To use the services of a DLL, say @code{API.dll}, in your Ada application
23391 The Ada spec for the routines and/or variables you want to access in
23392 @code{API.dll}. If not available this Ada spec must be built from the C/C++
23393 header files provided with the DLL.
23396 The import library (@code{libAPI.dll.a} or @code{API.lib}). As previously
23397 mentioned an import library is a statically linked library containing the
23398 import table which will be filled at load time to point to the actual
23399 @code{API.dll} routines. Sometimes you don’t have an import library for the
23400 DLL you want to use. The following sections will explain how to build
23401 one. Note that this is optional.
23404 The actual DLL, @code{API.dll}.
23407 Once you have all the above, to compile an Ada application that uses the
23408 services of @code{API.dll} and whose main subprogram is @code{My_Ada_App},
23409 you simply issue the command
23414 $ gnatmake my_ada_app -largs -lAPI
23418 The argument @code{-largs -lAPI} at the end of the @code{gnatmake} command
23419 tells the GNAT linker to look for an import library. The linker will
23420 look for a library name in this specific order:
23426 @code{libAPI.dll.a}
23444 The first three are the GNU style import libraries. The third is the
23445 Microsoft style import libraries. The last two are the actual DLL names.
23447 Note that if the Ada package spec for @code{API.dll} contains the
23453 pragma Linker_Options ("-lAPI");
23457 you do not have to add @code{-largs -lAPI} at the end of the
23458 @code{gnatmake} command.
23460 If any one of the items above is missing you will have to create it
23461 yourself. The following sections explain how to do so using as an
23462 example a fictitious DLL called @code{API.dll}.
23465 * Creating an Ada Spec for the DLL Services::
23466 * Creating an Import Library::
23470 @node Creating an Ada Spec for the DLL Services,Creating an Import Library,,Using DLLs with GNAT
23471 @anchor{gnat_ugn/platform_specific_information creating-an-ada-spec-for-the-dll-services}@anchor{1dd}@anchor{gnat_ugn/platform_specific_information id21}@anchor{1de}
23472 @subsubsection Creating an Ada Spec for the DLL Services
23475 A DLL typically comes with a C/C++ header file which provides the
23476 definitions of the routines and variables exported by the DLL. The Ada
23477 equivalent of this header file is a package spec that contains definitions
23478 for the imported entities. If the DLL you intend to use does not come with
23479 an Ada spec you have to generate one such spec yourself. For example if
23480 the header file of @code{API.dll} is a file @code{api.h} containing the
23481 following two definitions:
23491 then the equivalent Ada spec could be:
23496 with Interfaces.C.Strings;
23501 function Get (Str : C.Strings.Chars_Ptr) return C.int;
23504 pragma Import (C, Get);
23505 pragma Import (DLL, Some_Var);
23510 @node Creating an Import Library,,Creating an Ada Spec for the DLL Services,Using DLLs with GNAT
23511 @anchor{gnat_ugn/platform_specific_information creating-an-import-library}@anchor{1df}@anchor{gnat_ugn/platform_specific_information id22}@anchor{1e0}
23512 @subsubsection Creating an Import Library
23515 @geindex Import library
23517 If a Microsoft-style import library @code{API.lib} or a GNAT-style
23518 import library @code{libAPI.dll.a} or @code{libAPI.a} is available
23519 with @code{API.dll} you can skip this section. You can also skip this
23520 section if @code{API.dll} or @code{libAPI.dll} is built with GNU tools
23521 as in this case it is possible to link directly against the
23522 DLL. Otherwise read on.
23524 @geindex Definition file
23525 @anchor{gnat_ugn/platform_specific_information the-definition-file}@anchor{1db}
23526 @subsubheading The Definition File
23529 As previously mentioned, and unlike Unix systems, the list of symbols
23530 that are exported from a DLL must be provided explicitly in Windows.
23531 The main goal of a definition file is precisely that: list the symbols
23532 exported by a DLL. A definition file (usually a file with a @code{.def}
23533 suffix) has the following structure:
23538 [LIBRARY `@w{`}name`@w{`}]
23539 [DESCRIPTION `@w{`}string`@w{`}]
23541 `@w{`}symbol1`@w{`}
23542 `@w{`}symbol2`@w{`}
23550 @item @emph{LIBRARY name}
23552 This section, which is optional, gives the name of the DLL.
23554 @item @emph{DESCRIPTION string}
23556 This section, which is optional, gives a description string that will be
23557 embedded in the import library.
23559 @item @emph{EXPORTS}
23561 This section gives the list of exported symbols (procedures, functions or
23562 variables). For instance in the case of @code{API.dll} the @code{EXPORTS}
23563 section of @code{API.def} looks like:
23572 Note that you must specify the correct suffix (@code{@@@emph{nn}})
23573 (see @ref{1d0,,Windows Calling Conventions}) for a Stdcall
23574 calling convention function in the exported symbols list.
23576 There can actually be other sections in a definition file, but these
23577 sections are not relevant to the discussion at hand.
23578 @anchor{gnat_ugn/platform_specific_information create-def-file-automatically}@anchor{1e1}
23579 @subsubheading Creating a Definition File Automatically
23582 You can automatically create the definition file @code{API.def}
23583 (see @ref{1db,,The Definition File}) from a DLL.
23584 For that use the @code{dlltool} program as follows:
23589 $ dlltool API.dll -z API.def --export-all-symbols
23592 Note that if some routines in the DLL have the @code{Stdcall} convention
23593 (@ref{1d0,,Windows Calling Conventions}) with stripped @code{@@@emph{nn}}
23594 suffix then you’ll have to edit @code{api.def} to add it, and specify
23595 @code{-k} to @code{gnatdll} when creating the import library.
23597 Here are some hints to find the right @code{@@@emph{nn}} suffix.
23603 If you have the Microsoft import library (.lib), it is possible to get
23604 the right symbols by using Microsoft @code{dumpbin} tool (see the
23605 corresponding Microsoft documentation for further details).
23608 $ dumpbin /exports api.lib
23612 If you have a message about a missing symbol at link time the compiler
23613 tells you what symbol is expected. You just have to go back to the
23614 definition file and add the right suffix.
23617 @anchor{gnat_ugn/platform_specific_information gnat-style-import-library}@anchor{1e2}
23618 @subsubheading GNAT-Style Import Library
23621 To create a static import library from @code{API.dll} with the GNAT tools
23622 you should create the .def file, then use @code{gnatdll} tool
23623 (see @ref{1e3,,Using gnatdll}) as follows:
23628 $ gnatdll -e API.def -d API.dll
23631 @code{gnatdll} takes as input a definition file @code{API.def} and the
23632 name of the DLL containing the services listed in the definition file
23633 @code{API.dll}. The name of the static import library generated is
23634 computed from the name of the definition file as follows: if the
23635 definition file name is @code{xyz.def}, the import library name will
23636 be @code{libxyz.a}. Note that in the previous example option
23637 @code{-e} could have been removed because the name of the definition
23638 file (before the @code{.def} suffix) is the same as the name of the
23639 DLL (@ref{1e3,,Using gnatdll} for more information about @code{gnatdll}).
23641 @anchor{gnat_ugn/platform_specific_information msvs-style-import-library}@anchor{1e4}
23642 @subsubheading Microsoft-Style Import Library
23645 A Microsoft import library is needed only if you plan to make an
23646 Ada DLL available to applications developed with Microsoft
23647 tools (@ref{1cc,,Mixed-Language Programming on Windows}).
23649 To create a Microsoft-style import library for @code{API.dll} you
23650 should create the .def file, then build the actual import library using
23651 Microsoft’s @code{lib} utility:
23656 $ lib -machine:IX86 -def:API.def -out:API.lib
23659 If you use the above command the definition file @code{API.def} must
23660 contain a line giving the name of the DLL:
23666 See the Microsoft documentation for further details about the usage of
23670 @node Building DLLs with GNAT Project files,Building DLLs with GNAT,Using DLLs with GNAT,Mixed-Language Programming on Windows
23671 @anchor{gnat_ugn/platform_specific_information building-dlls-with-gnat-project-files}@anchor{1ce}@anchor{gnat_ugn/platform_specific_information id23}@anchor{1e5}
23672 @subsubsection Building DLLs with GNAT Project files
23678 There is nothing specific to Windows in the build process.
23679 See the @emph{Library Projects} section in the @emph{GNAT Project Manager}
23680 chapter of the @emph{GPRbuild User’s Guide}.
23682 Due to a system limitation, it is not possible under Windows to create threads
23683 when inside the @code{DllMain} routine which is used for auto-initialization
23684 of shared libraries, so it is not possible to have library level tasks in SALs.
23686 @node Building DLLs with GNAT,Building DLLs with gnatdll,Building DLLs with GNAT Project files,Mixed-Language Programming on Windows
23687 @anchor{gnat_ugn/platform_specific_information building-dlls-with-gnat}@anchor{1e6}@anchor{gnat_ugn/platform_specific_information id24}@anchor{1e7}
23688 @subsubsection Building DLLs with GNAT
23694 This section explain how to build DLLs using the GNAT built-in DLL
23695 support. With the following procedure it is straight forward to build
23696 and use DLLs with GNAT.
23702 Building object files.
23703 The first step is to build all objects files that are to be included
23704 into the DLL. This is done by using the standard @code{gnatmake} tool.
23708 To build the DLL you must use the @code{gcc} @code{-shared} and
23709 @code{-shared-libgcc} options. It is quite simple to use this method:
23712 $ gcc -shared -shared-libgcc -o api.dll obj1.o obj2.o ...
23715 It is important to note that in this case all symbols found in the
23716 object files are automatically exported. It is possible to restrict
23717 the set of symbols to export by passing to @code{gcc} a definition
23718 file (see @ref{1db,,The Definition File}).
23722 $ gcc -shared -shared-libgcc -o api.dll api.def obj1.o obj2.o ...
23725 If you use a definition file you must export the elaboration procedures
23726 for every package that required one. Elaboration procedures are named
23727 using the package name followed by “_E”.
23730 Preparing DLL to be used.
23731 For the DLL to be used by client programs the bodies must be hidden
23732 from it and the .ali set with read-only attribute. This is very important
23733 otherwise GNAT will recompile all packages and will not actually use
23734 the code in the DLL. For example:
23738 $ copy *.ads *.ali api.dll apilib
23739 $ attrib +R apilib\\*.ali
23743 At this point it is possible to use the DLL by directly linking
23744 against it. Note that you must use the GNAT shared runtime when using
23745 GNAT shared libraries. This is achieved by using the @code{-shared} binder
23751 $ gnatmake main -Iapilib -bargs -shared -largs -Lapilib -lAPI
23755 @node Building DLLs with gnatdll,Ada DLLs and Finalization,Building DLLs with GNAT,Mixed-Language Programming on Windows
23756 @anchor{gnat_ugn/platform_specific_information building-dlls-with-gnatdll}@anchor{1e8}@anchor{gnat_ugn/platform_specific_information id25}@anchor{1e9}
23757 @subsubsection Building DLLs with gnatdll
23763 Note that it is preferred to use GNAT Project files
23764 (@ref{1ce,,Building DLLs with GNAT Project files}) or the built-in GNAT
23765 DLL support (@ref{1e6,,Building DLLs with GNAT}) or to build DLLs.
23767 This section explains how to build DLLs containing Ada code using
23768 @code{gnatdll}. These DLLs will be referred to as Ada DLLs in the
23769 remainder of this section.
23771 The steps required to build an Ada DLL that is to be used by Ada as well as
23772 non-Ada applications are as follows:
23778 You need to mark each Ada entity exported by the DLL with a @code{C} or
23779 @code{Stdcall} calling convention to avoid any Ada name mangling for the
23780 entities exported by the DLL
23781 (see @ref{1ea,,Exporting Ada Entities}). You can
23782 skip this step if you plan to use the Ada DLL only from Ada applications.
23785 Your Ada code must export an initialization routine which calls the routine
23786 @code{adainit} generated by @code{gnatbind} to perform the elaboration of
23787 the Ada code in the DLL (@ref{1eb,,Ada DLLs and Elaboration}). The initialization
23788 routine exported by the Ada DLL must be invoked by the clients of the DLL
23789 to initialize the DLL.
23792 When useful, the DLL should also export a finalization routine which calls
23793 routine @code{adafinal} generated by @code{gnatbind} to perform the
23794 finalization of the Ada code in the DLL (@ref{1ec,,Ada DLLs and Finalization}).
23795 The finalization routine exported by the Ada DLL must be invoked by the
23796 clients of the DLL when the DLL services are no further needed.
23799 You must provide a spec for the services exported by the Ada DLL in each
23800 of the programming languages to which you plan to make the DLL available.
23803 You must provide a definition file listing the exported entities
23804 (@ref{1db,,The Definition File}).
23807 Finally you must use @code{gnatdll} to produce the DLL and the import
23808 library (@ref{1e3,,Using gnatdll}).
23811 Note that a relocatable DLL stripped using the @code{strip}
23812 binutils tool will not be relocatable anymore. To build a DLL without
23813 debug information pass @code{-largs -s} to @code{gnatdll}. This
23814 restriction does not apply to a DLL built using a Library Project.
23815 See the @emph{Library Projects} section in the @emph{GNAT Project Manager}
23816 chapter of the @emph{GPRbuild User’s Guide}.
23818 @c Limitations_When_Using_Ada_DLLs_from Ada:
23821 * Limitations When Using Ada DLLs from Ada::
23822 * Exporting Ada Entities::
23823 * Ada DLLs and Elaboration::
23827 @node Limitations When Using Ada DLLs from Ada,Exporting Ada Entities,,Building DLLs with gnatdll
23828 @anchor{gnat_ugn/platform_specific_information limitations-when-using-ada-dlls-from-ada}@anchor{1ed}
23829 @subsubsection Limitations When Using Ada DLLs from Ada
23832 When using Ada DLLs from Ada applications there is a limitation users
23833 should be aware of. Because on Windows the GNAT run-time is not in a DLL of
23834 its own, each Ada DLL includes a part of the GNAT run-time. Specifically,
23835 each Ada DLL includes the services of the GNAT run-time that are necessary
23836 to the Ada code inside the DLL. As a result, when an Ada program uses an
23837 Ada DLL there are two independent GNAT run-times: one in the Ada DLL and
23838 one in the main program.
23840 It is therefore not possible to exchange GNAT run-time objects between the
23841 Ada DLL and the main Ada program. Example of GNAT run-time objects are file
23842 handles (e.g., @code{Text_IO.File_Type}), tasks types, protected objects
23845 It is completely safe to exchange plain elementary, array or record types,
23846 Windows object handles, etc.
23848 @node Exporting Ada Entities,Ada DLLs and Elaboration,Limitations When Using Ada DLLs from Ada,Building DLLs with gnatdll
23849 @anchor{gnat_ugn/platform_specific_information exporting-ada-entities}@anchor{1ea}@anchor{gnat_ugn/platform_specific_information id26}@anchor{1ee}
23850 @subsubsection Exporting Ada Entities
23853 @geindex Export table
23855 Building a DLL is a way to encapsulate a set of services usable from any
23856 application. As a result, the Ada entities exported by a DLL should be
23857 exported with the @code{C} or @code{Stdcall} calling conventions to avoid
23858 any Ada name mangling. As an example here is an Ada package
23859 @code{API}, spec and body, exporting two procedures, a function, and a
23865 with Interfaces.C; use Interfaces;
23867 Count : C.int := 0;
23868 function Factorial (Val : C.int) return C.int;
23870 procedure Initialize_API;
23871 procedure Finalize_API;
23872 -- Initialization & Finalization routines. More in the next section.
23874 pragma Export (C, Initialize_API);
23875 pragma Export (C, Finalize_API);
23876 pragma Export (C, Count);
23877 pragma Export (C, Factorial);
23882 package body API is
23883 function Factorial (Val : C.int) return C.int is
23886 Count := Count + 1;
23887 for K in 1 .. Val loop
23893 procedure Initialize_API is
23895 pragma Import (C, Adainit);
23898 end Initialize_API;
23900 procedure Finalize_API is
23901 procedure Adafinal;
23902 pragma Import (C, Adafinal);
23910 If the Ada DLL you are building will only be used by Ada applications
23911 you do not have to export Ada entities with a @code{C} or @code{Stdcall}
23912 convention. As an example, the previous package could be written as
23919 Count : Integer := 0;
23920 function Factorial (Val : Integer) return Integer;
23922 procedure Initialize_API;
23923 procedure Finalize_API;
23924 -- Initialization and Finalization routines.
23929 package body API is
23930 function Factorial (Val : Integer) return Integer is
23931 Fact : Integer := 1;
23933 Count := Count + 1;
23934 for K in 1 .. Val loop
23941 -- The remainder of this package body is unchanged.
23946 Note that if you do not export the Ada entities with a @code{C} or
23947 @code{Stdcall} convention you will have to provide the mangled Ada names
23948 in the definition file of the Ada DLL
23949 (@ref{1ef,,Creating the Definition File}).
23951 @node Ada DLLs and Elaboration,,Exporting Ada Entities,Building DLLs with gnatdll
23952 @anchor{gnat_ugn/platform_specific_information ada-dlls-and-elaboration}@anchor{1eb}@anchor{gnat_ugn/platform_specific_information id27}@anchor{1f0}
23953 @subsubsection Ada DLLs and Elaboration
23956 @geindex DLLs and elaboration
23958 The DLL that you are building contains your Ada code as well as all the
23959 routines in the Ada library that are needed by it. The first thing a
23960 user of your DLL must do is elaborate the Ada code
23961 (@ref{f,,Elaboration Order Handling in GNAT}).
23963 To achieve this you must export an initialization routine
23964 (@code{Initialize_API} in the previous example), which must be invoked
23965 before using any of the DLL services. This elaboration routine must call
23966 the Ada elaboration routine @code{adainit} generated by the GNAT binder
23967 (@ref{a0,,Binding with Non-Ada Main Programs}). See the body of
23968 @code{Initialize_Api} for an example. Note that the GNAT binder is
23969 automatically invoked during the DLL build process by the @code{gnatdll}
23970 tool (@ref{1e3,,Using gnatdll}).
23972 When a DLL is loaded, Windows systematically invokes a routine called
23973 @code{DllMain}. It would therefore be possible to call @code{adainit}
23974 directly from @code{DllMain} without having to provide an explicit
23975 initialization routine. Unfortunately, it is not possible to call
23976 @code{adainit} from the @code{DllMain} if your program has library level
23977 tasks because access to the @code{DllMain} entry point is serialized by
23978 the system (that is, only a single thread can execute ‘through’ it at a
23979 time), which means that the GNAT run-time will deadlock waiting for the
23980 newly created task to complete its initialization.
23982 @node Ada DLLs and Finalization,Creating a Spec for Ada DLLs,Building DLLs with gnatdll,Mixed-Language Programming on Windows
23983 @anchor{gnat_ugn/platform_specific_information ada-dlls-and-finalization}@anchor{1ec}@anchor{gnat_ugn/platform_specific_information id28}@anchor{1f1}
23984 @subsubsection Ada DLLs and Finalization
23987 @geindex DLLs and finalization
23989 When the services of an Ada DLL are no longer needed, the client code should
23990 invoke the DLL finalization routine, if available. The DLL finalization
23991 routine is in charge of releasing all resources acquired by the DLL. In the
23992 case of the Ada code contained in the DLL, this is achieved by calling
23993 routine @code{adafinal} generated by the GNAT binder
23994 (@ref{a0,,Binding with Non-Ada Main Programs}).
23995 See the body of @code{Finalize_Api} for an
23996 example. As already pointed out the GNAT binder is automatically invoked
23997 during the DLL build process by the @code{gnatdll} tool
23998 (@ref{1e3,,Using gnatdll}).
24000 @node Creating a Spec for Ada DLLs,GNAT and Windows Resources,Ada DLLs and Finalization,Mixed-Language Programming on Windows
24001 @anchor{gnat_ugn/platform_specific_information creating-a-spec-for-ada-dlls}@anchor{1f2}@anchor{gnat_ugn/platform_specific_information id29}@anchor{1f3}
24002 @subsubsection Creating a Spec for Ada DLLs
24005 To use the services exported by the Ada DLL from another programming
24006 language (e.g., C), you have to translate the specs of the exported Ada
24007 entities in that language. For instance in the case of @code{API.dll},
24008 the corresponding C header file could look like:
24013 extern int *_imp__count;
24014 #define count (*_imp__count)
24015 int factorial (int);
24019 It is important to understand that when building an Ada DLL to be used by
24020 other Ada applications, you need two different specs for the packages
24021 contained in the DLL: one for building the DLL and the other for using
24022 the DLL. This is because the @code{DLL} calling convention is needed to
24023 use a variable defined in a DLL, but when building the DLL, the variable
24024 must have either the @code{Ada} or @code{C} calling convention. As an
24025 example consider a DLL comprising the following package @code{API}:
24031 Count : Integer := 0;
24033 -- Remainder of the package omitted.
24038 After producing a DLL containing package @code{API}, the spec that
24039 must be used to import @code{API.Count} from Ada code outside of the
24047 pragma Import (DLL, Count);
24053 * Creating the Definition File::
24058 @node Creating the Definition File,Using gnatdll,,Creating a Spec for Ada DLLs
24059 @anchor{gnat_ugn/platform_specific_information creating-the-definition-file}@anchor{1ef}@anchor{gnat_ugn/platform_specific_information id30}@anchor{1f4}
24060 @subsubsection Creating the Definition File
24063 The definition file is the last file needed to build the DLL. It lists
24064 the exported symbols. As an example, the definition file for a DLL
24065 containing only package @code{API} (where all the entities are exported
24066 with a @code{C} calling convention) is:
24079 If the @code{C} calling convention is missing from package @code{API},
24080 then the definition file contains the mangled Ada names of the above
24081 entities, which in this case are:
24090 api__initialize_api
24094 @node Using gnatdll,,Creating the Definition File,Creating a Spec for Ada DLLs
24095 @anchor{gnat_ugn/platform_specific_information id31}@anchor{1f5}@anchor{gnat_ugn/platform_specific_information using-gnatdll}@anchor{1e3}
24096 @subsubsection Using @code{gnatdll}
24101 @code{gnatdll} is a tool to automate the DLL build process once all the Ada
24102 and non-Ada sources that make up your DLL have been compiled.
24103 @code{gnatdll} is actually in charge of two distinct tasks: build the
24104 static import library for the DLL and the actual DLL. The form of the
24105 @code{gnatdll} command is
24110 $ gnatdll [ switches ] list-of-files [ -largs opts ]
24114 where @code{list-of-files} is a list of ALI and object files. The object
24115 file list must be the exact list of objects corresponding to the non-Ada
24116 sources whose services are to be included in the DLL. The ALI file list
24117 must be the exact list of ALI files for the corresponding Ada sources
24118 whose services are to be included in the DLL. If @code{list-of-files} is
24119 missing, only the static import library is generated.
24121 You may specify any of the following switches to @code{gnatdll}:
24125 @geindex -a (gnatdll)
24131 @item @code{-a[@emph{address}]}
24133 Build a non-relocatable DLL at @code{address}. If @code{address} is not
24134 specified the default address @code{0x11000000} will be used. By default,
24135 when this switch is missing, @code{gnatdll} builds relocatable DLL. We
24136 advise the reader to build relocatable DLL.
24138 @geindex -b (gnatdll)
24140 @item @code{-b @emph{address}}
24142 Set the relocatable DLL base address. By default the address is
24145 @geindex -bargs (gnatdll)
24147 @item @code{-bargs @emph{opts}}
24149 Binder options. Pass @code{opts} to the binder.
24151 @geindex -d (gnatdll)
24153 @item @code{-d @emph{dllfile}}
24155 @code{dllfile} is the name of the DLL. This switch must be present for
24156 @code{gnatdll} to do anything. The name of the generated import library is
24157 obtained algorithmically from @code{dllfile} as shown in the following
24158 example: if @code{dllfile} is @code{xyz.dll}, the import library name is
24159 @code{libxyz.dll.a}. The name of the definition file to use (if not specified
24160 by option @code{-e}) is obtained algorithmically from @code{dllfile}
24161 as shown in the following example:
24162 if @code{dllfile} is @code{xyz.dll}, the definition
24163 file used is @code{xyz.def}.
24165 @geindex -e (gnatdll)
24167 @item @code{-e @emph{deffile}}
24169 @code{deffile} is the name of the definition file.
24171 @geindex -g (gnatdll)
24175 Generate debugging information. This information is stored in the object
24176 file and copied from there to the final DLL file by the linker,
24177 where it can be read by the debugger. You must use the
24178 @code{-g} switch if you plan on using the debugger or the symbolic
24181 @geindex -h (gnatdll)
24185 Help mode. Displays @code{gnatdll} switch usage information.
24187 @geindex -I (gnatdll)
24189 @item @code{-I@emph{dir}}
24191 Direct @code{gnatdll} to search the @code{dir} directory for source and
24192 object files needed to build the DLL.
24193 (@ref{73,,Search Paths and the Run-Time Library (RTL)}).
24195 @geindex -k (gnatdll)
24199 Removes the @code{@@@emph{nn}} suffix from the import library’s exported
24200 names, but keeps them for the link names. You must specify this
24201 option if you want to use a @code{Stdcall} function in a DLL for which
24202 the @code{@@@emph{nn}} suffix has been removed. This is the case for most
24203 of the Windows NT DLL for example. This option has no effect when
24204 @code{-n} option is specified.
24206 @geindex -l (gnatdll)
24208 @item @code{-l @emph{file}}
24210 The list of ALI and object files used to build the DLL are listed in
24211 @code{file}, instead of being given in the command line. Each line in
24212 @code{file} contains the name of an ALI or object file.
24214 @geindex -n (gnatdll)
24218 No Import. Do not create the import library.
24220 @geindex -q (gnatdll)
24224 Quiet mode. Do not display unnecessary messages.
24226 @geindex -v (gnatdll)
24230 Verbose mode. Display extra information.
24232 @geindex -largs (gnatdll)
24234 @item @code{-largs @emph{opts}}
24236 Linker options. Pass @code{opts} to the linker.
24239 @subsubheading @code{gnatdll} Example
24242 As an example the command to build a relocatable DLL from @code{api.adb}
24243 once @code{api.adb} has been compiled and @code{api.def} created is
24248 $ gnatdll -d api.dll api.ali
24252 The above command creates two files: @code{libapi.dll.a} (the import
24253 library) and @code{api.dll} (the actual DLL). If you want to create
24254 only the DLL, just type:
24259 $ gnatdll -d api.dll -n api.ali
24263 Alternatively if you want to create just the import library, type:
24268 $ gnatdll -d api.dll
24272 @subsubheading @code{gnatdll} behind the Scenes
24275 This section details the steps involved in creating a DLL. @code{gnatdll}
24276 does these steps for you. Unless you are interested in understanding what
24277 goes on behind the scenes, you should skip this section.
24279 We use the previous example of a DLL containing the Ada package @code{API},
24280 to illustrate the steps necessary to build a DLL. The starting point is a
24281 set of objects that will make up the DLL and the corresponding ALI
24282 files. In the case of this example this means that @code{api.o} and
24283 @code{api.ali} are available. To build a relocatable DLL, @code{gnatdll} does
24290 @code{gnatdll} builds the base file (@code{api.base}). A base file gives
24291 the information necessary to generate relocation information for the
24296 $ gnatlink api -o api.jnk -mdll -Wl,--base-file,api.base
24299 In addition to the base file, the @code{gnatlink} command generates an
24300 output file @code{api.jnk} which can be discarded. The @code{-mdll} switch
24301 asks @code{gnatlink} to generate the routines @code{DllMain} and
24302 @code{DllMainCRTStartup} that are called by the Windows loader when the DLL
24303 is loaded into memory.
24306 @code{gnatdll} uses @code{dlltool} (see @ref{1f6,,Using dlltool}) to build the
24307 export table (@code{api.exp}). The export table contains the relocation
24308 information in a form which can be used during the final link to ensure
24309 that the Windows loader is able to place the DLL anywhere in memory.
24312 $ dlltool --dllname api.dll --def api.def --base-file api.base \\
24313 --output-exp api.exp
24317 @code{gnatdll} builds the base file using the new export table. Note that
24318 @code{gnatbind} must be called once again since the binder generated file
24319 has been deleted during the previous call to @code{gnatlink}.
24323 $ gnatlink api -o api.jnk api.exp -mdll
24324 -Wl,--base-file,api.base
24328 @code{gnatdll} builds the new export table using the new base file and
24329 generates the DLL import library @code{libAPI.dll.a}.
24332 $ dlltool --dllname api.dll --def api.def --base-file api.base \\
24333 --output-exp api.exp --output-lib libAPI.a
24337 Finally @code{gnatdll} builds the relocatable DLL using the final export
24342 $ gnatlink api api.exp -o api.dll -mdll
24345 @anchor{gnat_ugn/platform_specific_information using-dlltool}@anchor{1f6}
24346 @subsubheading Using @code{dlltool}
24349 @code{dlltool} is the low-level tool used by @code{gnatdll} to build
24350 DLLs and static import libraries. This section summarizes the most
24351 common @code{dlltool} switches. The form of the @code{dlltool} command
24357 $ dlltool [`switches`]
24361 @code{dlltool} switches include:
24363 @geindex --base-file (dlltool)
24368 @item @code{--base-file @emph{basefile}}
24370 Read the base file @code{basefile} generated by the linker. This switch
24371 is used to create a relocatable DLL.
24374 @geindex --def (dlltool)
24379 @item @code{--def @emph{deffile}}
24381 Read the definition file.
24384 @geindex --dllname (dlltool)
24389 @item @code{--dllname @emph{name}}
24391 Gives the name of the DLL. This switch is used to embed the name of the
24392 DLL in the static import library generated by @code{dlltool} with switch
24393 @code{--output-lib}.
24396 @geindex -k (dlltool)
24403 Kill @code{@@@emph{nn}} from exported names
24404 (@ref{1d0,,Windows Calling Conventions}
24405 for a discussion about @code{Stdcall}-style symbols.
24408 @geindex --help (dlltool)
24413 @item @code{--help}
24415 Prints the @code{dlltool} switches with a concise description.
24418 @geindex --output-exp (dlltool)
24423 @item @code{--output-exp @emph{exportfile}}
24425 Generate an export file @code{exportfile}. The export file contains the
24426 export table (list of symbols in the DLL) and is used to create the DLL.
24429 @geindex --output-lib (dlltool)
24434 @item @code{--output-lib @emph{libfile}}
24436 Generate a static import library @code{libfile}.
24439 @geindex -v (dlltool)
24449 @geindex --as (dlltool)
24454 @item @code{--as @emph{assembler-name}}
24456 Use @code{assembler-name} as the assembler. The default is @code{as}.
24459 @node GNAT and Windows Resources,Using GNAT DLLs from Microsoft Visual Studio Applications,Creating a Spec for Ada DLLs,Mixed-Language Programming on Windows
24460 @anchor{gnat_ugn/platform_specific_information gnat-and-windows-resources}@anchor{1f7}@anchor{gnat_ugn/platform_specific_information id32}@anchor{1f8}
24461 @subsubsection GNAT and Windows Resources
24467 Resources are an easy way to add Windows specific objects to your
24468 application. The objects that can be added as resources include:
24498 version information
24501 For example, a version information resource can be defined as follow and
24502 embedded into an executable or DLL:
24504 A version information resource can be used to embed information into an
24505 executable or a DLL. These information can be viewed using the file properties
24506 from the Windows Explorer. Here is an example of a version information
24513 FILEVERSION 1,0,0,0
24514 PRODUCTVERSION 1,0,0,0
24516 BLOCK "StringFileInfo"
24520 VALUE "CompanyName", "My Company Name"
24521 VALUE "FileDescription", "My application"
24522 VALUE "FileVersion", "1.0"
24523 VALUE "InternalName", "my_app"
24524 VALUE "LegalCopyright", "My Name"
24525 VALUE "OriginalFilename", "my_app.exe"
24526 VALUE "ProductName", "My App"
24527 VALUE "ProductVersion", "1.0"
24531 BLOCK "VarFileInfo"
24533 VALUE "Translation", 0x809, 1252
24539 The value @code{0809} (langID) is for the U.K English language and
24540 @code{04E4} (charsetID), which is equal to @code{1252} decimal, for
24543 This section explains how to build, compile and use resources. Note that this
24544 section does not cover all resource objects, for a complete description see
24545 the corresponding Microsoft documentation.
24548 * Building Resources::
24549 * Compiling Resources::
24550 * Using Resources::
24554 @node Building Resources,Compiling Resources,,GNAT and Windows Resources
24555 @anchor{gnat_ugn/platform_specific_information building-resources}@anchor{1f9}@anchor{gnat_ugn/platform_specific_information id33}@anchor{1fa}
24556 @subsubsection Building Resources
24562 A resource file is an ASCII file. By convention resource files have an
24563 @code{.rc} extension.
24564 The easiest way to build a resource file is to use Microsoft tools
24565 such as @code{imagedit.exe} to build bitmaps, icons and cursors and
24566 @code{dlgedit.exe} to build dialogs.
24567 It is always possible to build an @code{.rc} file yourself by writing a
24570 It is not our objective to explain how to write a resource file. A
24571 complete description of the resource script language can be found in the
24572 Microsoft documentation.
24574 @node Compiling Resources,Using Resources,Building Resources,GNAT and Windows Resources
24575 @anchor{gnat_ugn/platform_specific_information compiling-resources}@anchor{1fb}@anchor{gnat_ugn/platform_specific_information id34}@anchor{1fc}
24576 @subsubsection Compiling Resources
24586 This section describes how to build a GNAT-compatible (COFF) object file
24587 containing the resources. This is done using the Resource Compiler
24588 @code{windres} as follows:
24593 $ windres -i myres.rc -o myres.o
24597 By default @code{windres} will run @code{gcc} to preprocess the @code{.rc}
24598 file. You can specify an alternate preprocessor (usually named
24599 @code{cpp.exe}) using the @code{windres} @code{--preprocessor}
24600 parameter. A list of all possible options may be obtained by entering
24601 the command @code{windres} @code{--help}.
24603 It is also possible to use the Microsoft resource compiler @code{rc.exe}
24604 to produce a @code{.res} file (binary resource file). See the
24605 corresponding Microsoft documentation for further details. In this case
24606 you need to use @code{windres} to translate the @code{.res} file to a
24607 GNAT-compatible object file as follows:
24612 $ windres -i myres.res -o myres.o
24616 @node Using Resources,,Compiling Resources,GNAT and Windows Resources
24617 @anchor{gnat_ugn/platform_specific_information id35}@anchor{1fd}@anchor{gnat_ugn/platform_specific_information using-resources}@anchor{1fe}
24618 @subsubsection Using Resources
24624 To include the resource file in your program just add the
24625 GNAT-compatible object file for the resource(s) to the linker
24626 arguments. With @code{gnatmake} this is done by using the @code{-largs}
24632 $ gnatmake myprog -largs myres.o
24636 @node Using GNAT DLLs from Microsoft Visual Studio Applications,Debugging a DLL,GNAT and Windows Resources,Mixed-Language Programming on Windows
24637 @anchor{gnat_ugn/platform_specific_information using-gnat-dll-from-msvs}@anchor{1ff}@anchor{gnat_ugn/platform_specific_information using-gnat-dlls-from-microsoft-visual-studio-applications}@anchor{200}
24638 @subsubsection Using GNAT DLLs from Microsoft Visual Studio Applications
24641 @geindex Microsoft Visual Studio
24642 @geindex use with GNAT DLLs
24644 This section describes a common case of mixed GNAT/Microsoft Visual Studio
24645 application development, where the main program is developed using MSVS, and
24646 is linked with a DLL developed using GNAT. Such a mixed application should
24647 be developed following the general guidelines outlined above; below is the
24648 cookbook-style sequence of steps to follow:
24654 First develop and build the GNAT shared library using a library project
24655 (let’s assume the project is @code{mylib.gpr}, producing the library @code{libmylib.dll}):
24661 $ gprbuild -p mylib.gpr
24669 Produce a .def file for the symbols you need to interface with, either by
24670 hand or automatically with possibly some manual adjustments
24671 (see @ref{1e1,,Creating Definition File Automatically}):
24677 $ dlltool libmylib.dll -z libmylib.def --export-all-symbols
24685 Make sure that MSVS command-line tools are accessible on the path.
24688 Create the Microsoft-style import library (see @ref{1e4,,MSVS-Style Import Library}):
24694 $ lib -machine:IX86 -def:libmylib.def -out:libmylib.lib
24698 If you are using a 64-bit toolchain, the above becomes…
24703 $ lib -machine:X64 -def:libmylib.def -out:libmylib.lib
24717 $ cl /O2 /MD main.c libmylib.lib
24725 Before running the executable, make sure you have set the PATH to the DLL,
24726 or copy the DLL into into the directory containing the .exe.
24729 @node Debugging a DLL,Setting Stack Size from gnatlink,Using GNAT DLLs from Microsoft Visual Studio Applications,Mixed-Language Programming on Windows
24730 @anchor{gnat_ugn/platform_specific_information debugging-a-dll}@anchor{201}@anchor{gnat_ugn/platform_specific_information id36}@anchor{202}
24731 @subsubsection Debugging a DLL
24734 @geindex DLL debugging
24736 Debugging a DLL is similar to debugging a standard program. But
24737 we have to deal with two different executable parts: the DLL and the
24738 program that uses it. We have the following four possibilities:
24744 The program and the DLL are built with GCC/GNAT.
24747 The program is built with foreign tools and the DLL is built with
24751 The program is built with GCC/GNAT and the DLL is built with
24755 In this section we address only cases one and two above.
24756 There is no point in trying to debug
24757 a DLL with GNU/GDB, if there is no GDB-compatible debugging
24758 information in it. To do so you must use a debugger compatible with the
24759 tools suite used to build the DLL.
24762 * Program and DLL Both Built with GCC/GNAT::
24763 * Program Built with Foreign Tools and DLL Built with GCC/GNAT::
24767 @node Program and DLL Both Built with GCC/GNAT,Program Built with Foreign Tools and DLL Built with GCC/GNAT,,Debugging a DLL
24768 @anchor{gnat_ugn/platform_specific_information id37}@anchor{203}@anchor{gnat_ugn/platform_specific_information program-and-dll-both-built-with-gcc-gnat}@anchor{204}
24769 @subsubsection Program and DLL Both Built with GCC/GNAT
24772 This is the simplest case. Both the DLL and the program have @code{GDB}
24773 compatible debugging information. It is then possible to break anywhere in
24774 the process. Let’s suppose here that the main procedure is named
24775 @code{ada_main} and that in the DLL there is an entry point named
24778 The DLL (@ref{1da,,Introduction to Dynamic Link Libraries (DLLs)}) and
24779 program must have been built with the debugging information (see GNAT -g
24780 switch). Here are the step-by-step instructions for debugging it:
24786 Launch @code{GDB} on the main program.
24793 Start the program and stop at the beginning of the main procedure
24799 This step is required to be able to set a breakpoint inside the DLL. As long
24800 as the program is not run, the DLL is not loaded. This has the
24801 consequence that the DLL debugging information is also not loaded, so it is not
24802 possible to set a breakpoint in the DLL.
24805 Set a breakpoint inside the DLL
24808 (gdb) break ada_dll
24813 At this stage a breakpoint is set inside the DLL. From there on
24814 you can use the standard approach to debug the whole program
24815 (@ref{14d,,Running and Debugging Ada Programs}).
24817 @node Program Built with Foreign Tools and DLL Built with GCC/GNAT,,Program and DLL Both Built with GCC/GNAT,Debugging a DLL
24818 @anchor{gnat_ugn/platform_specific_information id38}@anchor{205}@anchor{gnat_ugn/platform_specific_information program-built-with-foreign-tools-and-dll-built-with-gcc-gnat}@anchor{206}
24819 @subsubsection Program Built with Foreign Tools and DLL Built with GCC/GNAT
24822 In this case things are slightly more complex because it is not possible to
24823 start the main program and then break at the beginning to load the DLL and the
24824 associated DLL debugging information. It is not possible to break at the
24825 beginning of the program because there is no @code{GDB} debugging information,
24826 and therefore there is no direct way of getting initial control. This
24827 section addresses this issue by describing some methods that can be used
24828 to break somewhere in the DLL to debug it.
24830 First suppose that the main procedure is named @code{main} (this is for
24831 example some C code built with Microsoft Visual C) and that there is a
24832 DLL named @code{test.dll} containing an Ada entry point named
24835 The DLL (see @ref{1da,,Introduction to Dynamic Link Libraries (DLLs)}) must have
24836 been built with debugging information (see the GNAT @code{-g} option).
24838 @subsubheading Debugging the DLL Directly
24845 Find out the executable starting address
24848 $ objdump --file-header main.exe
24851 The starting address is reported on the last line. For example:
24854 main.exe: file format pei-i386
24855 architecture: i386, flags 0x0000010a:
24856 EXEC_P, HAS_DEBUG, D_PAGED
24857 start address 0x00401010
24861 Launch the debugger on the executable.
24868 Set a breakpoint at the starting address, and launch the program.
24871 $ (gdb) break *0x00401010
24875 The program will stop at the given address.
24878 Set a breakpoint on a DLL subroutine.
24881 (gdb) break ada_dll.adb:45
24884 Or if you want to break using a symbol on the DLL, you need first to
24885 select the Ada language (language used by the DLL).
24888 (gdb) set language ada
24889 (gdb) break ada_dll
24893 Continue the program.
24899 This will run the program until it reaches the breakpoint that has been
24900 set. From that point you can use the standard way to debug a program
24901 as described in (@ref{14d,,Running and Debugging Ada Programs}).
24904 It is also possible to debug the DLL by attaching to a running process.
24906 @subsubheading Attaching to a Running Process
24909 @geindex DLL debugging
24910 @geindex attach to process
24912 With @code{GDB} it is always possible to debug a running process by
24913 attaching to it. It is possible to debug a DLL this way. The limitation
24914 of this approach is that the DLL must run long enough to perform the
24915 attach operation. It may be useful for instance to insert a time wasting
24916 loop in the code of the DLL to meet this criterion.
24922 Launch the main program @code{main.exe}.
24929 Use the Windows @emph{Task Manager} to find the process ID. Let’s say
24930 that the process PID for @code{main.exe} is 208.
24940 Attach to the running process to be debugged.
24947 Load the process debugging information.
24950 (gdb) symbol-file main.exe
24954 Break somewhere in the DLL.
24957 (gdb) break ada_dll
24961 Continue process execution.
24968 This last step will resume the process execution, and stop at
24969 the breakpoint we have set. From there you can use the standard
24970 approach to debug a program as described in
24971 @ref{14d,,Running and Debugging Ada Programs}.
24973 @node Setting Stack Size from gnatlink,Setting Heap Size from gnatlink,Debugging a DLL,Mixed-Language Programming on Windows
24974 @anchor{gnat_ugn/platform_specific_information id39}@anchor{207}@anchor{gnat_ugn/platform_specific_information setting-stack-size-from-gnatlink}@anchor{127}
24975 @subsubsection Setting Stack Size from @code{gnatlink}
24978 It is possible to specify the program stack size at link time. On modern
24979 versions of Windows, starting with XP, this is mostly useful to set the size of
24980 the main stack (environment task). The other task stacks are set with pragma
24981 Storage_Size or with the @emph{gnatbind -d} command.
24983 Since older versions of Windows (2000, NT4, etc.) do not allow setting the
24984 reserve size of individual tasks, the link-time stack size applies to all
24985 tasks, and pragma Storage_Size has no effect.
24986 In particular, Stack Overflow checks are made against this
24987 link-time specified size.
24989 This setting can be done with @code{gnatlink} using either of the following:
24995 @code{-Xlinker} linker option
24998 $ gnatlink hello -Xlinker --stack=0x10000,0x1000
25001 This sets the stack reserve size to 0x10000 bytes and the stack commit
25002 size to 0x1000 bytes.
25005 @code{-Wl} linker option
25008 $ gnatlink hello -Wl,--stack=0x1000000
25011 This sets the stack reserve size to 0x1000000 bytes. Note that with
25012 @code{-Wl} option it is not possible to set the stack commit size
25013 because the comma is a separator for this option.
25016 @node Setting Heap Size from gnatlink,,Setting Stack Size from gnatlink,Mixed-Language Programming on Windows
25017 @anchor{gnat_ugn/platform_specific_information id40}@anchor{208}@anchor{gnat_ugn/platform_specific_information setting-heap-size-from-gnatlink}@anchor{128}
25018 @subsubsection Setting Heap Size from @code{gnatlink}
25021 Under Windows systems, it is possible to specify the program heap size from
25022 @code{gnatlink} using either of the following:
25028 @code{-Xlinker} linker option
25031 $ gnatlink hello -Xlinker --heap=0x10000,0x1000
25034 This sets the heap reserve size to 0x10000 bytes and the heap commit
25035 size to 0x1000 bytes.
25038 @code{-Wl} linker option
25041 $ gnatlink hello -Wl,--heap=0x1000000
25044 This sets the heap reserve size to 0x1000000 bytes. Note that with
25045 @code{-Wl} option it is not possible to set the heap commit size
25046 because the comma is a separator for this option.
25049 @node Windows Specific Add-Ons,,Mixed-Language Programming on Windows,Microsoft Windows Topics
25050 @anchor{gnat_ugn/platform_specific_information win32-specific-addons}@anchor{209}@anchor{gnat_ugn/platform_specific_information windows-specific-add-ons}@anchor{20a}
25051 @subsection Windows Specific Add-Ons
25054 This section describes the Windows specific add-ons.
25062 @node Win32Ada,wPOSIX,,Windows Specific Add-Ons
25063 @anchor{gnat_ugn/platform_specific_information id41}@anchor{20b}@anchor{gnat_ugn/platform_specific_information win32ada}@anchor{20c}
25064 @subsubsection Win32Ada
25067 Win32Ada is a binding for the Microsoft Win32 API. This binding can be
25068 easily installed from the provided installer. To use the Win32Ada
25069 binding you need to use a project file, and adding a single with_clause
25070 will give you full access to the Win32Ada binding sources and ensure
25071 that the proper libraries are passed to the linker.
25078 for Sources use ...;
25083 To build the application you just need to call gprbuild for the
25084 application’s project, here p.gpr:
25093 @node wPOSIX,,Win32Ada,Windows Specific Add-Ons
25094 @anchor{gnat_ugn/platform_specific_information id42}@anchor{20d}@anchor{gnat_ugn/platform_specific_information wposix}@anchor{20e}
25095 @subsubsection wPOSIX
25098 wPOSIX is a minimal POSIX binding whose goal is to help with building
25099 cross-platforms applications. This binding is not complete though, as
25100 the Win32 API does not provide the necessary support for all POSIX APIs.
25102 To use the wPOSIX binding you need to use a project file, and adding
25103 a single with_clause will give you full access to the wPOSIX binding
25104 sources and ensure that the proper libraries are passed to the linker.
25111 for Sources use ...;
25116 To build the application you just need to call gprbuild for the
25117 application’s project, here p.gpr:
25126 @node Mac OS Topics,,Microsoft Windows Topics,Platform-Specific Information
25127 @anchor{gnat_ugn/platform_specific_information id43}@anchor{20f}@anchor{gnat_ugn/platform_specific_information mac-os-topics}@anchor{210}
25128 @section Mac OS Topics
25133 This section describes topics that are specific to Apple’s OS X
25137 * Codesigning the Debugger::
25141 @node Codesigning the Debugger,,,Mac OS Topics
25142 @anchor{gnat_ugn/platform_specific_information codesigning-the-debugger}@anchor{211}
25143 @subsection Codesigning the Debugger
25146 The Darwin Kernel requires the debugger to have special permissions
25147 before it is allowed to control other processes. These permissions
25148 are granted by codesigning the GDB executable. Without these
25149 permissions, the debugger will report error messages such as:
25152 Starting program: /x/y/foo
25153 Unable to find Mach task port for process-id 28885: (os/kern) failure (0x5).
25154 (please check gdb is codesigned - see taskgated(8))
25157 Codesigning requires a certificate. The following procedure explains
25164 Start the Keychain Access application (in
25165 /Applications/Utilities/Keychain Access.app)
25168 Select the Keychain Access -> Certificate Assistant ->
25169 Create a Certificate… menu
25178 Choose a name for the new certificate (this procedure will use
25179 “gdb-cert” as an example)
25182 Set “Identity Type” to “Self Signed Root”
25185 Set “Certificate Type” to “Code Signing”
25188 Activate the “Let me override defaults” option
25192 Click several times on “Continue” until the “Specify a Location
25193 For The Certificate” screen appears, then set “Keychain” to “System”
25196 Click on “Continue” until the certificate is created
25199 Finally, in the view, double-click on the new certificate,
25200 and set “When using this certificate” to “Always Trust”
25203 Exit the Keychain Access application and restart the computer
25204 (this is unfortunately required)
25207 Once a certificate has been created, the debugger can be codesigned
25208 as follow. In a Terminal, run the following command:
25213 $ codesign -f -s "gdb-cert" <gnat_install_prefix>/bin/gdb
25217 where “gdb-cert” should be replaced by the actual certificate
25218 name chosen above, and <gnat_install_prefix> should be replaced by
25219 the location where you installed GNAT. Also, be sure that users are
25220 in the Unix group @code{_developer}.
25222 @node Example of Binder Output File,Elaboration Order Handling in GNAT,Platform-Specific Information,Top
25223 @anchor{gnat_ugn/example_of_binder_output doc}@anchor{212}@anchor{gnat_ugn/example_of_binder_output example-of-binder-output-file}@anchor{e}@anchor{gnat_ugn/example_of_binder_output id1}@anchor{213}
25224 @chapter Example of Binder Output File
25227 @geindex Binder output (example)
25229 This Appendix displays the source code for the output file
25230 generated by @emph{gnatbind} for a simple ‘Hello World’ program.
25231 Comments have been added for clarification purposes.
25234 -- The package is called Ada_Main unless this name is actually used
25235 -- as a unit name in the partition, in which case some other unique
25240 package ada_main is
25241 pragma Warnings (Off);
25243 -- The main program saves the parameters (argument count,
25244 -- argument values, environment pointer) in global variables
25245 -- for later access by other units including
25246 -- Ada.Command_Line.
25248 gnat_argc : Integer;
25249 gnat_argv : System.Address;
25250 gnat_envp : System.Address;
25252 -- The actual variables are stored in a library routine. This
25253 -- is useful for some shared library situations, where there
25254 -- are problems if variables are not in the library.
25256 pragma Import (C, gnat_argc);
25257 pragma Import (C, gnat_argv);
25258 pragma Import (C, gnat_envp);
25260 -- The exit status is similarly an external location
25262 gnat_exit_status : Integer;
25263 pragma Import (C, gnat_exit_status);
25265 GNAT_Version : constant String :=
25266 "GNAT Version: Pro 7.4.0w (20141119-49)" & ASCII.NUL;
25267 pragma Export (C, GNAT_Version, "__gnat_version");
25269 Ada_Main_Program_Name : constant String := "_ada_hello" & ASCII.NUL;
25270 pragma Export (C, Ada_Main_Program_Name, "__gnat_ada_main_program_name");
25272 -- This is the generated adainit routine that performs
25273 -- initialization at the start of execution. In the case
25274 -- where Ada is the main program, this main program makes
25275 -- a call to adainit at program startup.
25278 pragma Export (C, adainit, "adainit");
25280 -- This is the generated adafinal routine that performs
25281 -- finalization at the end of execution. In the case where
25282 -- Ada is the main program, this main program makes a call
25283 -- to adafinal at program termination.
25285 procedure adafinal;
25286 pragma Export (C, adafinal, "adafinal");
25288 -- This routine is called at the start of execution. It is
25289 -- a dummy routine that is used by the debugger to breakpoint
25290 -- at the start of execution.
25292 -- This is the actual generated main program (it would be
25293 -- suppressed if the no main program switch were used). As
25294 -- required by standard system conventions, this program has
25295 -- the external name main.
25299 argv : System.Address;
25300 envp : System.Address)
25302 pragma Export (C, main, "main");
25304 -- The following set of constants give the version
25305 -- identification values for every unit in the bound
25306 -- partition. This identification is computed from all
25307 -- dependent semantic units, and corresponds to the
25308 -- string that would be returned by use of the
25309 -- Body_Version or Version attributes.
25311 -- The following Export pragmas export the version numbers
25312 -- with symbolic names ending in B (for body) or S
25313 -- (for spec) so that they can be located in a link. The
25314 -- information provided here is sufficient to track down
25315 -- the exact versions of units used in a given build.
25317 type Version_32 is mod 2 ** 32;
25318 u00001 : constant Version_32 := 16#8ad6e54a#;
25319 pragma Export (C, u00001, "helloB");
25320 u00002 : constant Version_32 := 16#fbff4c67#;
25321 pragma Export (C, u00002, "system__standard_libraryB");
25322 u00003 : constant Version_32 := 16#1ec6fd90#;
25323 pragma Export (C, u00003, "system__standard_libraryS");
25324 u00004 : constant Version_32 := 16#3ffc8e18#;
25325 pragma Export (C, u00004, "adaS");
25326 u00005 : constant Version_32 := 16#28f088c2#;
25327 pragma Export (C, u00005, "ada__text_ioB");
25328 u00006 : constant Version_32 := 16#f372c8ac#;
25329 pragma Export (C, u00006, "ada__text_ioS");
25330 u00007 : constant Version_32 := 16#2c143749#;
25331 pragma Export (C, u00007, "ada__exceptionsB");
25332 u00008 : constant Version_32 := 16#f4f0cce8#;
25333 pragma Export (C, u00008, "ada__exceptionsS");
25334 u00009 : constant Version_32 := 16#a46739c0#;
25335 pragma Export (C, u00009, "ada__exceptions__last_chance_handlerB");
25336 u00010 : constant Version_32 := 16#3aac8c92#;
25337 pragma Export (C, u00010, "ada__exceptions__last_chance_handlerS");
25338 u00011 : constant Version_32 := 16#1d274481#;
25339 pragma Export (C, u00011, "systemS");
25340 u00012 : constant Version_32 := 16#a207fefe#;
25341 pragma Export (C, u00012, "system__soft_linksB");
25342 u00013 : constant Version_32 := 16#467d9556#;
25343 pragma Export (C, u00013, "system__soft_linksS");
25344 u00014 : constant Version_32 := 16#b01dad17#;
25345 pragma Export (C, u00014, "system__parametersB");
25346 u00015 : constant Version_32 := 16#630d49fe#;
25347 pragma Export (C, u00015, "system__parametersS");
25348 u00016 : constant Version_32 := 16#b19b6653#;
25349 pragma Export (C, u00016, "system__secondary_stackB");
25350 u00017 : constant Version_32 := 16#b6468be8#;
25351 pragma Export (C, u00017, "system__secondary_stackS");
25352 u00018 : constant Version_32 := 16#39a03df9#;
25353 pragma Export (C, u00018, "system__storage_elementsB");
25354 u00019 : constant Version_32 := 16#30e40e85#;
25355 pragma Export (C, u00019, "system__storage_elementsS");
25356 u00020 : constant Version_32 := 16#41837d1e#;
25357 pragma Export (C, u00020, "system__stack_checkingB");
25358 u00021 : constant Version_32 := 16#93982f69#;
25359 pragma Export (C, u00021, "system__stack_checkingS");
25360 u00022 : constant Version_32 := 16#393398c1#;
25361 pragma Export (C, u00022, "system__exception_tableB");
25362 u00023 : constant Version_32 := 16#b33e2294#;
25363 pragma Export (C, u00023, "system__exception_tableS");
25364 u00024 : constant Version_32 := 16#ce4af020#;
25365 pragma Export (C, u00024, "system__exceptionsB");
25366 u00025 : constant Version_32 := 16#75442977#;
25367 pragma Export (C, u00025, "system__exceptionsS");
25368 u00026 : constant Version_32 := 16#37d758f1#;
25369 pragma Export (C, u00026, "system__exceptions__machineS");
25370 u00027 : constant Version_32 := 16#b895431d#;
25371 pragma Export (C, u00027, "system__exceptions_debugB");
25372 u00028 : constant Version_32 := 16#aec55d3f#;
25373 pragma Export (C, u00028, "system__exceptions_debugS");
25374 u00029 : constant Version_32 := 16#570325c8#;
25375 pragma Export (C, u00029, "system__img_intB");
25376 u00030 : constant Version_32 := 16#1ffca443#;
25377 pragma Export (C, u00030, "system__img_intS");
25378 u00031 : constant Version_32 := 16#b98c3e16#;
25379 pragma Export (C, u00031, "system__tracebackB");
25380 u00032 : constant Version_32 := 16#831a9d5a#;
25381 pragma Export (C, u00032, "system__tracebackS");
25382 u00033 : constant Version_32 := 16#9ed49525#;
25383 pragma Export (C, u00033, "system__traceback_entriesB");
25384 u00034 : constant Version_32 := 16#1d7cb2f1#;
25385 pragma Export (C, u00034, "system__traceback_entriesS");
25386 u00035 : constant Version_32 := 16#8c33a517#;
25387 pragma Export (C, u00035, "system__wch_conB");
25388 u00036 : constant Version_32 := 16#065a6653#;
25389 pragma Export (C, u00036, "system__wch_conS");
25390 u00037 : constant Version_32 := 16#9721e840#;
25391 pragma Export (C, u00037, "system__wch_stwB");
25392 u00038 : constant Version_32 := 16#2b4b4a52#;
25393 pragma Export (C, u00038, "system__wch_stwS");
25394 u00039 : constant Version_32 := 16#92b797cb#;
25395 pragma Export (C, u00039, "system__wch_cnvB");
25396 u00040 : constant Version_32 := 16#09eddca0#;
25397 pragma Export (C, u00040, "system__wch_cnvS");
25398 u00041 : constant Version_32 := 16#6033a23f#;
25399 pragma Export (C, u00041, "interfacesS");
25400 u00042 : constant Version_32 := 16#ece6fdb6#;
25401 pragma Export (C, u00042, "system__wch_jisB");
25402 u00043 : constant Version_32 := 16#899dc581#;
25403 pragma Export (C, u00043, "system__wch_jisS");
25404 u00044 : constant Version_32 := 16#10558b11#;
25405 pragma Export (C, u00044, "ada__streamsB");
25406 u00045 : constant Version_32 := 16#2e6701ab#;
25407 pragma Export (C, u00045, "ada__streamsS");
25408 u00046 : constant Version_32 := 16#db5c917c#;
25409 pragma Export (C, u00046, "ada__io_exceptionsS");
25410 u00047 : constant Version_32 := 16#12c8cd7d#;
25411 pragma Export (C, u00047, "ada__tagsB");
25412 u00048 : constant Version_32 := 16#ce72c228#;
25413 pragma Export (C, u00048, "ada__tagsS");
25414 u00049 : constant Version_32 := 16#c3335bfd#;
25415 pragma Export (C, u00049, "system__htableB");
25416 u00050 : constant Version_32 := 16#99e5f76b#;
25417 pragma Export (C, u00050, "system__htableS");
25418 u00051 : constant Version_32 := 16#089f5cd0#;
25419 pragma Export (C, u00051, "system__string_hashB");
25420 u00052 : constant Version_32 := 16#3bbb9c15#;
25421 pragma Export (C, u00052, "system__string_hashS");
25422 u00053 : constant Version_32 := 16#807fe041#;
25423 pragma Export (C, u00053, "system__unsigned_typesS");
25424 u00054 : constant Version_32 := 16#d27be59e#;
25425 pragma Export (C, u00054, "system__val_lluB");
25426 u00055 : constant Version_32 := 16#fa8db733#;
25427 pragma Export (C, u00055, "system__val_lluS");
25428 u00056 : constant Version_32 := 16#27b600b2#;
25429 pragma Export (C, u00056, "system__val_utilB");
25430 u00057 : constant Version_32 := 16#b187f27f#;
25431 pragma Export (C, u00057, "system__val_utilS");
25432 u00058 : constant Version_32 := 16#d1060688#;
25433 pragma Export (C, u00058, "system__case_utilB");
25434 u00059 : constant Version_32 := 16#392e2d56#;
25435 pragma Export (C, u00059, "system__case_utilS");
25436 u00060 : constant Version_32 := 16#84a27f0d#;
25437 pragma Export (C, u00060, "interfaces__c_streamsB");
25438 u00061 : constant Version_32 := 16#8bb5f2c0#;
25439 pragma Export (C, u00061, "interfaces__c_streamsS");
25440 u00062 : constant Version_32 := 16#6db6928f#;
25441 pragma Export (C, u00062, "system__crtlS");
25442 u00063 : constant Version_32 := 16#4e6a342b#;
25443 pragma Export (C, u00063, "system__file_ioB");
25444 u00064 : constant Version_32 := 16#ba56a5e4#;
25445 pragma Export (C, u00064, "system__file_ioS");
25446 u00065 : constant Version_32 := 16#b7ab275c#;
25447 pragma Export (C, u00065, "ada__finalizationB");
25448 u00066 : constant Version_32 := 16#19f764ca#;
25449 pragma Export (C, u00066, "ada__finalizationS");
25450 u00067 : constant Version_32 := 16#95817ed8#;
25451 pragma Export (C, u00067, "system__finalization_rootB");
25452 u00068 : constant Version_32 := 16#52d53711#;
25453 pragma Export (C, u00068, "system__finalization_rootS");
25454 u00069 : constant Version_32 := 16#769e25e6#;
25455 pragma Export (C, u00069, "interfaces__cB");
25456 u00070 : constant Version_32 := 16#4a38bedb#;
25457 pragma Export (C, u00070, "interfaces__cS");
25458 u00071 : constant Version_32 := 16#07e6ee66#;
25459 pragma Export (C, u00071, "system__os_libB");
25460 u00072 : constant Version_32 := 16#d7b69782#;
25461 pragma Export (C, u00072, "system__os_libS");
25462 u00073 : constant Version_32 := 16#1a817b8e#;
25463 pragma Export (C, u00073, "system__stringsB");
25464 u00074 : constant Version_32 := 16#639855e7#;
25465 pragma Export (C, u00074, "system__stringsS");
25466 u00075 : constant Version_32 := 16#e0b8de29#;
25467 pragma Export (C, u00075, "system__file_control_blockS");
25468 u00076 : constant Version_32 := 16#b5b2aca1#;
25469 pragma Export (C, u00076, "system__finalization_mastersB");
25470 u00077 : constant Version_32 := 16#69316dc1#;
25471 pragma Export (C, u00077, "system__finalization_mastersS");
25472 u00078 : constant Version_32 := 16#57a37a42#;
25473 pragma Export (C, u00078, "system__address_imageB");
25474 u00079 : constant Version_32 := 16#bccbd9bb#;
25475 pragma Export (C, u00079, "system__address_imageS");
25476 u00080 : constant Version_32 := 16#7268f812#;
25477 pragma Export (C, u00080, "system__img_boolB");
25478 u00081 : constant Version_32 := 16#e8fe356a#;
25479 pragma Export (C, u00081, "system__img_boolS");
25480 u00082 : constant Version_32 := 16#d7aac20c#;
25481 pragma Export (C, u00082, "system__ioB");
25482 u00083 : constant Version_32 := 16#8365b3ce#;
25483 pragma Export (C, u00083, "system__ioS");
25484 u00084 : constant Version_32 := 16#6d4d969a#;
25485 pragma Export (C, u00084, "system__storage_poolsB");
25486 u00085 : constant Version_32 := 16#e87cc305#;
25487 pragma Export (C, u00085, "system__storage_poolsS");
25488 u00086 : constant Version_32 := 16#e34550ca#;
25489 pragma Export (C, u00086, "system__pool_globalB");
25490 u00087 : constant Version_32 := 16#c88d2d16#;
25491 pragma Export (C, u00087, "system__pool_globalS");
25492 u00088 : constant Version_32 := 16#9d39c675#;
25493 pragma Export (C, u00088, "system__memoryB");
25494 u00089 : constant Version_32 := 16#445a22b5#;
25495 pragma Export (C, u00089, "system__memoryS");
25496 u00090 : constant Version_32 := 16#6a859064#;
25497 pragma Export (C, u00090, "system__storage_pools__subpoolsB");
25498 u00091 : constant Version_32 := 16#e3b008dc#;
25499 pragma Export (C, u00091, "system__storage_pools__subpoolsS");
25500 u00092 : constant Version_32 := 16#63f11652#;
25501 pragma Export (C, u00092, "system__storage_pools__subpools__finalizationB");
25502 u00093 : constant Version_32 := 16#fe2f4b3a#;
25503 pragma Export (C, u00093, "system__storage_pools__subpools__finalizationS");
25505 -- BEGIN ELABORATION ORDER
25509 -- system.case_util%s
25510 -- system.case_util%b
25512 -- system.img_bool%s
25513 -- system.img_bool%b
25514 -- system.img_int%s
25515 -- system.img_int%b
25518 -- system.parameters%s
25519 -- system.parameters%b
25521 -- interfaces.c_streams%s
25522 -- interfaces.c_streams%b
25523 -- system.standard_library%s
25524 -- system.exceptions_debug%s
25525 -- system.exceptions_debug%b
25526 -- system.storage_elements%s
25527 -- system.storage_elements%b
25528 -- system.stack_checking%s
25529 -- system.stack_checking%b
25530 -- system.string_hash%s
25531 -- system.string_hash%b
25533 -- system.strings%s
25534 -- system.strings%b
25536 -- system.traceback_entries%s
25537 -- system.traceback_entries%b
25538 -- ada.exceptions%s
25539 -- system.soft_links%s
25540 -- system.unsigned_types%s
25541 -- system.val_llu%s
25542 -- system.val_util%s
25543 -- system.val_util%b
25544 -- system.val_llu%b
25545 -- system.wch_con%s
25546 -- system.wch_con%b
25547 -- system.wch_cnv%s
25548 -- system.wch_jis%s
25549 -- system.wch_jis%b
25550 -- system.wch_cnv%b
25551 -- system.wch_stw%s
25552 -- system.wch_stw%b
25553 -- ada.exceptions.last_chance_handler%s
25554 -- ada.exceptions.last_chance_handler%b
25555 -- system.address_image%s
25556 -- system.exception_table%s
25557 -- system.exception_table%b
25558 -- ada.io_exceptions%s
25563 -- system.exceptions%s
25564 -- system.exceptions%b
25565 -- system.exceptions.machine%s
25566 -- system.finalization_root%s
25567 -- system.finalization_root%b
25568 -- ada.finalization%s
25569 -- ada.finalization%b
25570 -- system.storage_pools%s
25571 -- system.storage_pools%b
25572 -- system.finalization_masters%s
25573 -- system.storage_pools.subpools%s
25574 -- system.storage_pools.subpools.finalization%s
25575 -- system.storage_pools.subpools.finalization%b
25578 -- system.standard_library%b
25579 -- system.pool_global%s
25580 -- system.pool_global%b
25581 -- system.file_control_block%s
25582 -- system.file_io%s
25583 -- system.secondary_stack%s
25584 -- system.file_io%b
25585 -- system.storage_pools.subpools%b
25586 -- system.finalization_masters%b
25589 -- system.soft_links%b
25591 -- system.secondary_stack%b
25592 -- system.address_image%b
25593 -- system.traceback%s
25594 -- ada.exceptions%b
25595 -- system.traceback%b
25599 -- END ELABORATION ORDER
25606 -- The following source file name pragmas allow the generated file
25607 -- names to be unique for different main programs. They are needed
25608 -- since the package name will always be Ada_Main.
25610 pragma Source_File_Name (ada_main, Spec_File_Name => "b~hello.ads");
25611 pragma Source_File_Name (ada_main, Body_File_Name => "b~hello.adb");
25613 pragma Suppress (Overflow_Check);
25614 with Ada.Exceptions;
25616 -- Generated package body for Ada_Main starts here
25618 package body ada_main is
25619 pragma Warnings (Off);
25621 -- These values are reference counter associated to units which have
25622 -- been elaborated. It is also used to avoid elaborating the
25623 -- same unit twice.
25625 E72 : Short_Integer; pragma Import (Ada, E72, "system__os_lib_E");
25626 E13 : Short_Integer; pragma Import (Ada, E13, "system__soft_links_E");
25627 E23 : Short_Integer; pragma Import (Ada, E23, "system__exception_table_E");
25628 E46 : Short_Integer; pragma Import (Ada, E46, "ada__io_exceptions_E");
25629 E48 : Short_Integer; pragma Import (Ada, E48, "ada__tags_E");
25630 E45 : Short_Integer; pragma Import (Ada, E45, "ada__streams_E");
25631 E70 : Short_Integer; pragma Import (Ada, E70, "interfaces__c_E");
25632 E25 : Short_Integer; pragma Import (Ada, E25, "system__exceptions_E");
25633 E68 : Short_Integer; pragma Import (Ada, E68, "system__finalization_root_E");
25634 E66 : Short_Integer; pragma Import (Ada, E66, "ada__finalization_E");
25635 E85 : Short_Integer; pragma Import (Ada, E85, "system__storage_pools_E");
25636 E77 : Short_Integer; pragma Import (Ada, E77, "system__finalization_masters_E");
25637 E91 : Short_Integer; pragma Import (Ada, E91, "system__storage_pools__subpools_E");
25638 E87 : Short_Integer; pragma Import (Ada, E87, "system__pool_global_E");
25639 E75 : Short_Integer; pragma Import (Ada, E75, "system__file_control_block_E");
25640 E64 : Short_Integer; pragma Import (Ada, E64, "system__file_io_E");
25641 E17 : Short_Integer; pragma Import (Ada, E17, "system__secondary_stack_E");
25642 E06 : Short_Integer; pragma Import (Ada, E06, "ada__text_io_E");
25644 Local_Priority_Specific_Dispatching : constant String := "";
25645 Local_Interrupt_States : constant String := "";
25647 Is_Elaborated : Boolean := False;
25649 procedure finalize_library is
25654 pragma Import (Ada, F1, "ada__text_io__finalize_spec");
25662 pragma Import (Ada, F2, "system__file_io__finalize_body");
25669 pragma Import (Ada, F3, "system__file_control_block__finalize_spec");
25677 pragma Import (Ada, F4, "system__pool_global__finalize_spec");
25683 pragma Import (Ada, F5, "system__storage_pools__subpools__finalize_spec");
25689 pragma Import (Ada, F6, "system__finalization_masters__finalize_spec");
25694 procedure Reraise_Library_Exception_If_Any;
25695 pragma Import (Ada, Reraise_Library_Exception_If_Any, "__gnat_reraise_library_exception_if_any");
25697 Reraise_Library_Exception_If_Any;
25699 end finalize_library;
25705 procedure adainit is
25707 Main_Priority : Integer;
25708 pragma Import (C, Main_Priority, "__gl_main_priority");
25709 Time_Slice_Value : Integer;
25710 pragma Import (C, Time_Slice_Value, "__gl_time_slice_val");
25711 WC_Encoding : Character;
25712 pragma Import (C, WC_Encoding, "__gl_wc_encoding");
25713 Locking_Policy : Character;
25714 pragma Import (C, Locking_Policy, "__gl_locking_policy");
25715 Queuing_Policy : Character;
25716 pragma Import (C, Queuing_Policy, "__gl_queuing_policy");
25717 Task_Dispatching_Policy : Character;
25718 pragma Import (C, Task_Dispatching_Policy, "__gl_task_dispatching_policy");
25719 Priority_Specific_Dispatching : System.Address;
25720 pragma Import (C, Priority_Specific_Dispatching, "__gl_priority_specific_dispatching");
25721 Num_Specific_Dispatching : Integer;
25722 pragma Import (C, Num_Specific_Dispatching, "__gl_num_specific_dispatching");
25723 Main_CPU : Integer;
25724 pragma Import (C, Main_CPU, "__gl_main_cpu");
25725 Interrupt_States : System.Address;
25726 pragma Import (C, Interrupt_States, "__gl_interrupt_states");
25727 Num_Interrupt_States : Integer;
25728 pragma Import (C, Num_Interrupt_States, "__gl_num_interrupt_states");
25729 Unreserve_All_Interrupts : Integer;
25730 pragma Import (C, Unreserve_All_Interrupts, "__gl_unreserve_all_interrupts");
25731 Detect_Blocking : Integer;
25732 pragma Import (C, Detect_Blocking, "__gl_detect_blocking");
25733 Default_Stack_Size : Integer;
25734 pragma Import (C, Default_Stack_Size, "__gl_default_stack_size");
25735 Leap_Seconds_Support : Integer;
25736 pragma Import (C, Leap_Seconds_Support, "__gl_leap_seconds_support");
25738 procedure Runtime_Initialize;
25739 pragma Import (C, Runtime_Initialize, "__gnat_runtime_initialize");
25741 Finalize_Library_Objects : No_Param_Proc;
25742 pragma Import (C, Finalize_Library_Objects, "__gnat_finalize_library_objects");
25744 -- Start of processing for adainit
25748 -- Record various information for this partition. The values
25749 -- are derived by the binder from information stored in the ali
25750 -- files by the compiler.
25752 if Is_Elaborated then
25755 Is_Elaborated := True;
25756 Main_Priority := -1;
25757 Time_Slice_Value := -1;
25758 WC_Encoding := 'b';
25759 Locking_Policy := ' ';
25760 Queuing_Policy := ' ';
25761 Task_Dispatching_Policy := ' ';
25762 Priority_Specific_Dispatching :=
25763 Local_Priority_Specific_Dispatching'Address;
25764 Num_Specific_Dispatching := 0;
25766 Interrupt_States := Local_Interrupt_States'Address;
25767 Num_Interrupt_States := 0;
25768 Unreserve_All_Interrupts := 0;
25769 Detect_Blocking := 0;
25770 Default_Stack_Size := -1;
25771 Leap_Seconds_Support := 0;
25773 Runtime_Initialize;
25775 Finalize_Library_Objects := finalize_library'access;
25777 -- Now we have the elaboration calls for all units in the partition.
25778 -- The Elab_Spec and Elab_Body attributes generate references to the
25779 -- implicit elaboration procedures generated by the compiler for
25780 -- each unit that requires elaboration. Increment a counter of
25781 -- reference for each unit.
25783 System.Soft_Links'Elab_Spec;
25784 System.Exception_Table'Elab_Body;
25786 Ada.Io_Exceptions'Elab_Spec;
25788 Ada.Tags'Elab_Spec;
25789 Ada.Streams'Elab_Spec;
25791 Interfaces.C'Elab_Spec;
25792 System.Exceptions'Elab_Spec;
25794 System.Finalization_Root'Elab_Spec;
25796 Ada.Finalization'Elab_Spec;
25798 System.Storage_Pools'Elab_Spec;
25800 System.Finalization_Masters'Elab_Spec;
25801 System.Storage_Pools.Subpools'Elab_Spec;
25802 System.Pool_Global'Elab_Spec;
25804 System.File_Control_Block'Elab_Spec;
25806 System.File_Io'Elab_Body;
25809 System.Finalization_Masters'Elab_Body;
25812 Ada.Tags'Elab_Body;
25814 System.Soft_Links'Elab_Body;
25816 System.Os_Lib'Elab_Body;
25818 System.Secondary_Stack'Elab_Body;
25820 Ada.Text_Io'Elab_Spec;
25821 Ada.Text_Io'Elab_Body;
25829 procedure adafinal is
25830 procedure s_stalib_adafinal;
25831 pragma Import (C, s_stalib_adafinal, "system__standard_library__adafinal");
25833 procedure Runtime_Finalize;
25834 pragma Import (C, Runtime_Finalize, "__gnat_runtime_finalize");
25837 if not Is_Elaborated then
25840 Is_Elaborated := False;
25845 -- We get to the main program of the partition by using
25846 -- pragma Import because if we try to with the unit and
25847 -- call it Ada style, then not only do we waste time
25848 -- recompiling it, but also, we don't really know the right
25849 -- switches (e.g.@@: identifier character set) to be used
25852 procedure Ada_Main_Program;
25853 pragma Import (Ada, Ada_Main_Program, "_ada_hello");
25859 -- main is actually a function, as in the ANSI C standard,
25860 -- defined to return the exit status. The three parameters
25861 -- are the argument count, argument values and environment
25866 argv : System.Address;
25867 envp : System.Address)
25870 -- The initialize routine performs low level system
25871 -- initialization using a standard library routine which
25872 -- sets up signal handling and performs any other
25873 -- required setup. The routine can be found in file
25876 procedure initialize;
25877 pragma Import (C, initialize, "__gnat_initialize");
25879 -- The finalize routine performs low level system
25880 -- finalization using a standard library routine. The
25881 -- routine is found in file a-final.c and in the standard
25882 -- distribution is a dummy routine that does nothing, so
25883 -- really this is a hook for special user finalization.
25885 procedure finalize;
25886 pragma Import (C, finalize, "__gnat_finalize");
25888 -- The following is to initialize the SEH exceptions
25890 SEH : aliased array (1 .. 2) of Integer;
25892 Ensure_Reference : aliased System.Address := Ada_Main_Program_Name'Address;
25893 pragma Volatile (Ensure_Reference);
25895 -- Start of processing for main
25898 -- Save global variables
25904 -- Call low level system initialization
25906 Initialize (SEH'Address);
25908 -- Call our generated Ada initialization routine
25912 -- Now we call the main program of the partition
25916 -- Perform Ada finalization
25920 -- Perform low level system finalization
25924 -- Return the proper exit status
25925 return (gnat_exit_status);
25928 -- This section is entirely comments, so it has no effect on the
25929 -- compilation of the Ada_Main package. It provides the list of
25930 -- object files and linker options, as well as some standard
25931 -- libraries needed for the link. The gnatlink utility parses
25932 -- this b~hello.adb file to read these comment lines to generate
25933 -- the appropriate command line arguments for the call to the
25934 -- system linker. The BEGIN/END lines are used for sentinels for
25935 -- this parsing operation.
25937 -- The exact file names will of course depend on the environment,
25938 -- host/target and location of files on the host system.
25940 -- BEGIN Object file/option list
25943 -- -L/usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/
25944 -- /usr/local/gnat/lib/gcc-lib/i686-pc-linux-gnu/2.8.1/adalib/libgnat.a
25945 -- END Object file/option list
25950 The Ada code in the above example is exactly what is generated by the
25951 binder. We have added comments to more clearly indicate the function
25952 of each part of the generated @code{Ada_Main} package.
25954 The code is standard Ada in all respects, and can be processed by any
25955 tools that handle Ada. In particular, it is possible to use the debugger
25956 in Ada mode to debug the generated @code{Ada_Main} package. For example,
25957 suppose that for reasons that you do not understand, your program is crashing
25958 during elaboration of the body of @code{Ada.Text_IO}. To locate this bug,
25959 you can place a breakpoint on the call:
25964 Ada.Text_Io'Elab_Body;
25968 and trace the elaboration routine for this package to find out where
25969 the problem might be (more usually of course you would be debugging
25970 elaboration code in your own application).
25972 @c -- Example: A |withing| unit has a |with| clause, it |withs| a |withed| unit
25974 @node Elaboration Order Handling in GNAT,Inline Assembler,Example of Binder Output File,Top
25975 @anchor{gnat_ugn/elaboration_order_handling_in_gnat doc}@anchor{214}@anchor{gnat_ugn/elaboration_order_handling_in_gnat elaboration-order-handling-in-gnat}@anchor{f}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id1}@anchor{215}
25976 @chapter Elaboration Order Handling in GNAT
25979 @geindex Order of elaboration
25981 @geindex Elaboration control
25983 This appendix describes the handling of elaboration code in Ada and GNAT, and
25984 discusses how the order of elaboration of program units can be controlled in
25985 GNAT, either automatically or with explicit programming features.
25988 * Elaboration Code::
25989 * Elaboration Order::
25990 * Checking the Elaboration Order::
25991 * Controlling the Elaboration Order in Ada::
25992 * Controlling the Elaboration Order in GNAT::
25993 * Mixing Elaboration Models::
25994 * ABE Diagnostics::
25995 * SPARK Diagnostics::
25996 * Elaboration Circularities::
25997 * Resolving Elaboration Circularities::
25998 * Elaboration-related Compiler Switches::
25999 * Summary of Procedures for Elaboration Control::
26000 * Inspecting the Chosen Elaboration Order::
26004 @node Elaboration Code,Elaboration Order,,Elaboration Order Handling in GNAT
26005 @anchor{gnat_ugn/elaboration_order_handling_in_gnat elaboration-code}@anchor{216}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id2}@anchor{217}
26006 @section Elaboration Code
26009 Ada defines the term @emph{execution} as the process by which a construct achieves
26010 its run-time effect. This process is also referred to as @strong{elaboration} for
26011 declarations and @emph{evaluation} for expressions.
26013 The execution model in Ada allows for certain sections of an Ada program to be
26014 executed prior to execution of the program itself, primarily with the intent of
26015 initializing data. These sections are referred to as @strong{elaboration code}.
26016 Elaboration code is executed as follows:
26022 All partitions of an Ada program are executed in parallel with one another,
26023 possibly in a separate address space, and possibly on a separate computer.
26026 The execution of a partition involves running the environment task for that
26030 The environment task executes all elaboration code (if available) for all
26031 units within that partition. This code is said to be executed at
26032 @strong{elaboration time}.
26035 The environment task executes the Ada program (if available) for that
26039 In addition to the Ada terminology, this appendix defines the following terms:
26047 The act of calling a subprogram, instantiating a generic, or activating a
26053 A construct that is elaborated or invoked by elaboration code is referred to
26054 as an @emph{elaboration scenario} or simply a @strong{scenario}. GNAT recognizes the
26055 following scenarios:
26061 @code{'Access} of entries, operators, and subprograms
26064 Activation of tasks
26067 Calls to entries, operators, and subprograms
26070 Instantiations of generic templates
26076 A construct elaborated by a scenario is referred to as @emph{elaboration target}
26077 or simply @strong{target}. GNAT recognizes the following targets:
26083 For @code{'Access} of entries, operators, and subprograms, the target is the
26084 entry, operator, or subprogram being aliased.
26087 For activation of tasks, the target is the task body
26090 For calls to entries, operators, and subprograms, the target is the entry,
26091 operator, or subprogram being invoked.
26094 For instantiations of generic templates, the target is the generic template
26095 being instantiated.
26099 Elaboration code may appear in two distinct contexts:
26105 @emph{Library level}
26107 A scenario appears at the library level when it is encapsulated by a package
26108 [body] compilation unit, ignoring any other package [body] declarations in
26117 Val : ... := Server.Func;
26122 In the example above, the call to @code{Server.Func} is an elaboration scenario
26123 because it appears at the library level of package @code{Client}. Note that the
26124 declaration of package @code{Nested} is ignored according to the definition
26125 given above. As a result, the call to @code{Server.Func} will be invoked when
26126 the spec of unit @code{Client} is elaborated.
26129 @emph{Package body statements}
26131 A scenario appears within the statement sequence of a package body when it is
26132 bounded by the region starting from the @code{begin} keyword of the package body
26133 and ending at the @code{end} keyword of the package body.
26136 package body Client is
26146 In the example above, the call to @code{Proc} is an elaboration scenario because
26147 it appears within the statement sequence of package body @code{Client}. As a
26148 result, the call to @code{Proc} will be invoked when the body of @code{Client} is
26152 @node Elaboration Order,Checking the Elaboration Order,Elaboration Code,Elaboration Order Handling in GNAT
26153 @anchor{gnat_ugn/elaboration_order_handling_in_gnat elaboration-order}@anchor{218}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id3}@anchor{219}
26154 @section Elaboration Order
26157 The sequence by which the elaboration code of all units within a partition is
26158 executed is referred to as @strong{elaboration order}.
26160 Within a single unit, elaboration code is executed in sequential order.
26165 package body Client is
26166 Result : ... := Server.Func;
26169 package Inst is new Server.Gen;
26171 Inst.Eval (Result);
26179 In the example above, the elaboration order within package body @code{Client} is
26186 The object declaration of @code{Result} is elaborated.
26192 Function @code{Server.Func} is invoked.
26196 The subprogram body of @code{Proc} is elaborated.
26199 Procedure @code{Proc} is invoked.
26205 Generic unit @code{Server.Gen} is instantiated as @code{Inst}.
26208 Instance @code{Inst} is elaborated.
26211 Procedure @code{Inst.Eval} is invoked.
26215 The elaboration order of all units within a partition depends on the following
26222 @emph{with}ed units
26231 preelaborability of units
26234 presence of elaboration-control pragmas
26237 invocations performed in elaboration code
26240 A program may have several elaboration orders depending on its structure.
26246 function Func (Index : Integer) return Integer;
26251 package body Server is
26252 Results : array (1 .. 5) of Integer := (1, 2, 3, 4, 5);
26254 function Func (Index : Integer) return Integer is
26256 return Results (Index);
26264 Val : constant Integer := Server.Func (3);
26270 procedure Main is begin null; end Main;
26274 The following elaboration order exhibits a fundamental problem referred to as
26275 @emph{access-before-elaboration} or simply @strong{ABE}.
26287 The elaboration of @code{Server}’s spec materializes function @code{Func}, making it
26288 callable. The elaboration of @code{Client}’s spec elaborates the declaration of
26289 @code{Val}. This invokes function @code{Server.Func}, however the body of
26290 @code{Server.Func} has not been elaborated yet because @code{Server}’s body comes
26291 after @code{Client}’s spec in the elaboration order. As a result, the value of
26292 constant @code{Val} is now undefined.
26294 Without any guarantees from the language, an undetected ABE problem may hinder
26295 proper initialization of data, which in turn may lead to undefined behavior at
26296 run time. To prevent such ABE problems, Ada employs dynamic checks in the same
26297 vein as index or null exclusion checks. A failed ABE check raises exception
26298 @code{Program_Error}.
26300 The following elaboration order avoids the ABE problem and the program can be
26301 successfully elaborated.
26313 Ada states that a total elaboration order must exist, but it does not define
26314 what this order is. A compiler is thus tasked with choosing a suitable
26315 elaboration order which satisfies the dependencies imposed by @emph{with} clauses,
26316 unit categorization, elaboration-control pragmas, and invocations performed in
26317 elaboration code. Ideally an order that avoids ABE problems should be chosen,
26318 however a compiler may not always find such an order due to complications with
26319 respect to control and data flow.
26321 @node Checking the Elaboration Order,Controlling the Elaboration Order in Ada,Elaboration Order,Elaboration Order Handling in GNAT
26322 @anchor{gnat_ugn/elaboration_order_handling_in_gnat checking-the-elaboration-order}@anchor{21a}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id4}@anchor{21b}
26323 @section Checking the Elaboration Order
26326 To avoid placing the entire elaboration-order burden on the programmer, Ada
26327 provides three lines of defense:
26333 @emph{Static semantics}
26335 Static semantic rules restrict the possible choice of elaboration order. For
26336 instance, if unit Client @emph{with}s unit Server, then the spec of Server is
26337 always elaborated prior to Client. The same principle applies to child units
26338 - the spec of a parent unit is always elaborated prior to the child unit.
26341 @emph{Dynamic semantics}
26343 Dynamic checks are performed at run time, to ensure that a target is
26344 elaborated prior to a scenario that invokes it, thus avoiding ABE problems.
26345 A failed run-time check raises exception @code{Program_Error}. The following
26346 restrictions apply:
26352 @emph{Restrictions on calls}
26354 An entry, operator, or subprogram can be called from elaboration code only
26355 when the corresponding body has been elaborated.
26358 @emph{Restrictions on instantiations}
26360 A generic unit can be instantiated by elaboration code only when the
26361 corresponding body has been elaborated.
26364 @emph{Restrictions on task activation}
26366 A task can be activated by elaboration code only when the body of the
26367 associated task type has been elaborated.
26370 The restrictions above can be summarized by the following rule:
26372 @emph{If a target has a body, then this body must be elaborated prior to the
26373 scenario that invokes the target.}
26376 @emph{Elaboration control}
26378 Pragmas are provided for the programmer to specify the desired elaboration
26382 @node Controlling the Elaboration Order in Ada,Controlling the Elaboration Order in GNAT,Checking the Elaboration Order,Elaboration Order Handling in GNAT
26383 @anchor{gnat_ugn/elaboration_order_handling_in_gnat controlling-the-elaboration-order-in-ada}@anchor{21c}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id5}@anchor{21d}
26384 @section Controlling the Elaboration Order in Ada
26387 Ada provides several idioms and pragmas to aid the programmer with specifying
26388 the desired elaboration order and avoiding ABE problems altogether.
26394 @emph{Packages without a body}
26396 A library package which does not require a completing body does not suffer
26402 type Element is private;
26403 package Containers is
26404 type Element_Array is array (1 .. 10) of Element;
26409 In the example above, package @code{Pack} does not require a body because it
26410 does not contain any constructs which require completion in a body. As a
26411 result, generic @code{Pack.Containers} can be instantiated without encountering
26415 @geindex pragma Pure
26423 Pragma @code{Pure} places sufficient restrictions on a unit to guarantee that no
26424 scenario within the unit can result in an ABE problem.
26427 @geindex pragma Preelaborate
26433 @emph{pragma Preelaborate}
26435 Pragma @code{Preelaborate} is slightly less restrictive than pragma @code{Pure},
26436 but still strong enough to prevent ABE problems within a unit.
26439 @geindex pragma Elaborate_Body
26445 @emph{pragma Elaborate_Body}
26447 Pragma @code{Elaborate_Body} requires that the body of a unit is elaborated
26448 immediately after its spec. This restriction guarantees that no client
26449 scenario can invoke a server target before the target body has been
26450 elaborated because the spec and body are effectively “glued” together.
26454 pragma Elaborate_Body;
26456 function Func return Integer;
26461 package body Server is
26462 function Func return Integer is
26472 Val : constant Integer := Server.Func;
26476 In the example above, pragma @code{Elaborate_Body} guarantees the following
26485 because the spec of @code{Server} must be elaborated prior to @code{Client} by
26486 virtue of the @emph{with} clause, and in addition the body of @code{Server} must be
26487 elaborated immediately after the spec of @code{Server}.
26489 Removing pragma @code{Elaborate_Body} could result in the following incorrect
26498 where @code{Client} invokes @code{Server.Func}, but the body of @code{Server.Func} has
26499 not been elaborated yet.
26502 The pragmas outlined above allow a server unit to guarantee safe elaboration
26503 use by client units. Thus it is a good rule to mark units as @code{Pure} or
26504 @code{Preelaborate}, and if this is not possible, mark them as @code{Elaborate_Body}.
26506 There are however situations where @code{Pure}, @code{Preelaborate}, and
26507 @code{Elaborate_Body} are not applicable. Ada provides another set of pragmas for
26508 use by client units to help ensure the elaboration safety of server units they
26511 @geindex pragma Elaborate (Unit)
26517 @emph{pragma Elaborate (Unit)}
26519 Pragma @code{Elaborate} can be placed in the context clauses of a unit, after a
26520 @emph{with} clause. It guarantees that both the spec and body of its argument will
26521 be elaborated prior to the unit with the pragma. Note that other unrelated
26522 units may be elaborated in between the spec and the body.
26526 function Func return Integer;
26531 package body Server is
26532 function Func return Integer is
26541 pragma Elaborate (Server);
26543 Val : constant Integer := Server.Func;
26547 In the example above, pragma @code{Elaborate} guarantees the following
26556 Removing pragma @code{Elaborate} could result in the following incorrect
26565 where @code{Client} invokes @code{Server.Func}, but the body of @code{Server.Func}
26566 has not been elaborated yet.
26569 @geindex pragma Elaborate_All (Unit)
26575 @emph{pragma Elaborate_All (Unit)}
26577 Pragma @code{Elaborate_All} is placed in the context clauses of a unit, after
26578 a @emph{with} clause. It guarantees that both the spec and body of its argument
26579 will be elaborated prior to the unit with the pragma, as well as all units
26580 @emph{with}ed by the spec and body of the argument, recursively. Note that other
26581 unrelated units may be elaborated in between the spec and the body.
26585 function Factorial (Val : Natural) return Natural;
26590 package body Math is
26591 function Factorial (Val : Natural) return Natural is
26599 package Computer is
26600 type Operation_Kind is (None, Op_Factorial);
26604 Op : Operation_Kind) return Natural;
26610 package body Computer is
26613 Op : Operation_Kind) return Natural
26615 if Op = Op_Factorial then
26616 return Math.Factorial (Val);
26626 pragma Elaborate_All (Computer);
26628 Val : constant Natural :=
26629 Computer.Compute (123, Computer.Op_Factorial);
26633 In the example above, pragma @code{Elaborate_All} can result in the following
26644 Note that there are several allowable suborders for the specs and bodies of
26645 @code{Math} and @code{Computer}, but the point is that these specs and bodies will
26646 be elaborated prior to @code{Client}.
26648 Removing pragma @code{Elaborate_All} could result in the following incorrect
26659 where @code{Client} invokes @code{Computer.Compute}, which in turn invokes
26660 @code{Math.Factorial}, but the body of @code{Math.Factorial} has not been
26664 All pragmas shown above can be summarized by the following rule:
26666 @emph{If a client unit elaborates a server target directly or indirectly, then if
26667 the server unit requires a body and does not have pragma Pure, Preelaborate,
26668 or Elaborate_Body, then the client unit should have pragma Elaborate or
26669 Elaborate_All for the server unit.}
26671 If the rule outlined above is not followed, then a program may fall in one of
26672 the following states:
26678 @emph{No elaboration order exists}
26680 In this case a compiler must diagnose the situation, and refuse to build an
26681 executable program.
26684 @emph{One or more incorrect elaboration orders exist}
26686 In this case a compiler can build an executable program, but
26687 @code{Program_Error} will be raised when the program is run.
26690 @emph{Several elaboration orders exist, some correct, some incorrect}
26692 In this case the programmer has not controlled the elaboration order. As a
26693 result, a compiler may or may not pick one of the correct orders, and the
26694 program may or may not raise @code{Program_Error} when it is run. This is the
26695 worst possible state because the program may fail on another compiler, or
26696 even another version of the same compiler.
26699 @emph{One or more correct orders exist}
26701 In this case a compiler can build an executable program, and the program is
26702 run successfully. This state may be guaranteed by following the outlined
26703 rules, or may be the result of good program architecture.
26706 Note that one additional advantage of using @code{Elaborate} and @code{Elaborate_All}
26707 is that the program continues to stay in the last state (one or more correct
26708 orders exist) even if maintenance changes the bodies of targets.
26710 @node Controlling the Elaboration Order in GNAT,Mixing Elaboration Models,Controlling the Elaboration Order in Ada,Elaboration Order Handling in GNAT
26711 @anchor{gnat_ugn/elaboration_order_handling_in_gnat controlling-the-elaboration-order-in-gnat}@anchor{21e}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id6}@anchor{21f}
26712 @section Controlling the Elaboration Order in GNAT
26715 In addition to Ada semantics and rules synthesized from them, GNAT offers
26716 three elaboration models to aid the programmer with specifying the correct
26717 elaboration order and to diagnose elaboration problems.
26719 @geindex Dynamic elaboration model
26725 @emph{Dynamic elaboration model}
26727 This is the most permissive of the three elaboration models and emulates the
26728 behavior specified by the Ada Reference Manual. When the dynamic model is in
26729 effect, GNAT makes the following assumptions:
26735 All code within all units in a partition is considered to be elaboration
26739 Some of the invocations in elaboration code may not take place at run time
26740 due to conditional execution.
26743 GNAT performs extensive diagnostics on a unit-by-unit basis for all scenarios
26744 that invoke internal targets. In addition, GNAT generates run-time checks for
26745 all external targets and for all scenarios that may exhibit ABE problems.
26747 The elaboration order is obtained by honoring all @emph{with} clauses, purity and
26748 preelaborability of units, and elaboration-control pragmas. The dynamic model
26749 attempts to take all invocations in elaboration code into account. If an
26750 invocation leads to a circularity, GNAT ignores the invocation based on the
26751 assumptions stated above. An order obtained using the dynamic model may fail
26752 an ABE check at run time when GNAT ignored an invocation.
26754 The dynamic model is enabled with compiler switch @code{-gnatE}.
26757 @geindex Static elaboration model
26763 @emph{Static elaboration model}
26765 This is the middle ground of the three models. When the static model is in
26766 effect, GNAT makes the following assumptions:
26772 Only code at the library level and in package body statements within all
26773 units in a partition is considered to be elaboration code.
26776 All invocations in elaboration will take place at run time, regardless of
26777 conditional execution.
26780 GNAT performs extensive diagnostics on a unit-by-unit basis for all scenarios
26781 that invoke internal targets. In addition, GNAT generates run-time checks for
26782 all external targets and for all scenarios that may exhibit ABE problems.
26784 The elaboration order is obtained by honoring all @emph{with} clauses, purity and
26785 preelaborability of units, presence of elaboration-control pragmas, and all
26786 invocations in elaboration code. An order obtained using the static model is
26787 guaranteed to be ABE problem-free, excluding dispatching calls and
26788 access-to-subprogram types.
26790 The static model is the default model in GNAT.
26793 @geindex SPARK elaboration model
26799 @emph{SPARK elaboration model}
26801 This is the most conservative of the three models and enforces the SPARK
26802 rules of elaboration as defined in the SPARK Reference Manual, section 7.7.
26803 The SPARK model is in effect only when a scenario and a target reside in a
26804 region subject to @code{SPARK_Mode On}, otherwise the dynamic or static model
26807 The SPARK model is enabled with compiler switch @code{-gnatd.v}.
26810 @geindex Legacy elaboration models
26816 @emph{Legacy elaboration models}
26818 In addition to the three elaboration models outlined above, GNAT provides the
26819 following legacy models:
26825 @cite{Legacy elaboration-checking model} available in pre-18.x versions of GNAT.
26826 This model is enabled with compiler switch @code{-gnatH}.
26829 @cite{Legacy elaboration-order model} available in pre-20.x versions of GNAT.
26830 This model is enabled with binder switch @code{-H}.
26834 @geindex Relaxed elaboration mode
26836 The dynamic, legacy, and static models can be relaxed using compiler switch
26837 @code{-gnatJ}, making them more permissive. Note that in this mode, GNAT
26838 may not diagnose certain elaboration issues or install run-time checks.
26840 @node Mixing Elaboration Models,ABE Diagnostics,Controlling the Elaboration Order in GNAT,Elaboration Order Handling in GNAT
26841 @anchor{gnat_ugn/elaboration_order_handling_in_gnat id7}@anchor{220}@anchor{gnat_ugn/elaboration_order_handling_in_gnat mixing-elaboration-models}@anchor{221}
26842 @section Mixing Elaboration Models
26845 It is possible to mix units compiled with a different elaboration model,
26846 however the following rules must be observed:
26852 A client unit compiled with the dynamic model can only @emph{with} a server unit
26853 that meets at least one of the following criteria:
26859 The server unit is compiled with the dynamic model.
26862 The server unit is a GNAT implementation unit from the @code{Ada}, @code{GNAT},
26863 @code{Interfaces}, or @code{System} hierarchies.
26866 The server unit has pragma @code{Pure} or @code{Preelaborate}.
26869 The client unit has an explicit @code{Elaborate_All} pragma for the server
26874 These rules ensure that elaboration checks are not omitted. If the rules are
26875 violated, the binder emits a warning:
26880 warning: "x.ads" has dynamic elaboration checks and with's
26881 warning: "y.ads" which has static elaboration checks
26885 The warnings can be suppressed by binder switch @code{-ws}.
26887 @node ABE Diagnostics,SPARK Diagnostics,Mixing Elaboration Models,Elaboration Order Handling in GNAT
26888 @anchor{gnat_ugn/elaboration_order_handling_in_gnat abe-diagnostics}@anchor{222}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id8}@anchor{223}
26889 @section ABE Diagnostics
26892 GNAT performs extensive diagnostics on a unit-by-unit basis for all scenarios
26893 that invoke internal targets, regardless of whether the dynamic, SPARK, or
26894 static model is in effect.
26896 Note that GNAT emits warnings rather than hard errors whenever it encounters an
26897 elaboration problem. This is because the elaboration model in effect may be too
26898 conservative, or a particular scenario may not be invoked due conditional
26899 execution. The warnings can be suppressed selectively with @code{pragma Warnings
26900 (Off)} or globally with compiler switch @code{-gnatwL}.
26902 A @emph{guaranteed ABE} arises when the body of a target is not elaborated early
26903 enough, and causes @emph{all} scenarios that directly invoke the target to fail.
26908 package body Guaranteed_ABE is
26909 function ABE return Integer;
26911 Val : constant Integer := ABE;
26913 function ABE return Integer is
26917 end Guaranteed_ABE;
26921 In the example above, the elaboration of @code{Guaranteed_ABE}’s body elaborates
26922 the declaration of @code{Val}. This invokes function @code{ABE}, however the body of
26923 @code{ABE} has not been elaborated yet. GNAT emits the following diagnostic:
26928 4. Val : constant Integer := ABE;
26930 >>> warning: cannot call "ABE" before body seen
26931 >>> warning: Program_Error will be raised at run time
26935 A @emph{conditional ABE} arises when the body of a target is not elaborated early
26936 enough, and causes @emph{some} scenarios that directly invoke the target to fail.
26941 1. package body Conditional_ABE is
26942 2. procedure Force_Body is null;
26945 5. with function Func return Integer;
26947 7. Val : constant Integer := Func;
26950 10. function ABE return Integer;
26952 12. function Cause_ABE return Boolean is
26953 13. package Inst is new Gen (ABE);
26958 18. Val : constant Boolean := Cause_ABE;
26960 20. function ABE return Integer is
26965 25. Safe : constant Boolean := Cause_ABE;
26966 26. end Conditional_ABE;
26970 In the example above, the elaboration of package body @code{Conditional_ABE}
26971 elaborates the declaration of @code{Val}. This invokes function @code{Cause_ABE},
26972 which instantiates generic unit @code{Gen} as @code{Inst}. The elaboration of
26973 @code{Inst} invokes function @code{ABE}, however the body of @code{ABE} has not been
26974 elaborated yet. GNAT emits the following diagnostic:
26979 13. package Inst is new Gen (ABE);
26981 >>> warning: in instantiation at line 7
26982 >>> warning: cannot call "ABE" before body seen
26983 >>> warning: Program_Error may be raised at run time
26984 >>> warning: body of unit "Conditional_ABE" elaborated
26985 >>> warning: function "Cause_ABE" called at line 18
26986 >>> warning: function "ABE" called at line 7, instance at line 13
26990 Note that the same ABE problem does not occur with the elaboration of
26991 declaration @code{Safe} because the body of function @code{ABE} has already been
26992 elaborated at that point.
26994 @node SPARK Diagnostics,Elaboration Circularities,ABE Diagnostics,Elaboration Order Handling in GNAT
26995 @anchor{gnat_ugn/elaboration_order_handling_in_gnat id9}@anchor{224}@anchor{gnat_ugn/elaboration_order_handling_in_gnat spark-diagnostics}@anchor{225}
26996 @section SPARK Diagnostics
26999 GNAT enforces the SPARK rules of elaboration as defined in the SPARK Reference
27000 Manual section 7.7 when compiler switch @code{-gnatd.v} is in effect. Note
27001 that GNAT emits hard errors whenever it encounters a violation of the SPARK
27008 2. package body SPARK_Diagnostics with SPARK_Mode is
27009 3. Val : constant Integer := Server.Func;
27011 >>> call to "Func" during elaboration in SPARK
27012 >>> unit "SPARK_Diagnostics" requires pragma "Elaborate_All" for "Server"
27013 >>> body of unit "SPARK_Model" elaborated
27014 >>> function "Func" called at line 3
27016 4. end SPARK_Diagnostics;
27020 @node Elaboration Circularities,Resolving Elaboration Circularities,SPARK Diagnostics,Elaboration Order Handling in GNAT
27021 @anchor{gnat_ugn/elaboration_order_handling_in_gnat elaboration-circularities}@anchor{226}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id10}@anchor{227}
27022 @section Elaboration Circularities
27025 An @strong{elaboration circularity} occurs whenever the elaboration of a set of
27026 units enters a deadlocked state, where each unit is waiting for another unit
27027 to be elaborated. This situation may be the result of improper use of @emph{with}
27028 clauses, elaboration-control pragmas, or invocations in elaboration code.
27030 The following example exhibits an elaboration circularity.
27035 with B; pragma Elaborate (B);
27042 procedure Force_Body;
27049 procedure Force_Body is null;
27051 Elab : constant Integer := C.Func;
27057 function Func return Integer;
27064 function Func return Integer is
27072 The binder emits the following diagnostic:
27077 error: Elaboration circularity detected
27081 info: unit "a (spec)" depends on its own elaboration
27085 info: unit "a (spec)" has with clause and pragma Elaborate for unit "b (spec)"
27086 info: unit "b (body)" is in the closure of pragma Elaborate
27087 info: unit "b (body)" invokes a construct of unit "c (body)" at elaboration time
27088 info: unit "c (body)" has with clause for unit "a (spec)"
27092 info: remove pragma Elaborate for unit "b (body)" in unit "a (spec)"
27093 info: use the dynamic elaboration model (compiler switch -gnatE)
27097 The diagnostic consist of the following sections:
27105 This section provides a short explanation describing why the set of units
27106 could not be ordered.
27111 This section enumerates the units comprising the deadlocked set, along with
27112 their interdependencies.
27117 This section enumerates various tactics for eliminating the circularity.
27120 @node Resolving Elaboration Circularities,Elaboration-related Compiler Switches,Elaboration Circularities,Elaboration Order Handling in GNAT
27121 @anchor{gnat_ugn/elaboration_order_handling_in_gnat id11}@anchor{228}@anchor{gnat_ugn/elaboration_order_handling_in_gnat resolving-elaboration-circularities}@anchor{229}
27122 @section Resolving Elaboration Circularities
27125 The most desirable option from the point of view of long-term maintenance is to
27126 rearrange the program so that the elaboration problems are avoided. One useful
27127 technique is to place the elaboration code into separate child packages.
27128 Another is to move some of the initialization code to explicitly invoked
27129 subprograms, where the program controls the order of initialization explicitly.
27130 Although this is the most desirable option, it may be impractical and involve
27131 too much modification, especially in the case of complex legacy code.
27133 When faced with an elaboration circularity, the programmer should also consider
27134 the tactics given in the suggestions section of the circularity diagnostic.
27135 Depending on the units involved in the circularity, their @emph{with} clauses,
27136 purity, preelaborability, presence of elaboration-control pragmas and
27137 invocations at elaboration time, the binder may suggest one or more of the
27138 following tactics to eliminate the circularity:
27144 Pragma Elaborate elimination
27147 remove pragma Elaborate for unit "..." in unit "..."
27150 This tactic is suggested when the binder has determined that pragma
27157 Prevents a set of units from being elaborated.
27160 The removal of the pragma will not eliminate the semantic effects of the
27161 pragma. In other words, the argument of the pragma will still be elaborated
27162 prior to the unit containing the pragma.
27165 The removal of the pragma will enable the successful ordering of the units.
27168 The programmer should remove the pragma as advised, and rebuild the program.
27171 Pragma Elaborate_All elimination
27174 remove pragma Elaborate_All for unit "..." in unit "..."
27177 This tactic is suggested when the binder has determined that pragma
27178 @code{Elaborate_All}:
27184 Prevents a set of units from being elaborated.
27187 The removal of the pragma will not eliminate the semantic effects of the
27188 pragma. In other words, the argument of the pragma along with its @emph{with}
27189 closure will still be elaborated prior to the unit containing the pragma.
27192 The removal of the pragma will enable the successful ordering of the units.
27195 The programmer should remove the pragma as advised, and rebuild the program.
27198 Pragma Elaborate_All downgrade
27201 change pragma Elaborate_All for unit "..." to Elaborate in unit "..."
27204 This tactic is always suggested with the pragma @code{Elaborate_All} elimination
27205 tactic. It offers a different alernative of guaranteeing that the argument of
27206 the pragma will still be elaborated prior to the unit containing the pragma.
27208 The programmer should update the pragma as advised, and rebuild the program.
27211 Pragma Elaborate_Body elimination
27214 remove pragma Elaborate_Body in unit "..."
27217 This tactic is suggested when the binder has determined that pragma
27218 @code{Elaborate_Body}:
27224 Prevents a set of units from being elaborated.
27227 The removal of the pragma will enable the successful ordering of the units.
27230 Note that the binder cannot determine whether the pragma is required for
27231 other purposes, such as guaranteeing the initialization of a variable
27232 declared in the spec by elaboration code in the body.
27234 The programmer should remove the pragma as advised, and rebuild the program.
27237 Use of pragma Restrictions
27240 use pragma Restrictions (No_Entry_Calls_In_Elaboration_Code)
27243 This tactic is suggested when the binder has determined that a task
27244 activation at elaboration time:
27250 Prevents a set of units from being elaborated.
27253 Note that the binder cannot determine with certainty whether the task will
27254 block at elaboration time.
27256 The programmer should create a configuration file, place the pragma within,
27257 update the general compilation arguments, and rebuild the program.
27260 Use of dynamic elaboration model
27263 use the dynamic elaboration model (compiler switch -gnatE)
27266 This tactic is suggested when the binder has determined that an invocation at
27273 Prevents a set of units from being elaborated.
27276 The use of the dynamic model will enable the successful ordering of the
27280 The programmer has two options:
27286 Determine the units involved in the invocation using the detailed
27287 invocation information, and add compiler switch @code{-gnatE} to the
27288 compilation arguments of selected files only. This approach will yield
27289 safer elaboration orders compared to the other option because it will
27290 minimize the opportunities presented to the dynamic model for ignoring
27294 Add compiler switch @code{-gnatE} to the general compilation arguments.
27298 Use of detailed invocation information
27301 use detailed invocation information (compiler switch -gnatd_F)
27304 This tactic is always suggested with the use of the dynamic model tactic. It
27305 causes the circularity section of the circularity diagnostic to describe the
27306 flow of elaboration code from a unit to a unit, enumerating all such paths in
27309 The programmer should analyze this information to determine which units
27310 should be compiled with the dynamic model.
27313 Forced-dependency elimination
27316 remove the dependency of unit "..." on unit "..." from the argument of switch -f
27319 This tactic is suggested when the binder has determined that a dependency
27320 present in the forced-elaboration-order file indicated by binder switch
27327 Prevents a set of units from being elaborated.
27330 The removal of the dependency will enable the successful ordering of the
27334 The programmer should edit the forced-elaboration-order file, remove the
27335 dependency, and rebind the program.
27338 All forced-dependency elimination
27344 This tactic is suggested in case editing the forced-elaboration-order file is
27347 The programmer should remove binder switch @code{-f} from the binder
27348 arguments, and rebind.
27351 Multiple-circularities diagnostic
27354 diagnose all circularities (binder switch -d_C)
27357 By default, the binder will diagnose only the highest-precedence circularity.
27358 If the program contains multiple circularities, the binder will suggest the
27359 use of binder switch @code{-d_C} in order to obtain the diagnostics of all
27362 The programmer should add binder switch @code{-d_C} to the binder
27363 arguments, and rebind.
27366 If none of the tactics suggested by the binder eliminate the elaboration
27367 circularity, the programmer should consider using one of the legacy elaboration
27368 models, in the following order:
27374 Use the pre-20.x legacy elaboration-order model, with binder switch
27378 Use both pre-18.x and pre-20.x legacy elaboration models, with compiler
27379 switch @code{-gnatH} and binder switch @code{-H}.
27382 Use the relaxed static-elaboration model, with compiler switches
27383 @code{-gnatH} @code{-gnatJ} and binder switch @code{-H}.
27386 Use the relaxed dynamic-elaboration model, with compiler switches
27387 @code{-gnatH} @code{-gnatJ} @code{-gnatE} and binder switch
27391 @node Elaboration-related Compiler Switches,Summary of Procedures for Elaboration Control,Resolving Elaboration Circularities,Elaboration Order Handling in GNAT
27392 @anchor{gnat_ugn/elaboration_order_handling_in_gnat elaboration-related-compiler-switches}@anchor{22a}@anchor{gnat_ugn/elaboration_order_handling_in_gnat id12}@anchor{22b}
27393 @section Elaboration-related Compiler Switches
27396 GNAT has several switches that affect the elaboration model and consequently
27397 the elaboration order chosen by the binder.
27399 @geindex -gnatE (gnat)
27404 @item @code{-gnatE}
27406 Dynamic elaboration checking mode enabled
27408 When this switch is in effect, GNAT activates the dynamic model.
27411 @geindex -gnatel (gnat)
27416 @item @code{-gnatel}
27418 Turn on info messages on generated Elaborate[_All] pragmas
27420 This switch is only applicable to the pre-20.x legacy elaboration models.
27421 The post-20.x elaboration model no longer relies on implicitly generated
27422 @code{Elaborate} and @code{Elaborate_All} pragmas to order units.
27424 When this switch is in effect, GNAT will emit the following supplementary
27425 information depending on the elaboration model in effect.
27431 @emph{Dynamic model}
27433 GNAT will indicate missing @code{Elaborate} and @code{Elaborate_All} pragmas for
27434 all library-level scenarios within the partition.
27437 @emph{Static model}
27439 GNAT will indicate all scenarios invoked during elaboration. In addition,
27440 it will provide detailed traceback when an implicit @code{Elaborate} or
27441 @code{Elaborate_All} pragma is generated.
27446 GNAT will indicate how an elaboration requirement is met by the context of
27447 a unit. This diagnostic requires compiler switch @code{-gnatd.v}.
27450 1. with Server; pragma Elaborate_All (Server);
27451 2. package Client with SPARK_Mode is
27452 3. Val : constant Integer := Server.Func;
27454 >>> info: call to "Func" during elaboration in SPARK
27455 >>> info: "Elaborate_All" requirement for unit "Server" met by pragma at line 1
27462 @geindex -gnatH (gnat)
27467 @item @code{-gnatH}
27469 Legacy elaboration checking mode enabled
27471 When this switch is in effect, GNAT will utilize the pre-18.x elaboration
27475 @geindex -gnatJ (gnat)
27480 @item @code{-gnatJ}
27482 Relaxed elaboration checking mode enabled
27484 When this switch is in effect, GNAT will not process certain scenarios,
27485 resulting in a more permissive elaboration model. Note that this may
27486 eliminate some diagnostics and run-time checks.
27489 @geindex -gnatw.f (gnat)
27494 @item @code{-gnatw.f}
27496 Turn on warnings for suspicious Subp’Access
27498 When this switch is in effect, GNAT will treat @code{'Access} of an entry,
27499 operator, or subprogram as a potential call to the target and issue warnings:
27502 1. package body Attribute_Call is
27503 2. function Func return Integer;
27504 3. type Func_Ptr is access function return Integer;
27506 5. Ptr : constant Func_Ptr := Func'Access;
27508 >>> warning: "Access" attribute of "Func" before body seen
27509 >>> warning: possible Program_Error on later references
27510 >>> warning: body of unit "Attribute_Call" elaborated
27511 >>> warning: "Access" of "Func" taken at line 5
27514 7. function Func return Integer is
27518 11. end Attribute_Call;
27521 In the example above, the elaboration of declaration @code{Ptr} is assigned
27522 @code{Func'Access} before the body of @code{Func} has been elaborated.
27525 @geindex -gnatwl (gnat)
27530 @item @code{-gnatwl}
27532 Turn on warnings for elaboration problems
27534 When this switch is in effect, GNAT emits diagnostics in the form of warnings
27535 concerning various elaboration problems. The warnings are enabled by default.
27536 The switch is provided in case all warnings are suppressed, but elaboration
27537 warnings are still desired.
27539 @item @code{-gnatwL}
27541 Turn off warnings for elaboration problems
27543 When this switch is in effect, GNAT no longer emits any diagnostics in the
27544 form of warnings. Selective suppression of elaboration problems is possible
27545 using @code{pragma Warnings (Off)}.
27548 1. package body Selective_Suppression is
27549 2. function ABE return Integer;
27551 4. Val_1 : constant Integer := ABE;
27553 >>> warning: cannot call "ABE" before body seen
27554 >>> warning: Program_Error will be raised at run time
27557 6. pragma Warnings (Off);
27558 7. Val_2 : constant Integer := ABE;
27559 8. pragma Warnings (On);
27561 10. function ABE return Integer is
27565 14. end Selective_Suppression;
27568 Note that suppressing elaboration warnings does not eliminate run-time
27569 checks. The example above will still fail at run time with an ABE.
27572 @node Summary of Procedures for Elaboration Control,Inspecting the Chosen Elaboration Order,Elaboration-related Compiler Switches,Elaboration Order Handling in GNAT
27573 @anchor{gnat_ugn/elaboration_order_handling_in_gnat id13}@anchor{22c}@anchor{gnat_ugn/elaboration_order_handling_in_gnat summary-of-procedures-for-elaboration-control}@anchor{22d}
27574 @section Summary of Procedures for Elaboration Control
27577 A programmer should first compile the program with the default options, using
27578 none of the binder or compiler switches. If the binder succeeds in finding an
27579 elaboration order, then apart from possible cases involing dispatching calls
27580 and access-to-subprogram types, the program is free of elaboration errors.
27582 If it is important for the program to be portable to compilers other than GNAT,
27583 then the programmer should use compiler switch @code{-gnatel} and consider
27584 the messages about missing or implicitly created @code{Elaborate} and
27585 @code{Elaborate_All} pragmas.
27587 If the binder reports an elaboration circularity, the programmer has several
27594 Ensure that elaboration warnings are enabled. This will allow the static
27595 model to output trace information of elaboration issues. The trace
27596 information could shed light on previously unforeseen dependencies, as well
27597 as their origins. Elaboration warnings are enabled with compiler switch
27601 Cosider the tactics given in the suggestions section of the circularity
27605 If none of the steps outlined above resolve the circularity, use a more
27606 permissive elaboration model, in the following order:
27612 Use the pre-20.x legacy elaboration-order model, with binder switch
27616 Use both pre-18.x and pre-20.x legacy elaboration models, with compiler
27617 switch @code{-gnatH} and binder switch @code{-H}.
27620 Use the relaxed static elaboration model, with compiler switches
27621 @code{-gnatH} @code{-gnatJ} and binder switch @code{-H}.
27624 Use the relaxed dynamic elaboration model, with compiler switches
27625 @code{-gnatH} @code{-gnatJ} @code{-gnatE} and binder switch
27630 @node Inspecting the Chosen Elaboration Order,,Summary of Procedures for Elaboration Control,Elaboration Order Handling in GNAT
27631 @anchor{gnat_ugn/elaboration_order_handling_in_gnat id14}@anchor{22e}@anchor{gnat_ugn/elaboration_order_handling_in_gnat inspecting-the-chosen-elaboration-order}@anchor{22f}
27632 @section Inspecting the Chosen Elaboration Order
27635 To see the elaboration order chosen by the binder, inspect the contents of file
27636 @cite{b~xxx.adb}. On certain targets, this file appears as @cite{b_xxx.adb}. The
27637 elaboration order appears as a sequence of calls to @code{Elab_Body} and
27638 @code{Elab_Spec}, interspersed with assignments to @cite{Exxx} which indicates that a
27639 particular unit is elaborated. For example:
27644 System.Soft_Links'Elab_Body;
27646 System.Secondary_Stack'Elab_Body;
27648 System.Exception_Table'Elab_Body;
27650 Ada.Io_Exceptions'Elab_Spec;
27652 Ada.Tags'Elab_Spec;
27653 Ada.Streams'Elab_Spec;
27655 Interfaces.C'Elab_Spec;
27657 System.Finalization_Root'Elab_Spec;
27659 System.Os_Lib'Elab_Body;
27661 System.Finalization_Implementation'Elab_Spec;
27662 System.Finalization_Implementation'Elab_Body;
27664 Ada.Finalization'Elab_Spec;
27666 Ada.Finalization.List_Controller'Elab_Spec;
27668 System.File_Control_Block'Elab_Spec;
27670 System.File_Io'Elab_Body;
27672 Ada.Tags'Elab_Body;
27674 Ada.Text_Io'Elab_Spec;
27675 Ada.Text_Io'Elab_Body;
27680 Note also binder switch @code{-l}, which outputs the chosen elaboration
27681 order and provides a more readable form of the above:
27689 system.case_util (spec)
27690 system.case_util (body)
27691 system.concat_2 (spec)
27692 system.concat_2 (body)
27693 system.concat_3 (spec)
27694 system.concat_3 (body)
27695 system.htable (spec)
27696 system.parameters (spec)
27697 system.parameters (body)
27699 interfaces.c_streams (spec)
27700 interfaces.c_streams (body)
27701 system.restrictions (spec)
27702 system.restrictions (body)
27703 system.standard_library (spec)
27704 system.exceptions (spec)
27705 system.exceptions (body)
27706 system.storage_elements (spec)
27707 system.storage_elements (body)
27708 system.secondary_stack (spec)
27709 system.stack_checking (spec)
27710 system.stack_checking (body)
27711 system.string_hash (spec)
27712 system.string_hash (body)
27713 system.htable (body)
27714 system.strings (spec)
27715 system.strings (body)
27716 system.traceback (spec)
27717 system.traceback (body)
27718 system.traceback_entries (spec)
27719 system.traceback_entries (body)
27720 ada.exceptions (spec)
27721 ada.exceptions.last_chance_handler (spec)
27722 system.soft_links (spec)
27723 system.soft_links (body)
27724 ada.exceptions.last_chance_handler (body)
27725 system.secondary_stack (body)
27726 system.exception_table (spec)
27727 system.exception_table (body)
27728 ada.io_exceptions (spec)
27731 interfaces.c (spec)
27732 interfaces.c (body)
27733 system.finalization_root (spec)
27734 system.finalization_root (body)
27735 system.memory (spec)
27736 system.memory (body)
27737 system.standard_library (body)
27738 system.os_lib (spec)
27739 system.os_lib (body)
27740 system.unsigned_types (spec)
27741 system.stream_attributes (spec)
27742 system.stream_attributes (body)
27743 system.finalization_implementation (spec)
27744 system.finalization_implementation (body)
27745 ada.finalization (spec)
27746 ada.finalization (body)
27747 ada.finalization.list_controller (spec)
27748 ada.finalization.list_controller (body)
27749 system.file_control_block (spec)
27750 system.file_io (spec)
27751 system.file_io (body)
27752 system.val_uns (spec)
27753 system.val_util (spec)
27754 system.val_util (body)
27755 system.val_uns (body)
27756 system.wch_con (spec)
27757 system.wch_con (body)
27758 system.wch_cnv (spec)
27759 system.wch_jis (spec)
27760 system.wch_jis (body)
27761 system.wch_cnv (body)
27762 system.wch_stw (spec)
27763 system.wch_stw (body)
27765 ada.exceptions (body)
27773 @node Inline Assembler,GNU Free Documentation License,Elaboration Order Handling in GNAT,Top
27774 @anchor{gnat_ugn/inline_assembler doc}@anchor{230}@anchor{gnat_ugn/inline_assembler id1}@anchor{231}@anchor{gnat_ugn/inline_assembler inline-assembler}@anchor{10}
27775 @chapter Inline Assembler
27778 @geindex Inline Assembler
27780 If you need to write low-level software that interacts directly
27781 with the hardware, Ada provides two ways to incorporate assembly
27782 language code into your program. First, you can import and invoke
27783 external routines written in assembly language, an Ada feature fully
27784 supported by GNAT. However, for small sections of code it may be simpler
27785 or more efficient to include assembly language statements directly
27786 in your Ada source program, using the facilities of the implementation-defined
27787 package @code{System.Machine_Code}, which incorporates the gcc
27788 Inline Assembler. The Inline Assembler approach offers a number of advantages,
27789 including the following:
27795 No need to use non-Ada tools
27798 Consistent interface over different targets
27801 Automatic usage of the proper calling conventions
27804 Access to Ada constants and variables
27807 Definition of intrinsic routines
27810 Possibility of inlining a subprogram comprising assembler code
27813 Code optimizer can take Inline Assembler code into account
27816 This appendix presents a series of examples to show you how to use
27817 the Inline Assembler. Although it focuses on the Intel x86,
27818 the general approach applies also to other processors.
27819 It is assumed that you are familiar with Ada
27820 and with assembly language programming.
27823 * Basic Assembler Syntax::
27824 * A Simple Example of Inline Assembler::
27825 * Output Variables in Inline Assembler::
27826 * Input Variables in Inline Assembler::
27827 * Inlining Inline Assembler Code::
27828 * Other Asm Functionality::
27832 @node Basic Assembler Syntax,A Simple Example of Inline Assembler,,Inline Assembler
27833 @anchor{gnat_ugn/inline_assembler basic-assembler-syntax}@anchor{232}@anchor{gnat_ugn/inline_assembler id2}@anchor{233}
27834 @section Basic Assembler Syntax
27837 The assembler used by GNAT and gcc is based not on the Intel assembly
27838 language, but rather on a language that descends from the AT&T Unix
27839 assembler @code{as} (and which is often referred to as ‘AT&T syntax’).
27840 The following table summarizes the main features of @code{as} syntax
27841 and points out the differences from the Intel conventions.
27842 See the gcc @code{as} and @code{gas} (an @code{as} macro
27843 pre-processor) documentation for further information.
27847 @emph{Register names}@w{ }
27849 gcc / @code{as}: Prefix with ‘%’; for example @code{%eax}@w{ }
27850 Intel: No extra punctuation; for example @code{eax}@w{ }
27858 @emph{Immediate operand}@w{ }
27860 gcc / @code{as}: Prefix with ‘$’; for example @code{$4}@w{ }
27861 Intel: No extra punctuation; for example @code{4}@w{ }
27869 @emph{Address}@w{ }
27871 gcc / @code{as}: Prefix with ‘$’; for example @code{$loc}@w{ }
27872 Intel: No extra punctuation; for example @code{loc}@w{ }
27880 @emph{Memory contents}@w{ }
27882 gcc / @code{as}: No extra punctuation; for example @code{loc}@w{ }
27883 Intel: Square brackets; for example @code{[loc]}@w{ }
27891 @emph{Register contents}@w{ }
27893 gcc / @code{as}: Parentheses; for example @code{(%eax)}@w{ }
27894 Intel: Square brackets; for example @code{[eax]}@w{ }
27902 @emph{Hexadecimal numbers}@w{ }
27904 gcc / @code{as}: Leading ‘0x’ (C language syntax); for example @code{0xA0}@w{ }
27905 Intel: Trailing ‘h’; for example @code{A0h}@w{ }
27913 @emph{Operand size}@w{ }
27915 gcc / @code{as}: Explicit in op code; for example @code{movw} to move a 16-bit word@w{ }
27916 Intel: Implicit, deduced by assembler; for example @code{mov}@w{ }
27924 @emph{Instruction repetition}@w{ }
27926 gcc / @code{as}: Split into two lines; for example@w{ }
27931 Intel: Keep on one line; for example @code{rep stosl}@w{ }
27939 @emph{Order of operands}@w{ }
27941 gcc / @code{as}: Source first; for example @code{movw $4, %eax}@w{ }
27942 Intel: Destination first; for example @code{mov eax, 4}@w{ }
27948 @node A Simple Example of Inline Assembler,Output Variables in Inline Assembler,Basic Assembler Syntax,Inline Assembler
27949 @anchor{gnat_ugn/inline_assembler a-simple-example-of-inline-assembler}@anchor{234}@anchor{gnat_ugn/inline_assembler id3}@anchor{235}
27950 @section A Simple Example of Inline Assembler
27953 The following example will generate a single assembly language statement,
27954 @code{nop}, which does nothing. Despite its lack of run-time effect,
27955 the example will be useful in illustrating the basics of
27956 the Inline Assembler facility.
27961 with System.Machine_Code; use System.Machine_Code;
27962 procedure Nothing is
27969 @code{Asm} is a procedure declared in package @code{System.Machine_Code};
27970 here it takes one parameter, a @emph{template string} that must be a static
27971 expression and that will form the generated instruction.
27972 @code{Asm} may be regarded as a compile-time procedure that parses
27973 the template string and additional parameters (none here),
27974 from which it generates a sequence of assembly language instructions.
27976 The examples in this chapter will illustrate several of the forms
27977 for invoking @code{Asm}; a complete specification of the syntax
27978 is found in the @code{Machine_Code_Insertions} section of the
27979 @cite{GNAT Reference Manual}.
27981 Under the standard GNAT conventions, the @code{Nothing} procedure
27982 should be in a file named @code{nothing.adb}.
27983 You can build the executable in the usual way:
27992 However, the interesting aspect of this example is not its run-time behavior
27993 but rather the generated assembly code.
27994 To see this output, invoke the compiler as follows:
27999 $ gcc -c -S -fomit-frame-pointer -gnatp nothing.adb
28003 where the options are:
28014 compile only (no bind or link)
28023 generate assembler listing
28030 @item @code{-fomit-frame-pointer}
28032 do not set up separate stack frames
28039 @item @code{-gnatp}
28041 do not add runtime checks
28045 This gives a human-readable assembler version of the code. The resulting
28046 file will have the same name as the Ada source file, but with a @code{.s}
28047 extension. In our example, the file @code{nothing.s} has the following
28053 .file "nothing.adb"
28055 ___gnu_compiled_ada:
28058 .globl __ada_nothing
28070 The assembly code you included is clearly indicated by
28071 the compiler, between the @code{#APP} and @code{#NO_APP}
28072 delimiters. The character before the ‘APP’ and ‘NOAPP’
28073 can differ on different targets. For example, GNU/Linux uses ‘#APP’ while
28074 on NT you will see ‘/APP’.
28076 If you make a mistake in your assembler code (such as using the
28077 wrong size modifier, or using a wrong operand for the instruction) GNAT
28078 will report this error in a temporary file, which will be deleted when
28079 the compilation is finished. Generating an assembler file will help
28080 in such cases, since you can assemble this file separately using the
28081 @code{as} assembler that comes with gcc.
28083 Assembling the file using the command
28092 will give you error messages whose lines correspond to the assembler
28093 input file, so you can easily find and correct any mistakes you made.
28094 If there are no errors, @code{as} will generate an object file
28095 @code{nothing.out}.
28097 @node Output Variables in Inline Assembler,Input Variables in Inline Assembler,A Simple Example of Inline Assembler,Inline Assembler
28098 @anchor{gnat_ugn/inline_assembler id4}@anchor{236}@anchor{gnat_ugn/inline_assembler output-variables-in-inline-assembler}@anchor{237}
28099 @section Output Variables in Inline Assembler
28102 The examples in this section, showing how to access the processor flags,
28103 illustrate how to specify the destination operands for assembly language
28109 with Interfaces; use Interfaces;
28110 with Ada.Text_IO; use Ada.Text_IO;
28111 with System.Machine_Code; use System.Machine_Code;
28112 procedure Get_Flags is
28113 Flags : Unsigned_32;
28116 Asm ("pushfl" & LF & HT & -- push flags on stack
28117 "popl %%eax" & LF & HT & -- load eax with flags
28118 "movl %%eax, %0", -- store flags in variable
28119 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28120 Put_Line ("Flags register:" & Flags'Img);
28125 In order to have a nicely aligned assembly listing, we have separated
28126 multiple assembler statements in the Asm template string with linefeed
28127 (ASCII.LF) and horizontal tab (ASCII.HT) characters.
28128 The resulting section of the assembly output file is:
28136 movl %eax, -40(%ebp)
28141 It would have been legal to write the Asm invocation as:
28146 Asm ("pushfl popl %%eax movl %%eax, %0")
28150 but in the generated assembler file, this would come out as:
28156 pushfl popl %eax movl %eax, -40(%ebp)
28161 which is not so convenient for the human reader.
28163 We use Ada comments
28164 at the end of each line to explain what the assembler instructions
28165 actually do. This is a useful convention.
28167 When writing Inline Assembler instructions, you need to precede each register
28168 and variable name with a percent sign. Since the assembler already requires
28169 a percent sign at the beginning of a register name, you need two consecutive
28170 percent signs for such names in the Asm template string, thus @code{%%eax}.
28171 In the generated assembly code, one of the percent signs will be stripped off.
28173 Names such as @code{%0}, @code{%1}, @code{%2}, etc., denote input or output
28174 variables: operands you later define using @code{Input} or @code{Output}
28175 parameters to @code{Asm}.
28176 An output variable is illustrated in
28177 the third statement in the Asm template string:
28186 The intent is to store the contents of the eax register in a variable that can
28187 be accessed in Ada. Simply writing @code{movl %%eax, Flags} would not
28188 necessarily work, since the compiler might optimize by using a register
28189 to hold Flags, and the expansion of the @code{movl} instruction would not be
28190 aware of this optimization. The solution is not to store the result directly
28191 but rather to advise the compiler to choose the correct operand form;
28192 that is the purpose of the @code{%0} output variable.
28194 Information about the output variable is supplied in the @code{Outputs}
28195 parameter to @code{Asm}:
28200 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28204 The output is defined by the @code{Asm_Output} attribute of the target type;
28205 the general format is
28210 Type'Asm_Output (constraint_string, variable_name)
28214 The constraint string directs the compiler how
28215 to store/access the associated variable. In the example
28220 Unsigned_32'Asm_Output ("=m", Flags);
28224 the @code{"m"} (memory) constraint tells the compiler that the variable
28225 @code{Flags} should be stored in a memory variable, thus preventing
28226 the optimizer from keeping it in a register. In contrast,
28231 Unsigned_32'Asm_Output ("=r", Flags);
28235 uses the @code{"r"} (register) constraint, telling the compiler to
28236 store the variable in a register.
28238 If the constraint is preceded by the equal character ‘=’, it tells
28239 the compiler that the variable will be used to store data into it.
28241 In the @code{Get_Flags} example, we used the @code{"g"} (global) constraint,
28242 allowing the optimizer to choose whatever it deems best.
28244 There are a fairly large number of constraints, but the ones that are
28245 most useful (for the Intel x86 processor) are the following:
28250 @multitable {xxxxxxxx} {xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx}
28265 global (i.e., can be stored anywhere)
28337 use one of eax, ebx, ecx or edx
28345 use one of eax, ebx, ecx, edx, esi or edi
28351 The full set of constraints is described in the gcc and @code{as}
28352 documentation; note that it is possible to combine certain constraints
28353 in one constraint string.
28355 You specify the association of an output variable with an assembler operand
28356 through the @code{%@emph{n}} notation, where @emph{n} is a non-negative
28362 Asm ("pushfl" & LF & HT & -- push flags on stack
28363 "popl %%eax" & LF & HT & -- load eax with flags
28364 "movl %%eax, %0", -- store flags in variable
28365 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28369 @code{%0} will be replaced in the expanded code by the appropriate operand,
28371 the compiler decided for the @code{Flags} variable.
28373 In general, you may have any number of output variables:
28379 Count the operands starting at 0; thus @code{%0}, @code{%1}, etc.
28382 Specify the @code{Outputs} parameter as a parenthesized comma-separated list
28383 of @code{Asm_Output} attributes
28391 Asm ("movl %%eax, %0" & LF & HT &
28392 "movl %%ebx, %1" & LF & HT &
28394 Outputs => (Unsigned_32'Asm_Output ("=g", Var_A), -- %0 = Var_A
28395 Unsigned_32'Asm_Output ("=g", Var_B), -- %1 = Var_B
28396 Unsigned_32'Asm_Output ("=g", Var_C))); -- %2 = Var_C
28400 where @code{Var_A}, @code{Var_B}, and @code{Var_C} are variables
28401 in the Ada program.
28403 As a variation on the @code{Get_Flags} example, we can use the constraints
28404 string to direct the compiler to store the eax register into the @code{Flags}
28405 variable, instead of including the store instruction explicitly in the
28406 @code{Asm} template string:
28411 with Interfaces; use Interfaces;
28412 with Ada.Text_IO; use Ada.Text_IO;
28413 with System.Machine_Code; use System.Machine_Code;
28414 procedure Get_Flags_2 is
28415 Flags : Unsigned_32;
28418 Asm ("pushfl" & LF & HT & -- push flags on stack
28419 "popl %%eax", -- save flags in eax
28420 Outputs => Unsigned_32'Asm_Output ("=a", Flags));
28421 Put_Line ("Flags register:" & Flags'Img);
28426 The @code{"a"} constraint tells the compiler that the @code{Flags}
28427 variable will come from the eax register. Here is the resulting code:
28436 movl %eax,-40(%ebp)
28440 The compiler generated the store of eax into Flags after
28441 expanding the assembler code.
28443 Actually, there was no need to pop the flags into the eax register;
28444 more simply, we could just pop the flags directly into the program variable:
28449 with Interfaces; use Interfaces;
28450 with Ada.Text_IO; use Ada.Text_IO;
28451 with System.Machine_Code; use System.Machine_Code;
28452 procedure Get_Flags_3 is
28453 Flags : Unsigned_32;
28456 Asm ("pushfl" & LF & HT & -- push flags on stack
28457 "pop %0", -- save flags in Flags
28458 Outputs => Unsigned_32'Asm_Output ("=g", Flags));
28459 Put_Line ("Flags register:" & Flags'Img);
28464 @node Input Variables in Inline Assembler,Inlining Inline Assembler Code,Output Variables in Inline Assembler,Inline Assembler
28465 @anchor{gnat_ugn/inline_assembler id5}@anchor{238}@anchor{gnat_ugn/inline_assembler input-variables-in-inline-assembler}@anchor{239}
28466 @section Input Variables in Inline Assembler
28469 The example in this section illustrates how to specify the source operands
28470 for assembly language statements.
28471 The program simply increments its input value by 1:
28476 with Interfaces; use Interfaces;
28477 with Ada.Text_IO; use Ada.Text_IO;
28478 with System.Machine_Code; use System.Machine_Code;
28479 procedure Increment is
28481 function Incr (Value : Unsigned_32) return Unsigned_32 is
28482 Result : Unsigned_32;
28485 Outputs => Unsigned_32'Asm_Output ("=a", Result),
28486 Inputs => Unsigned_32'Asm_Input ("a", Value));
28490 Value : Unsigned_32;
28494 Put_Line ("Value before is" & Value'Img);
28495 Value := Incr (Value);
28496 Put_Line ("Value after is" & Value'Img);
28501 The @code{Outputs} parameter to @code{Asm} specifies
28502 that the result will be in the eax register and that it is to be stored
28503 in the @code{Result} variable.
28505 The @code{Inputs} parameter looks much like the @code{Outputs} parameter,
28506 but with an @code{Asm_Input} attribute.
28507 The @code{"="} constraint, indicating an output value, is not present.
28509 You can have multiple input variables, in the same way that you can have more
28510 than one output variable.
28512 The parameter count (%0, %1) etc, still starts at the first output statement,
28513 and continues with the input statements.
28515 Just as the @code{Outputs} parameter causes the register to be stored into the
28516 target variable after execution of the assembler statements, so does the
28517 @code{Inputs} parameter cause its variable to be loaded into the register
28518 before execution of the assembler statements.
28520 Thus the effect of the @code{Asm} invocation is:
28526 load the 32-bit value of @code{Value} into eax
28529 execute the @code{incl %eax} instruction
28532 store the contents of eax into the @code{Result} variable
28535 The resulting assembler file (with @code{-O2} optimization) contains:
28540 _increment__incr.1:
28553 @node Inlining Inline Assembler Code,Other Asm Functionality,Input Variables in Inline Assembler,Inline Assembler
28554 @anchor{gnat_ugn/inline_assembler id6}@anchor{23a}@anchor{gnat_ugn/inline_assembler inlining-inline-assembler-code}@anchor{23b}
28555 @section Inlining Inline Assembler Code
28558 For a short subprogram such as the @code{Incr} function in the previous
28559 section, the overhead of the call and return (creating / deleting the stack
28560 frame) can be significant, compared to the amount of code in the subprogram
28561 body. A solution is to apply Ada’s @code{Inline} pragma to the subprogram,
28562 which directs the compiler to expand invocations of the subprogram at the
28563 point(s) of call, instead of setting up a stack frame for out-of-line calls.
28564 Here is the resulting program:
28569 with Interfaces; use Interfaces;
28570 with Ada.Text_IO; use Ada.Text_IO;
28571 with System.Machine_Code; use System.Machine_Code;
28572 procedure Increment_2 is
28574 function Incr (Value : Unsigned_32) return Unsigned_32 is
28575 Result : Unsigned_32;
28578 Outputs => Unsigned_32'Asm_Output ("=a", Result),
28579 Inputs => Unsigned_32'Asm_Input ("a", Value));
28582 pragma Inline (Increment);
28584 Value : Unsigned_32;
28588 Put_Line ("Value before is" & Value'Img);
28589 Value := Increment (Value);
28590 Put_Line ("Value after is" & Value'Img);
28595 Compile the program with both optimization (@code{-O2}) and inlining
28596 (@code{-gnatn}) enabled.
28598 The @code{Incr} function is still compiled as usual, but at the
28599 point in @code{Increment} where our function used to be called:
28605 call _increment__incr.1
28609 the code for the function body directly appears:
28622 thus saving the overhead of stack frame setup and an out-of-line call.
28624 @node Other Asm Functionality,,Inlining Inline Assembler Code,Inline Assembler
28625 @anchor{gnat_ugn/inline_assembler id7}@anchor{23c}@anchor{gnat_ugn/inline_assembler other-asm-functionality}@anchor{23d}
28626 @section Other @code{Asm} Functionality
28629 This section describes two important parameters to the @code{Asm}
28630 procedure: @code{Clobber}, which identifies register usage;
28631 and @code{Volatile}, which inhibits unwanted optimizations.
28634 * The Clobber Parameter::
28635 * The Volatile Parameter::
28639 @node The Clobber Parameter,The Volatile Parameter,,Other Asm Functionality
28640 @anchor{gnat_ugn/inline_assembler id8}@anchor{23e}@anchor{gnat_ugn/inline_assembler the-clobber-parameter}@anchor{23f}
28641 @subsection The @code{Clobber} Parameter
28644 One of the dangers of intermixing assembly language and a compiled language
28645 such as Ada is that the compiler needs to be aware of which registers are
28646 being used by the assembly code. In some cases, such as the earlier examples,
28647 the constraint string is sufficient to indicate register usage (e.g.,
28649 the eax register). But more generally, the compiler needs an explicit
28650 identification of the registers that are used by the Inline Assembly
28653 Using a register that the compiler doesn’t know about
28654 could be a side effect of an instruction (like @code{mull}
28655 storing its result in both eax and edx).
28656 It can also arise from explicit register usage in your
28657 assembly code; for example:
28662 Asm ("movl %0, %%ebx" & LF & HT &
28664 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
28665 Inputs => Unsigned_32'Asm_Input ("g", Var_In));
28669 where the compiler (since it does not analyze the @code{Asm} template string)
28670 does not know you are using the ebx register.
28672 In such cases you need to supply the @code{Clobber} parameter to @code{Asm},
28673 to identify the registers that will be used by your assembly code:
28678 Asm ("movl %0, %%ebx" & LF & HT &
28680 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
28681 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
28686 The Clobber parameter is a static string expression specifying the
28687 register(s) you are using. Note that register names are @emph{not} prefixed
28688 by a percent sign. Also, if more than one register is used then their names
28689 are separated by commas; e.g., @code{"eax, ebx"}
28691 The @code{Clobber} parameter has several additional uses:
28697 Use ‘register’ name @code{cc} to indicate that flags might have changed
28700 Use ‘register’ name @code{memory} if you changed a memory location
28703 @node The Volatile Parameter,,The Clobber Parameter,Other Asm Functionality
28704 @anchor{gnat_ugn/inline_assembler id9}@anchor{240}@anchor{gnat_ugn/inline_assembler the-volatile-parameter}@anchor{241}
28705 @subsection The @code{Volatile} Parameter
28708 @geindex Volatile parameter
28710 Compiler optimizations in the presence of Inline Assembler may sometimes have
28711 unwanted effects. For example, when an @code{Asm} invocation with an input
28712 variable is inside a loop, the compiler might move the loading of the input
28713 variable outside the loop, regarding it as a one-time initialization.
28715 If this effect is not desired, you can disable such optimizations by setting
28716 the @code{Volatile} parameter to @code{True}; for example:
28721 Asm ("movl %0, %%ebx" & LF & HT &
28723 Outputs => Unsigned_32'Asm_Output ("=g", Var_Out),
28724 Inputs => Unsigned_32'Asm_Input ("g", Var_In),
28730 By default, @code{Volatile} is set to @code{False} unless there is no
28731 @code{Outputs} parameter.
28733 Although setting @code{Volatile} to @code{True} prevents unwanted
28734 optimizations, it will also disable other optimizations that might be
28735 important for efficiency. In general, you should set @code{Volatile}
28736 to @code{True} only if the compiler’s optimizations have created
28739 @node GNU Free Documentation License,Index,Inline Assembler,Top
28740 @anchor{share/gnu_free_documentation_license doc}@anchor{242}@anchor{share/gnu_free_documentation_license gnu-fdl}@anchor{1}@anchor{share/gnu_free_documentation_license gnu-free-documentation-license}@anchor{243}
28741 @chapter GNU Free Documentation License
28744 Version 1.3, 3 November 2008
28746 Copyright 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc
28747 @indicateurl{http://fsf.org/}
28749 Everyone is permitted to copy and distribute verbatim copies of this
28750 license document, but changing it is not allowed.
28754 The purpose of this License is to make a manual, textbook, or other
28755 functional and useful document “free” in the sense of freedom: to
28756 assure everyone the effective freedom to copy and redistribute it,
28757 with or without modifying it, either commercially or noncommercially.
28758 Secondarily, this License preserves for the author and publisher a way
28759 to get credit for their work, while not being considered responsible
28760 for modifications made by others.
28762 This License is a kind of “copyleft”, which means that derivative
28763 works of the document must themselves be free in the same sense. It
28764 complements the GNU General Public License, which is a copyleft
28765 license designed for free software.
28767 We have designed this License in order to use it for manuals for free
28768 software, because free software needs free documentation: a free
28769 program should come with manuals providing the same freedoms that the
28770 software does. But this License is not limited to software manuals;
28771 it can be used for any textual work, regardless of subject matter or
28772 whether it is published as a printed book. We recommend this License
28773 principally for works whose purpose is instruction or reference.
28775 @strong{1. APPLICABILITY AND DEFINITIONS}
28777 This License applies to any manual or other work, in any medium, that
28778 contains a notice placed by the copyright holder saying it can be
28779 distributed under the terms of this License. Such a notice grants a
28780 world-wide, royalty-free license, unlimited in duration, to use that
28781 work under the conditions stated herein. The @strong{Document}, below,
28782 refers to any such manual or work. Any member of the public is a
28783 licensee, and is addressed as “@strong{you}”. You accept the license if you
28784 copy, modify or distribute the work in a way requiring permission
28785 under copyright law.
28787 A “@strong{Modified Version}” of the Document means any work containing the
28788 Document or a portion of it, either copied verbatim, or with
28789 modifications and/or translated into another language.
28791 A “@strong{Secondary Section}” is a named appendix or a front-matter section of
28792 the Document that deals exclusively with the relationship of the
28793 publishers or authors of the Document to the Document’s overall subject
28794 (or to related matters) and contains nothing that could fall directly
28795 within that overall subject. (Thus, if the Document is in part a
28796 textbook of mathematics, a Secondary Section may not explain any
28797 mathematics.) The relationship could be a matter of historical
28798 connection with the subject or with related matters, or of legal,
28799 commercial, philosophical, ethical or political position regarding
28802 The “@strong{Invariant Sections}” are certain Secondary Sections whose titles
28803 are designated, as being those of Invariant Sections, in the notice
28804 that says that the Document is released under this License. If a
28805 section does not fit the above definition of Secondary then it is not
28806 allowed to be designated as Invariant. The Document may contain zero
28807 Invariant Sections. If the Document does not identify any Invariant
28808 Sections then there are none.
28810 The “@strong{Cover Texts}” are certain short passages of text that are listed,
28811 as Front-Cover Texts or Back-Cover Texts, in the notice that says that
28812 the Document is released under this License. A Front-Cover Text may
28813 be at most 5 words, and a Back-Cover Text may be at most 25 words.
28815 A “@strong{Transparent}” copy of the Document means a machine-readable copy,
28816 represented in a format whose specification is available to the
28817 general public, that is suitable for revising the document
28818 straightforwardly with generic text editors or (for images composed of
28819 pixels) generic paint programs or (for drawings) some widely available
28820 drawing editor, and that is suitable for input to text formatters or
28821 for automatic translation to a variety of formats suitable for input
28822 to text formatters. A copy made in an otherwise Transparent file
28823 format whose markup, or absence of markup, has been arranged to thwart
28824 or discourage subsequent modification by readers is not Transparent.
28825 An image format is not Transparent if used for any substantial amount
28826 of text. A copy that is not “Transparent” is called @strong{Opaque}.
28828 Examples of suitable formats for Transparent copies include plain
28829 ASCII without markup, Texinfo input format, LaTeX input format, SGML
28830 or XML using a publicly available DTD, and standard-conforming simple
28831 HTML, PostScript or PDF designed for human modification. Examples of
28832 transparent image formats include PNG, XCF and JPG. Opaque formats
28833 include proprietary formats that can be read and edited only by
28834 proprietary word processors, SGML or XML for which the DTD and/or
28835 processing tools are not generally available, and the
28836 machine-generated HTML, PostScript or PDF produced by some word
28837 processors for output purposes only.
28839 The “@strong{Title Page}” means, for a printed book, the title page itself,
28840 plus such following pages as are needed to hold, legibly, the material
28841 this License requires to appear in the title page. For works in
28842 formats which do not have any title page as such, “Title Page” means
28843 the text near the most prominent appearance of the work’s title,
28844 preceding the beginning of the body of the text.
28846 The “@strong{publisher}” means any person or entity that distributes
28847 copies of the Document to the public.
28849 A section “@strong{Entitled XYZ}” means a named subunit of the Document whose
28850 title either is precisely XYZ or contains XYZ in parentheses following
28851 text that translates XYZ in another language. (Here XYZ stands for a
28852 specific section name mentioned below, such as “@strong{Acknowledgements}”,
28853 “@strong{Dedications}”, “@strong{Endorsements}”, or “@strong{History}”.)
28854 To “@strong{Preserve the Title}”
28855 of such a section when you modify the Document means that it remains a
28856 section “Entitled XYZ” according to this definition.
28858 The Document may include Warranty Disclaimers next to the notice which
28859 states that this License applies to the Document. These Warranty
28860 Disclaimers are considered to be included by reference in this
28861 License, but only as regards disclaiming warranties: any other
28862 implication that these Warranty Disclaimers may have is void and has
28863 no effect on the meaning of this License.
28865 @strong{2. VERBATIM COPYING}
28867 You may copy and distribute the Document in any medium, either
28868 commercially or noncommercially, provided that this License, the
28869 copyright notices, and the license notice saying this License applies
28870 to the Document are reproduced in all copies, and that you add no other
28871 conditions whatsoever to those of this License. You may not use
28872 technical measures to obstruct or control the reading or further
28873 copying of the copies you make or distribute. However, you may accept
28874 compensation in exchange for copies. If you distribute a large enough
28875 number of copies you must also follow the conditions in section 3.
28877 You may also lend copies, under the same conditions stated above, and
28878 you may publicly display copies.
28880 @strong{3. COPYING IN QUANTITY}
28882 If you publish printed copies (or copies in media that commonly have
28883 printed covers) of the Document, numbering more than 100, and the
28884 Document’s license notice requires Cover Texts, you must enclose the
28885 copies in covers that carry, clearly and legibly, all these Cover
28886 Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on
28887 the back cover. Both covers must also clearly and legibly identify
28888 you as the publisher of these copies. The front cover must present
28889 the full title with all words of the title equally prominent and
28890 visible. You may add other material on the covers in addition.
28891 Copying with changes limited to the covers, as long as they preserve
28892 the title of the Document and satisfy these conditions, can be treated
28893 as verbatim copying in other respects.
28895 If the required texts for either cover are too voluminous to fit
28896 legibly, you should put the first ones listed (as many as fit
28897 reasonably) on the actual cover, and continue the rest onto adjacent
28900 If you publish or distribute Opaque copies of the Document numbering
28901 more than 100, you must either include a machine-readable Transparent
28902 copy along with each Opaque copy, or state in or with each Opaque copy
28903 a computer-network location from which the general network-using
28904 public has access to download using public-standard network protocols
28905 a complete Transparent copy of the Document, free of added material.
28906 If you use the latter option, you must take reasonably prudent steps,
28907 when you begin distribution of Opaque copies in quantity, to ensure
28908 that this Transparent copy will remain thus accessible at the stated
28909 location until at least one year after the last time you distribute an
28910 Opaque copy (directly or through your agents or retailers) of that
28911 edition to the public.
28913 It is requested, but not required, that you contact the authors of the
28914 Document well before redistributing any large number of copies, to give
28915 them a chance to provide you with an updated version of the Document.
28917 @strong{4. MODIFICATIONS}
28919 You may copy and distribute a Modified Version of the Document under
28920 the conditions of sections 2 and 3 above, provided that you release
28921 the Modified Version under precisely this License, with the Modified
28922 Version filling the role of the Document, thus licensing distribution
28923 and modification of the Modified Version to whoever possesses a copy
28924 of it. In addition, you must do these things in the Modified Version:
28930 Use in the Title Page (and on the covers, if any) a title distinct
28931 from that of the Document, and from those of previous versions
28932 (which should, if there were any, be listed in the History section
28933 of the Document). You may use the same title as a previous version
28934 if the original publisher of that version gives permission.
28937 List on the Title Page, as authors, one or more persons or entities
28938 responsible for authorship of the modifications in the Modified
28939 Version, together with at least five of the principal authors of the
28940 Document (all of its principal authors, if it has fewer than five),
28941 unless they release you from this requirement.
28944 State on the Title page the name of the publisher of the
28945 Modified Version, as the publisher.
28948 Preserve all the copyright notices of the Document.
28951 Add an appropriate copyright notice for your modifications
28952 adjacent to the other copyright notices.
28955 Include, immediately after the copyright notices, a license notice
28956 giving the public permission to use the Modified Version under the
28957 terms of this License, in the form shown in the Addendum below.
28960 Preserve in that license notice the full lists of Invariant Sections
28961 and required Cover Texts given in the Document’s license notice.
28964 Include an unaltered copy of this License.
28967 Preserve the section Entitled “History”, Preserve its Title, and add
28968 to it an item stating at least the title, year, new authors, and
28969 publisher of the Modified Version as given on the Title Page. If
28970 there is no section Entitled “History” in the Document, create one
28971 stating the title, year, authors, and publisher of the Document as
28972 given on its Title Page, then add an item describing the Modified
28973 Version as stated in the previous sentence.
28976 Preserve the network location, if any, given in the Document for
28977 public access to a Transparent copy of the Document, and likewise
28978 the network locations given in the Document for previous versions
28979 it was based on. These may be placed in the “History” section.
28980 You may omit a network location for a work that was published at
28981 least four years before the Document itself, or if the original
28982 publisher of the version it refers to gives permission.
28985 For any section Entitled “Acknowledgements” or “Dedications”,
28986 Preserve the Title of the section, and preserve in the section all
28987 the substance and tone of each of the contributor acknowledgements
28988 and/or dedications given therein.
28991 Preserve all the Invariant Sections of the Document,
28992 unaltered in their text and in their titles. Section numbers
28993 or the equivalent are not considered part of the section titles.
28996 Delete any section Entitled “Endorsements”. Such a section
28997 may not be included in the Modified Version.
29000 Do not retitle any existing section to be Entitled “Endorsements”
29001 or to conflict in title with any Invariant Section.
29004 Preserve any Warranty Disclaimers.
29007 If the Modified Version includes new front-matter sections or
29008 appendices that qualify as Secondary Sections and contain no material
29009 copied from the Document, you may at your option designate some or all
29010 of these sections as invariant. To do this, add their titles to the
29011 list of Invariant Sections in the Modified Version’s license notice.
29012 These titles must be distinct from any other section titles.
29014 You may add a section Entitled “Endorsements”, provided it contains
29015 nothing but endorsements of your Modified Version by various
29016 parties—for example, statements of peer review or that the text has
29017 been approved by an organization as the authoritative definition of a
29020 You may add a passage of up to five words as a Front-Cover Text, and a
29021 passage of up to 25 words as a Back-Cover Text, to the end of the list
29022 of Cover Texts in the Modified Version. Only one passage of
29023 Front-Cover Text and one of Back-Cover Text may be added by (or
29024 through arrangements made by) any one entity. If the Document already
29025 includes a cover text for the same cover, previously added by you or
29026 by arrangement made by the same entity you are acting on behalf of,
29027 you may not add another; but you may replace the old one, on explicit
29028 permission from the previous publisher that added the old one.
29030 The author(s) and publisher(s) of the Document do not by this License
29031 give permission to use their names for publicity for or to assert or
29032 imply endorsement of any Modified Version.
29034 @strong{5. COMBINING DOCUMENTS}
29036 You may combine the Document with other documents released under this
29037 License, under the terms defined in section 4 above for modified
29038 versions, provided that you include in the combination all of the
29039 Invariant Sections of all of the original documents, unmodified, and
29040 list them all as Invariant Sections of your combined work in its
29041 license notice, and that you preserve all their Warranty Disclaimers.
29043 The combined work need only contain one copy of this License, and
29044 multiple identical Invariant Sections may be replaced with a single
29045 copy. If there are multiple Invariant Sections with the same name but
29046 different contents, make the title of each such section unique by
29047 adding at the end of it, in parentheses, the name of the original
29048 author or publisher of that section if known, or else a unique number.
29049 Make the same adjustment to the section titles in the list of
29050 Invariant Sections in the license notice of the combined work.
29052 In the combination, you must combine any sections Entitled “History”
29053 in the various original documents, forming one section Entitled
29054 “History”; likewise combine any sections Entitled “Acknowledgements”,
29055 and any sections Entitled “Dedications”. You must delete all sections
29056 Entitled “Endorsements”.
29058 @strong{6. COLLECTIONS OF DOCUMENTS}
29060 You may make a collection consisting of the Document and other documents
29061 released under this License, and replace the individual copies of this
29062 License in the various documents with a single copy that is included in
29063 the collection, provided that you follow the rules of this License for
29064 verbatim copying of each of the documents in all other respects.
29066 You may extract a single document from such a collection, and distribute
29067 it individually under this License, provided you insert a copy of this
29068 License into the extracted document, and follow this License in all
29069 other respects regarding verbatim copying of that document.
29071 @strong{7. AGGREGATION WITH INDEPENDENT WORKS}
29073 A compilation of the Document or its derivatives with other separate
29074 and independent documents or works, in or on a volume of a storage or
29075 distribution medium, is called an “aggregate” if the copyright
29076 resulting from the compilation is not used to limit the legal rights
29077 of the compilation’s users beyond what the individual works permit.
29078 When the Document is included in an aggregate, this License does not
29079 apply to the other works in the aggregate which are not themselves
29080 derivative works of the Document.
29082 If the Cover Text requirement of section 3 is applicable to these
29083 copies of the Document, then if the Document is less than one half of
29084 the entire aggregate, the Document’s Cover Texts may be placed on
29085 covers that bracket the Document within the aggregate, or the
29086 electronic equivalent of covers if the Document is in electronic form.
29087 Otherwise they must appear on printed covers that bracket the whole
29090 @strong{8. TRANSLATION}
29092 Translation is considered a kind of modification, so you may
29093 distribute translations of the Document under the terms of section 4.
29094 Replacing Invariant Sections with translations requires special
29095 permission from their copyright holders, but you may include
29096 translations of some or all Invariant Sections in addition to the
29097 original versions of these Invariant Sections. You may include a
29098 translation of this License, and all the license notices in the
29099 Document, and any Warranty Disclaimers, provided that you also include
29100 the original English version of this License and the original versions
29101 of those notices and disclaimers. In case of a disagreement between
29102 the translation and the original version of this License or a notice
29103 or disclaimer, the original version will prevail.
29105 If a section in the Document is Entitled “Acknowledgements”,
29106 “Dedications”, or “History”, the requirement (section 4) to Preserve
29107 its Title (section 1) will typically require changing the actual
29110 @strong{9. TERMINATION}
29112 You may not copy, modify, sublicense, or distribute the Document
29113 except as expressly provided under this License. Any attempt
29114 otherwise to copy, modify, sublicense, or distribute it is void, and
29115 will automatically terminate your rights under this License.
29117 However, if you cease all violation of this License, then your license
29118 from a particular copyright holder is reinstated (a) provisionally,
29119 unless and until the copyright holder explicitly and finally
29120 terminates your license, and (b) permanently, if the copyright holder
29121 fails to notify you of the violation by some reasonable means prior to
29122 60 days after the cessation.
29124 Moreover, your license from a particular copyright holder is
29125 reinstated permanently if the copyright holder notifies you of the
29126 violation by some reasonable means, this is the first time you have
29127 received notice of violation of this License (for any work) from that
29128 copyright holder, and you cure the violation prior to 30 days after
29129 your receipt of the notice.
29131 Termination of your rights under this section does not terminate the
29132 licenses of parties who have received copies or rights from you under
29133 this License. If your rights have been terminated and not permanently
29134 reinstated, receipt of a copy of some or all of the same material does
29135 not give you any rights to use it.
29137 @strong{10. FUTURE REVISIONS OF THIS LICENSE}
29139 The Free Software Foundation may publish new, revised versions
29140 of the GNU Free Documentation License from time to time. Such new
29141 versions will be similar in spirit to the present version, but may
29142 differ in detail to address new problems or concerns. See
29143 @indicateurl{http://www.gnu.org/copyleft/}.
29145 Each version of the License is given a distinguishing version number.
29146 If the Document specifies that a particular numbered version of this
29147 License “or any later version” applies to it, you have the option of
29148 following the terms and conditions either of that specified version or
29149 of any later version that has been published (not as a draft) by the
29150 Free Software Foundation. If the Document does not specify a version
29151 number of this License, you may choose any version ever published (not
29152 as a draft) by the Free Software Foundation. If the Document
29153 specifies that a proxy can decide which future versions of this
29154 License can be used, that proxy’s public statement of acceptance of a
29155 version permanently authorizes you to choose that version for the
29158 @strong{11. RELICENSING}
29160 “Massive Multiauthor Collaboration Site” (or “MMC Site”) means any
29161 World Wide Web server that publishes copyrightable works and also
29162 provides prominent facilities for anybody to edit those works. A
29163 public wiki that anybody can edit is an example of such a server. A
29164 “Massive Multiauthor Collaboration” (or “MMC”) contained in the
29165 site means any set of copyrightable works thus published on the MMC
29168 “CC-BY-SA” means the Creative Commons Attribution-Share Alike 3.0
29169 license published by Creative Commons Corporation, a not-for-profit
29170 corporation with a principal place of business in San Francisco,
29171 California, as well as future copyleft versions of that license
29172 published by that same organization.
29174 “Incorporate” means to publish or republish a Document, in whole or
29175 in part, as part of another Document.
29177 An MMC is “eligible for relicensing” if it is licensed under this
29178 License, and if all works that were first published under this License
29179 somewhere other than this MMC, and subsequently incorporated in whole
29180 or in part into the MMC, (1) had no cover texts or invariant sections,
29181 and (2) were thus incorporated prior to November 1, 2008.
29183 The operator of an MMC Site may republish an MMC contained in the site
29184 under CC-BY-SA on the same site at any time before August 1, 2009,
29185 provided the MMC is eligible for relicensing.
29187 @strong{ADDENDUM: How to use this License for your documents}
29189 To use this License in a document you have written, include a copy of
29190 the License in the document and put the following copyright and
29191 license notices just after the title page:
29195 Copyright © YEAR YOUR NAME.
29196 Permission is granted to copy, distribute and/or modify this document
29197 under the terms of the GNU Free Documentation License, Version 1.3
29198 or any later version published by the Free Software Foundation;
29199 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts.
29200 A copy of the license is included in the section entitled “GNU
29201 Free Documentation License”.
29204 If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts,
29205 replace the “with … Texts.” line with this:
29209 with the Invariant Sections being LIST THEIR TITLES, with the
29210 Front-Cover Texts being LIST, and with the Back-Cover Texts being LIST.
29213 If you have Invariant Sections without Cover Texts, or some other
29214 combination of the three, merge those two alternatives to suit the
29217 If your document contains nontrivial examples of program code, we
29218 recommend releasing these examples in parallel under your choice of
29219 free software license, such as the GNU General Public License,
29220 to permit their use in free software.
29222 @node Index,,GNU Free Documentation License,Top
29229 @anchor{gnat_ugn/gnat_utility_programs switches-related-to-project-files}@w{ }