1 \input texinfo @c -*-texinfo-*-
5 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
7 @c GNAT DOCUMENTATION o
11 @c Copyright (C) 1995-2004 Free Software Foundation o
14 @c GNAT is maintained by Ada Core Technologies Inc (http://www.gnat.com). o
16 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
18 @setfilename gnat_rm.info
22 @settitle GNAT Reference Manual
24 @setchapternewpage odd
27 @include gcc-common.texi
29 @dircategory GNU Ada tools
31 * GNAT Reference Manual: (gnat_rm). Reference Manual for GNU Ada tools.
35 Copyright @copyright{} 1995-2004, Free Software Foundation
37 Permission is granted to copy, distribute and/or modify this document
38 under the terms of the GNU Free Documentation License, Version 1.2
39 or any later version published by the Free Software Foundation;
40 with the Invariant Sections being ``GNU Free Documentation License'',
41 with the Front-Cover Texts being ``GNAT Reference Manual'', and with
42 no Back-Cover Texts. A copy of the license is included in the section
43 entitled ``GNU Free Documentation License''.
48 @title GNAT Reference Manual
49 @subtitle GNAT, The GNU Ada 95 Compiler
50 @subtitle GCC version @value{version-GCC}
51 @author Ada Core Technologies, Inc.
54 @vskip 0pt plus 1filll
61 @node Top, About This Guide, (dir), (dir)
62 @top GNAT Reference Manual
68 GNAT, The GNU Ada 95 Compiler@*
69 GCC version @value{version-GCC}@*
72 Ada Core Technologies, Inc.
76 * Implementation Defined Pragmas::
77 * Implementation Defined Attributes::
78 * Implementation Advice::
79 * Implementation Defined Characteristics::
80 * Intrinsic Subprograms::
81 * Representation Clauses and Pragmas::
82 * Standard Library Routines::
83 * The Implementation of Standard I/O::
85 * Interfacing to Other Languages::
86 * Specialized Needs Annexes::
87 * Implementation of Specific Ada Features::
88 * Project File Reference::
89 * Obsolescent Features::
90 * GNU Free Documentation License::
93 --- The Detailed Node Listing ---
97 * What This Reference Manual Contains::
98 * Related Information::
100 Implementation Defined Pragmas
102 * Pragma Abort_Defer::
108 * Pragma C_Pass_By_Copy::
110 * Pragma Common_Object::
111 * Pragma Compile_Time_Warning::
112 * Pragma Complex_Representation::
113 * Pragma Component_Alignment::
114 * Pragma Convention_Identifier::
116 * Pragma CPP_Constructor::
117 * Pragma CPP_Virtual::
118 * Pragma CPP_Vtable::
120 * Pragma Elaboration_Checks::
122 * Pragma Export_Exception::
123 * Pragma Export_Function::
124 * Pragma Export_Object::
125 * Pragma Export_Procedure::
126 * Pragma Export_Value::
127 * Pragma Export_Valued_Procedure::
128 * Pragma Extend_System::
130 * Pragma External_Name_Casing::
131 * Pragma Finalize_Storage_Only::
132 * Pragma Float_Representation::
134 * Pragma Import_Exception::
135 * Pragma Import_Function::
136 * Pragma Import_Object::
137 * Pragma Import_Procedure::
138 * Pragma Import_Valued_Procedure::
139 * Pragma Initialize_Scalars::
140 * Pragma Inline_Always::
141 * Pragma Inline_Generic::
143 * Pragma Interface_Name::
144 * Pragma Interrupt_Handler::
145 * Pragma Interrupt_State::
146 * Pragma Keep_Names::
149 * Pragma Linker_Alias::
150 * Pragma Linker_Section::
151 * Pragma Long_Float::
152 * Pragma Machine_Attribute::
153 * Pragma Main_Storage::
155 * Pragma Normalize_Scalars::
156 * Pragma Obsolescent::
159 * Pragma Profile (Ravenscar)::
160 * Pragma Profile (Restricted)::
161 * Pragma Propagate_Exceptions::
162 * Pragma Psect_Object::
163 * Pragma Pure_Function::
164 * Pragma Restriction_Warnings::
165 * Pragma Source_File_Name::
166 * Pragma Source_File_Name_Project::
167 * Pragma Source_Reference::
168 * Pragma Stream_Convert::
169 * Pragma Style_Checks::
171 * Pragma Suppress_All::
172 * Pragma Suppress_Exception_Locations::
173 * Pragma Suppress_Initialization::
176 * Pragma Task_Storage::
177 * Pragma Thread_Body::
178 * Pragma Time_Slice::
180 * Pragma Unchecked_Union::
181 * Pragma Unimplemented_Unit::
182 * Pragma Universal_Data::
183 * Pragma Unreferenced::
184 * Pragma Unreserve_All_Interrupts::
185 * Pragma Unsuppress::
186 * Pragma Use_VADS_Size::
187 * Pragma Validity_Checks::
190 * Pragma Weak_External::
192 Implementation Defined Attributes
202 * Default_Bit_Order::
210 * Has_Access_Values::
211 * Has_Discriminants::
217 * Max_Interrupt_Priority::
219 * Maximum_Alignment::
223 * Passed_By_Reference::
234 * Unconstrained_Array::
235 * Universal_Literal_String::
236 * Unrestricted_Access::
242 The Implementation of Standard I/O
244 * Standard I/O Packages::
253 * Operations on C Streams::
254 * Interfacing to C Streams::
258 * Ada.Characters.Latin_9 (a-chlat9.ads)::
259 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
260 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
261 * Ada.Command_Line.Remove (a-colire.ads)::
262 * Ada.Command_Line.Environment (a-colien.ads)::
263 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
264 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
265 * Ada.Exceptions.Traceback (a-exctra.ads)::
266 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
267 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
268 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
269 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
270 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
271 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
272 * GNAT.Array_Split (g-arrspl.ads)::
273 * GNAT.AWK (g-awk.ads)::
274 * GNAT.Bounded_Buffers (g-boubuf.ads)::
275 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
276 * GNAT.Bubble_Sort (g-bubsor.ads)::
277 * GNAT.Bubble_Sort_A (g-busora.ads)::
278 * GNAT.Bubble_Sort_G (g-busorg.ads)::
279 * GNAT.Calendar (g-calend.ads)::
280 * GNAT.Calendar.Time_IO (g-catiio.ads)::
281 * GNAT.Case_Util (g-casuti.ads)::
282 * GNAT.CGI (g-cgi.ads)::
283 * GNAT.CGI.Cookie (g-cgicoo.ads)::
284 * GNAT.CGI.Debug (g-cgideb.ads)::
285 * GNAT.Command_Line (g-comlin.ads)::
286 * GNAT.Compiler_Version (g-comver.ads)::
287 * GNAT.Ctrl_C (g-ctrl_c.ads)::
288 * GNAT.CRC32 (g-crc32.ads)::
289 * GNAT.Current_Exception (g-curexc.ads)::
290 * GNAT.Debug_Pools (g-debpoo.ads)::
291 * GNAT.Debug_Utilities (g-debuti.ads)::
292 * GNAT.Directory_Operations (g-dirope.ads)::
293 * GNAT.Dynamic_HTables (g-dynhta.ads)::
294 * GNAT.Dynamic_Tables (g-dyntab.ads)::
295 * GNAT.Exception_Actions (g-excact.ads)::
296 * GNAT.Exception_Traces (g-exctra.ads)::
297 * GNAT.Exceptions (g-except.ads)::
298 * GNAT.Expect (g-expect.ads)::
299 * GNAT.Float_Control (g-flocon.ads)::
300 * GNAT.Heap_Sort (g-heasor.ads)::
301 * GNAT.Heap_Sort_A (g-hesora.ads)::
302 * GNAT.Heap_Sort_G (g-hesorg.ads)::
303 * GNAT.HTable (g-htable.ads)::
304 * GNAT.IO (g-io.ads)::
305 * GNAT.IO_Aux (g-io_aux.ads)::
306 * GNAT.Lock_Files (g-locfil.ads)::
307 * GNAT.MD5 (g-md5.ads)::
308 * GNAT.Memory_Dump (g-memdum.ads)::
309 * GNAT.Most_Recent_Exception (g-moreex.ads)::
310 * GNAT.OS_Lib (g-os_lib.ads)::
311 * GNAT.Perfect_Hash.Generators (g-pehage.ads)::
312 * GNAT.Regexp (g-regexp.ads)::
313 * GNAT.Registry (g-regist.ads)::
314 * GNAT.Regpat (g-regpat.ads)::
315 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
316 * GNAT.Semaphores (g-semaph.ads)::
317 * GNAT.Signals (g-signal.ads)::
318 * GNAT.Sockets (g-socket.ads)::
319 * GNAT.Source_Info (g-souinf.ads)::
320 * GNAT.Spell_Checker (g-speche.ads)::
321 * GNAT.Spitbol.Patterns (g-spipat.ads)::
322 * GNAT.Spitbol (g-spitbo.ads)::
323 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
324 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
325 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
326 * GNAT.Strings (g-string.ads)::
327 * GNAT.String_Split (g-strspl.ads)::
328 * GNAT.Table (g-table.ads)::
329 * GNAT.Task_Lock (g-tasloc.ads)::
330 * GNAT.Threads (g-thread.ads)::
331 * GNAT.Traceback (g-traceb.ads)::
332 * GNAT.Traceback.Symbolic (g-trasym.ads)::
333 * GNAT.Wide_String_Split (g-wistsp.ads)::
334 * Interfaces.C.Extensions (i-cexten.ads)::
335 * Interfaces.C.Streams (i-cstrea.ads)::
336 * Interfaces.CPP (i-cpp.ads)::
337 * Interfaces.Os2lib (i-os2lib.ads)::
338 * Interfaces.Os2lib.Errors (i-os2err.ads)::
339 * Interfaces.Os2lib.Synchronization (i-os2syn.ads)::
340 * Interfaces.Os2lib.Threads (i-os2thr.ads)::
341 * Interfaces.Packed_Decimal (i-pacdec.ads)::
342 * Interfaces.VxWorks (i-vxwork.ads)::
343 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
344 * System.Address_Image (s-addima.ads)::
345 * System.Assertions (s-assert.ads)::
346 * System.Memory (s-memory.ads)::
347 * System.Partition_Interface (s-parint.ads)::
348 * System.Restrictions (s-restri.ads)::
349 * System.Rident (s-rident.ads)::
350 * System.Task_Info (s-tasinf.ads)::
351 * System.Wch_Cnv (s-wchcnv.ads)::
352 * System.Wch_Con (s-wchcon.ads)::
356 * Text_IO Stream Pointer Positioning::
357 * Text_IO Reading and Writing Non-Regular Files::
359 * Treating Text_IO Files as Streams::
360 * Text_IO Extensions::
361 * Text_IO Facilities for Unbounded Strings::
365 * Wide_Text_IO Stream Pointer Positioning::
366 * Wide_Text_IO Reading and Writing Non-Regular Files::
368 Interfacing to Other Languages
371 * Interfacing to C++::
372 * Interfacing to COBOL::
373 * Interfacing to Fortran::
374 * Interfacing to non-GNAT Ada code::
376 Specialized Needs Annexes
378 Implementation of Specific Ada Features
379 * Machine Code Insertions::
380 * GNAT Implementation of Tasking::
381 * GNAT Implementation of Shared Passive Packages::
382 * Code Generation for Array Aggregates::
383 * The Size of Discriminated Records with Default Discriminants::
385 Project File Reference
389 GNU Free Documentation License
396 @node About This Guide
397 @unnumbered About This Guide
401 This manual contains useful information in writing programs using the
402 GNAT compiler. It includes information on implementation dependent
403 characteristics of GNAT, including all the information required by Annex
409 This manual contains useful information in writing programs using the
410 GNAT Pro compiler. It includes information on implementation dependent
411 characteristics of GNAT Pro, including all the information required by Annex
415 Ada 95 is designed to be highly portable.
416 In general, a program will have the same effect even when compiled by
417 different compilers on different platforms.
418 However, since Ada 95 is designed to be used in a
419 wide variety of applications, it also contains a number of system
420 dependent features to be used in interfacing to the external world.
421 @cindex Implementation-dependent features
424 Note: Any program that makes use of implementation-dependent features
425 may be non-portable. You should follow good programming practice and
426 isolate and clearly document any sections of your program that make use
427 of these features in a non-portable manner.
430 For ease of exposition, ``GNAT Pro'' will be referred to simply as
431 ``GNAT'' in the remainder of this document.
435 * What This Reference Manual Contains::
437 * Related Information::
440 @node What This Reference Manual Contains
441 @unnumberedsec What This Reference Manual Contains
444 This reference manual contains the following chapters:
448 @ref{Implementation Defined Pragmas}, lists GNAT implementation-dependent
449 pragmas, which can be used to extend and enhance the functionality of the
453 @ref{Implementation Defined Attributes}, lists GNAT
454 implementation-dependent attributes which can be used to extend and
455 enhance the functionality of the compiler.
458 @ref{Implementation Advice}, provides information on generally
459 desirable behavior which are not requirements that all compilers must
460 follow since it cannot be provided on all systems, or which may be
461 undesirable on some systems.
464 @ref{Implementation Defined Characteristics}, provides a guide to
465 minimizing implementation dependent features.
468 @ref{Intrinsic Subprograms}, describes the intrinsic subprograms
469 implemented by GNAT, and how they can be imported into user
470 application programs.
473 @ref{Representation Clauses and Pragmas}, describes in detail the
474 way that GNAT represents data, and in particular the exact set
475 of representation clauses and pragmas that is accepted.
478 @ref{Standard Library Routines}, provides a listing of packages and a
479 brief description of the functionality that is provided by Ada's
480 extensive set of standard library routines as implemented by GNAT@.
483 @ref{The Implementation of Standard I/O}, details how the GNAT
484 implementation of the input-output facilities.
487 @ref{The GNAT Library}, is a catalog of packages that complement
488 the Ada predefined library.
491 @ref{Interfacing to Other Languages}, describes how programs
492 written in Ada using GNAT can be interfaced to other programming
495 @ref{Specialized Needs Annexes}, describes the GNAT implementation of all
496 of the specialized needs annexes.
499 @ref{Implementation of Specific Ada Features}, discusses issues related
500 to GNAT's implementation of machine code insertions, tasking, and several
504 @ref{Project File Reference}, presents the syntax and semantics
508 @ref{Obsolescent Features} documents implementation dependent features,
509 including pragmas and attributes, which are considered obsolescent, since
510 there are other preferred ways of achieving the same results. These
511 obsolescent forms are retained for backwards compatibilty.
515 @cindex Ada 95 ISO/ANSI Standard
517 This reference manual assumes that you are familiar with Ada 95
518 language, as described in the International Standard
519 ANSI/ISO/IEC-8652:1995, Jan 1995.
522 @unnumberedsec Conventions
523 @cindex Conventions, typographical
524 @cindex Typographical conventions
527 Following are examples of the typographical and graphic conventions used
532 @code{Functions}, @code{utility program names}, @code{standard names},
539 @file{File Names}, @samp{button names}, and @samp{field names}.
548 [optional information or parameters]
551 Examples are described by text
553 and then shown this way.
558 Commands that are entered by the user are preceded in this manual by the
559 characters @samp{$ } (dollar sign followed by space). If your system uses this
560 sequence as a prompt, then the commands will appear exactly as you see them
561 in the manual. If your system uses some other prompt, then the command will
562 appear with the @samp{$} replaced by whatever prompt character you are using.
564 @node Related Information
565 @unnumberedsec Related Information
567 See the following documents for further information on GNAT:
571 @cite{GNAT User's Guide}, which provides information on how to use
572 the GNAT compiler system.
575 @cite{Ada 95 Reference Manual}, which contains all reference
576 material for the Ada 95 programming language.
579 @cite{Ada 95 Annotated Reference Manual}, which is an annotated version
580 of the standard reference manual cited above. The annotations describe
581 detailed aspects of the design decision, and in particular contain useful
582 sections on Ada 83 compatibility.
585 @cite{DEC Ada, Technical Overview and Comparison on DIGITAL Platforms},
586 which contains specific information on compatibility between GNAT and
590 @cite{DEC Ada, Language Reference Manual, part number AA-PYZAB-TK} which
591 describes in detail the pragmas and attributes provided by the DEC Ada 83
596 @node Implementation Defined Pragmas
597 @chapter Implementation Defined Pragmas
600 Ada 95 defines a set of pragmas that can be used to supply additional
601 information to the compiler. These language defined pragmas are
602 implemented in GNAT and work as described in the Ada 95 Reference
605 In addition, Ada 95 allows implementations to define additional pragmas
606 whose meaning is defined by the implementation. GNAT provides a number
607 of these implementation-dependent pragmas which can be used to extend
608 and enhance the functionality of the compiler. This section of the GNAT
609 Reference Manual describes these additional pragmas.
611 Note that any program using these pragmas may not be portable to other
612 compilers (although GNAT implements this set of pragmas on all
613 platforms). Therefore if portability to other compilers is an important
614 consideration, the use of these pragmas should be minimized.
617 * Pragma Abort_Defer::
623 * Pragma C_Pass_By_Copy::
625 * Pragma Common_Object::
626 * Pragma Compile_Time_Warning::
627 * Pragma Complex_Representation::
628 * Pragma Component_Alignment::
629 * Pragma Convention_Identifier::
631 * Pragma CPP_Constructor::
632 * Pragma CPP_Virtual::
633 * Pragma CPP_Vtable::
635 * Pragma Elaboration_Checks::
637 * Pragma Export_Exception::
638 * Pragma Export_Function::
639 * Pragma Export_Object::
640 * Pragma Export_Procedure::
641 * Pragma Export_Value::
642 * Pragma Export_Valued_Procedure::
643 * Pragma Extend_System::
645 * Pragma External_Name_Casing::
646 * Pragma Finalize_Storage_Only::
647 * Pragma Float_Representation::
649 * Pragma Import_Exception::
650 * Pragma Import_Function::
651 * Pragma Import_Object::
652 * Pragma Import_Procedure::
653 * Pragma Import_Valued_Procedure::
654 * Pragma Initialize_Scalars::
655 * Pragma Inline_Always::
656 * Pragma Inline_Generic::
658 * Pragma Interface_Name::
659 * Pragma Interrupt_Handler::
660 * Pragma Interrupt_State::
661 * Pragma Keep_Names::
664 * Pragma Linker_Alias::
665 * Pragma Linker_Section::
666 * Pragma Long_Float::
667 * Pragma Machine_Attribute::
668 * Pragma Main_Storage::
670 * Pragma Normalize_Scalars::
671 * Pragma Obsolescent::
674 * Pragma Profile (Ravenscar)::
675 * Pragma Profile (Restricted)::
676 * Pragma Propagate_Exceptions::
677 * Pragma Psect_Object::
678 * Pragma Pure_Function::
679 * Pragma Restriction_Warnings::
680 * Pragma Source_File_Name::
681 * Pragma Source_File_Name_Project::
682 * Pragma Source_Reference::
683 * Pragma Stream_Convert::
684 * Pragma Style_Checks::
686 * Pragma Suppress_All::
687 * Pragma Suppress_Exception_Locations::
688 * Pragma Suppress_Initialization::
691 * Pragma Task_Storage::
692 * Pragma Thread_Body::
693 * Pragma Time_Slice::
695 * Pragma Unchecked_Union::
696 * Pragma Unimplemented_Unit::
697 * Pragma Universal_Data::
698 * Pragma Unreferenced::
699 * Pragma Unreserve_All_Interrupts::
700 * Pragma Unsuppress::
701 * Pragma Use_VADS_Size::
702 * Pragma Validity_Checks::
705 * Pragma Weak_External::
708 @node Pragma Abort_Defer
709 @unnumberedsec Pragma Abort_Defer
711 @cindex Deferring aborts
719 This pragma must appear at the start of the statement sequence of a
720 handled sequence of statements (right after the @code{begin}). It has
721 the effect of deferring aborts for the sequence of statements (but not
722 for the declarations or handlers, if any, associated with this statement
726 @unnumberedsec Pragma Ada_83
735 A configuration pragma that establishes Ada 83 mode for the unit to
736 which it applies, regardless of the mode set by the command line
737 switches. In Ada 83 mode, GNAT attempts to be as compatible with
738 the syntax and semantics of Ada 83, as defined in the original Ada
739 83 Reference Manual as possible. In particular, the new Ada 95
740 keywords are not recognized, optional package bodies are allowed,
741 and generics may name types with unknown discriminants without using
742 the @code{(<>)} notation. In addition, some but not all of the additional
743 restrictions of Ada 83 are enforced.
745 Ada 83 mode is intended for two purposes. Firstly, it allows existing
746 legacy Ada 83 code to be compiled and adapted to GNAT with less effort.
747 Secondly, it aids in keeping code backwards compatible with Ada 83.
748 However, there is no guarantee that code that is processed correctly
749 by GNAT in Ada 83 mode will in fact compile and execute with an Ada
750 83 compiler, since GNAT does not enforce all the additional checks
754 @unnumberedsec Pragma Ada_95
763 A configuration pragma that establishes Ada 95 mode for the unit to which
764 it applies, regardless of the mode set by the command line switches.
765 This mode is set automatically for the @code{Ada} and @code{System}
766 packages and their children, so you need not specify it in these
767 contexts. This pragma is useful when writing a reusable component that
768 itself uses Ada 95 features, but which is intended to be usable from
769 either Ada 83 or Ada 95 programs.
771 @node Pragma Annotate
772 @unnumberedsec Pragma Annotate
777 pragma Annotate (IDENTIFIER @{, ARG@});
779 ARG ::= NAME | EXPRESSION
783 This pragma is used to annotate programs. @var{identifier} identifies
784 the type of annotation. GNAT verifies this is an identifier, but does
785 not otherwise analyze it. The @var{arg} argument
786 can be either a string literal or an
787 expression. String literals are assumed to be of type
788 @code{Standard.String}. Names of entities are simply analyzed as entity
789 names. All other expressions are analyzed as expressions, and must be
792 The analyzed pragma is retained in the tree, but not otherwise processed
793 by any part of the GNAT compiler. This pragma is intended for use by
794 external tools, including ASIS@.
797 @unnumberedsec Pragma Assert
804 [, static_string_EXPRESSION]);
808 The effect of this pragma depends on whether the corresponding command
809 line switch is set to activate assertions. The pragma expands into code
810 equivalent to the following:
813 if assertions-enabled then
814 if not boolean_EXPRESSION then
815 System.Assertions.Raise_Assert_Failure
822 The string argument, if given, is the message that will be associated
823 with the exception occurrence if the exception is raised. If no second
824 argument is given, the default message is @samp{@var{file}:@var{nnn}},
825 where @var{file} is the name of the source file containing the assert,
826 and @var{nnn} is the line number of the assert. A pragma is not a
827 statement, so if a statement sequence contains nothing but a pragma
828 assert, then a null statement is required in addition, as in:
833 pragma Assert (K > 3, "Bad value for K");
839 Note that, as with the @code{if} statement to which it is equivalent, the
840 type of the expression is either @code{Standard.Boolean}, or any type derived
841 from this standard type.
843 If assertions are disabled (switch @code{-gnata} not used), then there
844 is no effect (and in particular, any side effects from the expression
845 are suppressed). More precisely it is not quite true that the pragma
846 has no effect, since the expression is analyzed, and may cause types
847 to be frozen if they are mentioned here for the first time.
849 If assertions are enabled, then the given expression is tested, and if
850 it is @code{False} then @code{System.Assertions.Raise_Assert_Failure} is called
851 which results in the raising of @code{Assert_Failure} with the given message.
853 If the boolean expression has side effects, these side effects will turn
854 on and off with the setting of the assertions mode, resulting in
855 assertions that have an effect on the program. You should generally
856 avoid side effects in the expression arguments of this pragma. However,
857 the expressions are analyzed for semantic correctness whether or not
858 assertions are enabled, so turning assertions on and off cannot affect
859 the legality of a program.
861 @node Pragma Ast_Entry
862 @unnumberedsec Pragma Ast_Entry
868 pragma AST_Entry (entry_IDENTIFIER);
872 This pragma is implemented only in the OpenVMS implementation of GNAT@. The
873 argument is the simple name of a single entry; at most one @code{AST_Entry}
874 pragma is allowed for any given entry. This pragma must be used in
875 conjunction with the @code{AST_Entry} attribute, and is only allowed after
876 the entry declaration and in the same task type specification or single task
877 as the entry to which it applies. This pragma specifies that the given entry
878 may be used to handle an OpenVMS asynchronous system trap (@code{AST})
879 resulting from an OpenVMS system service call. The pragma does not affect
880 normal use of the entry. For further details on this pragma, see the
881 DEC Ada Language Reference Manual, section 9.12a.
883 @node Pragma C_Pass_By_Copy
884 @unnumberedsec Pragma C_Pass_By_Copy
885 @cindex Passing by copy
886 @findex C_Pass_By_Copy
890 pragma C_Pass_By_Copy
891 ([Max_Size =>] static_integer_EXPRESSION);
895 Normally the default mechanism for passing C convention records to C
896 convention subprograms is to pass them by reference, as suggested by RM
897 B.3(69). Use the configuration pragma @code{C_Pass_By_Copy} to change
898 this default, by requiring that record formal parameters be passed by
899 copy if all of the following conditions are met:
903 The size of the record type does not exceed@*@var{static_integer_expression}.
905 The record type has @code{Convention C}.
907 The formal parameter has this record type, and the subprogram has a
908 foreign (non-Ada) convention.
912 If these conditions are met the argument is passed by copy, i.e.@: in a
913 manner consistent with what C expects if the corresponding formal in the
914 C prototype is a struct (rather than a pointer to a struct).
916 You can also pass records by copy by specifying the convention
917 @code{C_Pass_By_Copy} for the record type, or by using the extended
918 @code{Import} and @code{Export} pragmas, which allow specification of
919 passing mechanisms on a parameter by parameter basis.
922 @unnumberedsec Pragma Comment
928 pragma Comment (static_string_EXPRESSION);
932 This is almost identical in effect to pragma @code{Ident}. It allows the
933 placement of a comment into the object file and hence into the
934 executable file if the operating system permits such usage. The
935 difference is that @code{Comment}, unlike @code{Ident}, has
936 no limitations on placement of the pragma (it can be placed
937 anywhere in the main source unit), and if more than one pragma
938 is used, all comments are retained.
940 @node Pragma Common_Object
941 @unnumberedsec Pragma Common_Object
942 @findex Common_Object
947 pragma Common_Object (
948 [Internal =>] LOCAL_NAME,
949 [, [External =>] EXTERNAL_SYMBOL]
950 [, [Size =>] EXTERNAL_SYMBOL] );
954 | static_string_EXPRESSION
958 This pragma enables the shared use of variables stored in overlaid
959 linker areas corresponding to the use of @code{COMMON}
960 in Fortran. The single
961 object @var{local_name} is assigned to the area designated by
962 the @var{External} argument.
963 You may define a record to correspond to a series
964 of fields. The @var{size} argument
965 is syntax checked in GNAT, but otherwise ignored.
967 @code{Common_Object} is not supported on all platforms. If no
968 support is available, then the code generator will issue a message
969 indicating that the necessary attribute for implementation of this
970 pragma is not available.
972 @node Pragma Compile_Time_Warning
973 @unnumberedsec Pragma Compile_Time_Warning
974 @findex Compile_Time_Warning
979 pragma Compile_Time_Warning
980 (boolean_EXPRESSION, static_string_EXPRESSION);
984 This pragma can be used to generate additional compile time warnings. It
985 is particularly useful in generics, where warnings can be issued for
986 specific problematic instantiations. The first parameter is a boolean
987 expression. The pragma is effective only if the value of this expression
988 is known at compile time, and has the value True. The set of expressions
989 whose values are known at compile time includes all static boolean
990 expressions, and also other values which the compiler can determine
991 at compile time (e.g. the size of a record type set by an explicit
992 size representation clause, or the value of a variable which was
993 initialized to a constant and is known not to have been modified).
994 If these conditions are met, a warning message is generated using
995 the value given as the second argument. This string value may contain
996 embedded ASCII.LF characters to break the message into multiple lines.
998 @node Pragma Complex_Representation
999 @unnumberedsec Pragma Complex_Representation
1000 @findex Complex_Representation
1004 @smallexample @c ada
1005 pragma Complex_Representation
1006 ([Entity =>] LOCAL_NAME);
1010 The @var{Entity} argument must be the name of a record type which has
1011 two fields of the same floating-point type. The effect of this pragma is
1012 to force gcc to use the special internal complex representation form for
1013 this record, which may be more efficient. Note that this may result in
1014 the code for this type not conforming to standard ABI (application
1015 binary interface) requirements for the handling of record types. For
1016 example, in some environments, there is a requirement for passing
1017 records by pointer, and the use of this pragma may result in passing
1018 this type in floating-point registers.
1020 @node Pragma Component_Alignment
1021 @unnumberedsec Pragma Component_Alignment
1022 @cindex Alignments of components
1023 @findex Component_Alignment
1027 @smallexample @c ada
1028 pragma Component_Alignment (
1029 [Form =>] ALIGNMENT_CHOICE
1030 [, [Name =>] type_LOCAL_NAME]);
1032 ALIGNMENT_CHOICE ::=
1040 Specifies the alignment of components in array or record types.
1041 The meaning of the @var{Form} argument is as follows:
1044 @findex Component_Size
1045 @item Component_Size
1046 Aligns scalar components and subcomponents of the array or record type
1047 on boundaries appropriate to their inherent size (naturally
1048 aligned). For example, 1-byte components are aligned on byte boundaries,
1049 2-byte integer components are aligned on 2-byte boundaries, 4-byte
1050 integer components are aligned on 4-byte boundaries and so on. These
1051 alignment rules correspond to the normal rules for C compilers on all
1052 machines except the VAX@.
1054 @findex Component_Size_4
1055 @item Component_Size_4
1056 Naturally aligns components with a size of four or fewer
1057 bytes. Components that are larger than 4 bytes are placed on the next
1060 @findex Storage_Unit
1062 Specifies that array or record components are byte aligned, i.e.@:
1063 aligned on boundaries determined by the value of the constant
1064 @code{System.Storage_Unit}.
1068 Specifies that array or record components are aligned on default
1069 boundaries, appropriate to the underlying hardware or operating system or
1070 both. For OpenVMS VAX systems, the @code{Default} choice is the same as
1071 the @code{Storage_Unit} choice (byte alignment). For all other systems,
1072 the @code{Default} choice is the same as @code{Component_Size} (natural
1077 If the @code{Name} parameter is present, @var{type_local_name} must
1078 refer to a local record or array type, and the specified alignment
1079 choice applies to the specified type. The use of
1080 @code{Component_Alignment} together with a pragma @code{Pack} causes the
1081 @code{Component_Alignment} pragma to be ignored. The use of
1082 @code{Component_Alignment} together with a record representation clause
1083 is only effective for fields not specified by the representation clause.
1085 If the @code{Name} parameter is absent, the pragma can be used as either
1086 a configuration pragma, in which case it applies to one or more units in
1087 accordance with the normal rules for configuration pragmas, or it can be
1088 used within a declarative part, in which case it applies to types that
1089 are declared within this declarative part, or within any nested scope
1090 within this declarative part. In either case it specifies the alignment
1091 to be applied to any record or array type which has otherwise standard
1094 If the alignment for a record or array type is not specified (using
1095 pragma @code{Pack}, pragma @code{Component_Alignment}, or a record rep
1096 clause), the GNAT uses the default alignment as described previously.
1098 @node Pragma Convention_Identifier
1099 @unnumberedsec Pragma Convention_Identifier
1100 @findex Convention_Identifier
1101 @cindex Conventions, synonyms
1105 @smallexample @c ada
1106 pragma Convention_Identifier (
1107 [Name =>] IDENTIFIER,
1108 [Convention =>] convention_IDENTIFIER);
1112 This pragma provides a mechanism for supplying synonyms for existing
1113 convention identifiers. The @code{Name} identifier can subsequently
1114 be used as a synonym for the given convention in other pragmas (including
1115 for example pragma @code{Import} or another @code{Convention_Identifier}
1116 pragma). As an example of the use of this, suppose you had legacy code
1117 which used Fortran77 as the identifier for Fortran. Then the pragma:
1119 @smallexample @c ada
1120 pragma Convention_Identifier (Fortran77, Fortran);
1124 would allow the use of the convention identifier @code{Fortran77} in
1125 subsequent code, avoiding the need to modify the sources. As another
1126 example, you could use this to parametrize convention requirements
1127 according to systems. Suppose you needed to use @code{Stdcall} on
1128 windows systems, and @code{C} on some other system, then you could
1129 define a convention identifier @code{Library} and use a single
1130 @code{Convention_Identifier} pragma to specify which convention
1131 would be used system-wide.
1133 @node Pragma CPP_Class
1134 @unnumberedsec Pragma CPP_Class
1136 @cindex Interfacing with C++
1140 @smallexample @c ada
1141 pragma CPP_Class ([Entity =>] LOCAL_NAME);
1145 The argument denotes an entity in the current declarative region
1146 that is declared as a tagged or untagged record type. It indicates that
1147 the type corresponds to an externally declared C++ class type, and is to
1148 be laid out the same way that C++ would lay out the type.
1150 If (and only if) the type is tagged, at least one component in the
1151 record must be of type @code{Interfaces.CPP.Vtable_Ptr}, corresponding
1152 to the C++ Vtable (or Vtables in the case of multiple inheritance) used
1155 Types for which @code{CPP_Class} is specified do not have assignment or
1156 equality operators defined (such operations can be imported or declared
1157 as subprograms as required). Initialization is allowed only by
1158 constructor functions (see pragma @code{CPP_Constructor}).
1160 Pragma @code{CPP_Class} is intended primarily for automatic generation
1161 using an automatic binding generator tool.
1162 See @ref{Interfacing to C++} for related information.
1164 @node Pragma CPP_Constructor
1165 @unnumberedsec Pragma CPP_Constructor
1166 @cindex Interfacing with C++
1167 @findex CPP_Constructor
1171 @smallexample @c ada
1172 pragma CPP_Constructor ([Entity =>] LOCAL_NAME);
1176 This pragma identifies an imported function (imported in the usual way
1177 with pragma @code{Import}) as corresponding to a C++
1178 constructor. The argument is a name that must have been
1179 previously mentioned in a pragma @code{Import}
1180 with @code{Convention} = @code{CPP}, and must be of one of the following
1185 @code{function @var{Fname} return @var{T}'Class}
1188 @code{function @var{Fname} (@dots{}) return @var{T}'Class}
1192 where @var{T} is a tagged type to which the pragma @code{CPP_Class} applies.
1194 The first form is the default constructor, used when an object of type
1195 @var{T} is created on the Ada side with no explicit constructor. Other
1196 constructors (including the copy constructor, which is simply a special
1197 case of the second form in which the one and only argument is of type
1198 @var{T}), can only appear in two contexts:
1202 On the right side of an initialization of an object of type @var{T}.
1204 In an extension aggregate for an object of a type derived from @var{T}.
1208 Although the constructor is described as a function that returns a value
1209 on the Ada side, it is typically a procedure with an extra implicit
1210 argument (the object being initialized) at the implementation
1211 level. GNAT issues the appropriate call, whatever it is, to get the
1212 object properly initialized.
1214 In the case of derived objects, you may use one of two possible forms
1215 for declaring and creating an object:
1218 @item @code{New_Object : Derived_T}
1219 @item @code{New_Object : Derived_T := (@var{constructor-call with} @dots{})}
1223 In the first case the default constructor is called and extension fields
1224 if any are initialized according to the default initialization
1225 expressions in the Ada declaration. In the second case, the given
1226 constructor is called and the extension aggregate indicates the explicit
1227 values of the extension fields.
1229 If no constructors are imported, it is impossible to create any objects
1230 on the Ada side. If no default constructor is imported, only the
1231 initialization forms using an explicit call to a constructor are
1234 Pragma @code{CPP_Constructor} is intended primarily for automatic generation
1235 using an automatic binding generator tool.
1236 See @ref{Interfacing to C++} for more related information.
1238 @node Pragma CPP_Virtual
1239 @unnumberedsec Pragma CPP_Virtual
1240 @cindex Interfacing to C++
1245 @smallexample @c ada
1248 [, [Vtable_Ptr =>] vtable_ENTITY,]
1249 [, [Position =>] static_integer_EXPRESSION]);
1253 This pragma serves the same function as pragma @code{Import} in that
1254 case of a virtual function imported from C++. The @var{Entity} argument
1256 primitive subprogram of a tagged type to which pragma @code{CPP_Class}
1257 applies. The @var{Vtable_Ptr} argument specifies
1258 the Vtable_Ptr component which contains the
1259 entry for this virtual function. The @var{Position} argument
1260 is the sequential number
1261 counting virtual functions for this Vtable starting at 1.
1263 The @code{Vtable_Ptr} and @code{Position} arguments may be omitted if
1264 there is one Vtable_Ptr present (single inheritance case) and all
1265 virtual functions are imported. In that case the compiler can deduce both
1268 No @code{External_Name} or @code{Link_Name} arguments are required for a
1269 virtual function, since it is always accessed indirectly via the
1270 appropriate Vtable entry.
1272 Pragma @code{CPP_Virtual} is intended primarily for automatic generation
1273 using an automatic binding generator tool.
1274 See @ref{Interfacing to C++} for related information.
1276 @node Pragma CPP_Vtable
1277 @unnumberedsec Pragma CPP_Vtable
1278 @cindex Interfacing with C++
1283 @smallexample @c ada
1286 [Vtable_Ptr =>] vtable_ENTITY,
1287 [Entry_Count =>] static_integer_EXPRESSION);
1291 Given a record to which the pragma @code{CPP_Class} applies,
1292 this pragma can be specified for each component of type
1293 @code{CPP.Interfaces.Vtable_Ptr}.
1294 @var{Entity} is the tagged type, @var{Vtable_Ptr}
1295 is the record field of type @code{Vtable_Ptr}, and @var{Entry_Count} is
1296 the number of virtual functions on the C++ side. Not all of these
1297 functions need to be imported on the Ada side.
1299 You may omit the @code{CPP_Vtable} pragma if there is only one
1300 @code{Vtable_Ptr} component in the record and all virtual functions are
1301 imported on the Ada side (the default value for the entry count in this
1302 case is simply the total number of virtual functions).
1304 Pragma @code{CPP_Vtable} is intended primarily for automatic generation
1305 using an automatic binding generator tool.
1306 See @ref{Interfacing to C++} for related information.
1309 @unnumberedsec Pragma Debug
1314 @smallexample @c ada
1315 pragma Debug (PROCEDURE_CALL_WITHOUT_SEMICOLON);
1317 PROCEDURE_CALL_WITHOUT_SEMICOLON ::=
1319 | PROCEDURE_PREFIX ACTUAL_PARAMETER_PART
1323 The argument has the syntactic form of an expression, meeting the
1324 syntactic requirements for pragmas.
1326 If assertions are not enabled on the command line, this pragma has no
1327 effect. If asserts are enabled, the semantics of the pragma is exactly
1328 equivalent to the procedure call statement corresponding to the argument
1329 with a terminating semicolon. Pragmas are permitted in sequences of
1330 declarations, so you can use pragma @code{Debug} to intersperse calls to
1331 debug procedures in the middle of declarations.
1333 @node Pragma Elaboration_Checks
1334 @unnumberedsec Pragma Elaboration_Checks
1335 @cindex Elaboration control
1336 @findex Elaboration_Checks
1340 @smallexample @c ada
1341 pragma Elaboration_Checks (Dynamic | Static);
1345 This is a configuration pragma that provides control over the
1346 elaboration model used by the compilation affected by the
1347 pragma. If the parameter is @code{Dynamic},
1348 then the dynamic elaboration
1349 model described in the Ada Reference Manual is used, as though
1350 the @code{-gnatE} switch had been specified on the command
1351 line. If the parameter is @code{Static}, then the default GNAT static
1352 model is used. This configuration pragma overrides the setting
1353 of the command line. For full details on the elaboration models
1354 used by the GNAT compiler, see section ``Elaboration Order
1355 Handling in GNAT'' in the @cite{GNAT User's Guide}.
1357 @node Pragma Eliminate
1358 @unnumberedsec Pragma Eliminate
1359 @cindex Elimination of unused subprograms
1364 @smallexample @c ada
1366 [Unit_Name =>] IDENTIFIER |
1367 SELECTED_COMPONENT);
1370 [Unit_Name =>] IDENTIFIER |
1372 [Entity =>] IDENTIFIER |
1373 SELECTED_COMPONENT |
1375 [,OVERLOADING_RESOLUTION]);
1377 OVERLOADING_RESOLUTION ::= PARAMETER_AND_RESULT_TYPE_PROFILE |
1380 PARAMETER_AND_RESULT_TYPE_PROFILE ::= PROCEDURE_PROFILE |
1383 PROCEDURE_PROFILE ::= Parameter_Types => PARAMETER_TYPES
1385 FUNCTION_PROFILE ::= [Parameter_Types => PARAMETER_TYPES,]
1386 Result_Type => result_SUBTYPE_NAME]
1388 PARAMETER_TYPES ::= (SUBTYPE_NAME @{, SUBTYPE_NAME@})
1389 SUBTYPE_NAME ::= STRING_VALUE
1391 SOURCE_LOCATION ::= Source_Location => SOURCE_TRACE
1392 SOURCE_TRACE ::= STRING_VALUE
1394 STRING_VALUE ::= STRING_LITERAL @{& STRING_LITERAL@}
1398 This pragma indicates that the given entity is not used outside the
1399 compilation unit it is defined in. The entity must be an explicitly declared
1400 subprogram; this includes generic subprogram instances and
1401 subprograms declared in generic package instances.
1403 If the entity to be eliminated is a library level subprogram, then
1404 the first form of pragma @code{Eliminate} is used with only a single argument.
1405 In this form, the @code{Unit_Name} argument specifies the name of the
1406 library level unit to be eliminated.
1408 In all other cases, both @code{Unit_Name} and @code{Entity} arguments
1409 are required. If item is an entity of a library package, then the first
1410 argument specifies the unit name, and the second argument specifies
1411 the particular entity. If the second argument is in string form, it must
1412 correspond to the internal manner in which GNAT stores entity names (see
1413 compilation unit Namet in the compiler sources for details).
1415 The remaining parameters (OVERLOADING_RESOLUTION) are optionally used
1416 to distinguish between overloaded subprograms. If a pragma does not contain
1417 the OVERLOADING_RESOLUTION parameter(s), it is applied to all the overloaded
1418 subprograms denoted by the first two parameters.
1420 Use PARAMETER_AND_RESULT_TYPE_PROFILE to specify the profile of the subprogram
1421 to be eliminated in a manner similar to that used for the extended
1422 @code{Import} and @code{Export} pragmas, except that the subtype names are
1423 always given as strings. At the moment, this form of distinguishing
1424 overloaded subprograms is implemented only partially, so we do not recommend
1425 using it for practical subprogram elimination.
1427 Note, that in case of a parameterless procedure its profile is represented
1428 as @code{Parameter_Types => ("")}
1430 Alternatively, the @code{Source_Location} parameter is used to specify
1431 which overloaded alternative is to be eliminated by pointing to the
1432 location of the DEFINING_PROGRAM_UNIT_NAME of this subprogram in the
1433 source text. The string literal (or concatenation of string literals)
1434 given as SOURCE_TRACE must have the following format:
1436 @smallexample @c ada
1437 SOURCE_TRACE ::= SOURCE_LOCATION@{LBRACKET SOURCE_LOCATION RBRACKET@}
1442 SOURCE_LOCATION ::= FILE_NAME:LINE_NUMBER
1443 FILE_NAME ::= STRING_LITERAL
1444 LINE_NUMBER ::= DIGIT @{DIGIT@}
1447 SOURCE_TRACE should be the short name of the source file (with no directory
1448 information), and LINE_NUMBER is supposed to point to the line where the
1449 defining name of the subprogram is located.
1451 For the subprograms that are not a part of generic instantiations, only one
1452 SOURCE_LOCATION is used. If a subprogram is declared in a package
1453 instantiation, SOURCE_TRACE contains two SOURCE_LOCATIONs, the first one is
1454 the location of the (DEFINING_PROGRAM_UNIT_NAME of the) instantiation, and the
1455 second one denotes the declaration of the corresponding subprogram in the
1456 generic package. This approach is recursively used to create SOURCE_LOCATIONs
1457 in case of nested instantiations.
1459 The effect of the pragma is to allow the compiler to eliminate
1460 the code or data associated with the named entity. Any reference to
1461 an eliminated entity outside the compilation unit it is defined in,
1462 causes a compile time or link time error.
1464 The intention of pragma @code{Eliminate} is to allow a program to be compiled
1465 in a system independent manner, with unused entities eliminated, without
1466 the requirement of modifying the source text. Normally the required set
1467 of @code{Eliminate} pragmas is constructed automatically using the gnatelim
1468 tool. Elimination of unused entities local to a compilation unit is
1469 automatic, without requiring the use of pragma @code{Eliminate}.
1471 Note that the reason this pragma takes string literals where names might
1472 be expected is that a pragma @code{Eliminate} can appear in a context where the
1473 relevant names are not visible.
1475 Note that any change in the source files that includes removing, splitting of
1476 adding lines may make the set of Eliminate pragmas using SOURCE_LOCATION
1479 @node Pragma Export_Exception
1480 @unnumberedsec Pragma Export_Exception
1482 @findex Export_Exception
1486 @smallexample @c ada
1487 pragma Export_Exception (
1488 [Internal =>] LOCAL_NAME,
1489 [, [External =>] EXTERNAL_SYMBOL,]
1490 [, [Form =>] Ada | VMS]
1491 [, [Code =>] static_integer_EXPRESSION]);
1495 | static_string_EXPRESSION
1499 This pragma is implemented only in the OpenVMS implementation of GNAT@. It
1500 causes the specified exception to be propagated outside of the Ada program,
1501 so that it can be handled by programs written in other OpenVMS languages.
1502 This pragma establishes an external name for an Ada exception and makes the
1503 name available to the OpenVMS Linker as a global symbol. For further details
1504 on this pragma, see the
1505 DEC Ada Language Reference Manual, section 13.9a3.2.
1507 @node Pragma Export_Function
1508 @unnumberedsec Pragma Export_Function
1509 @cindex Argument passing mechanisms
1510 @findex Export_Function
1515 @smallexample @c ada
1516 pragma Export_Function (
1517 [Internal =>] LOCAL_NAME,
1518 [, [External =>] EXTERNAL_SYMBOL]
1519 [, [Parameter_Types =>] PARAMETER_TYPES]
1520 [, [Result_Type =>] result_SUBTYPE_MARK]
1521 [, [Mechanism =>] MECHANISM]
1522 [, [Result_Mechanism =>] MECHANISM_NAME]);
1526 | static_string_EXPRESSION
1531 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1535 | subtype_Name ' Access
1539 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1541 MECHANISM_ASSOCIATION ::=
1542 [formal_parameter_NAME =>] MECHANISM_NAME
1550 Use this pragma to make a function externally callable and optionally
1551 provide information on mechanisms to be used for passing parameter and
1552 result values. We recommend, for the purposes of improving portability,
1553 this pragma always be used in conjunction with a separate pragma
1554 @code{Export}, which must precede the pragma @code{Export_Function}.
1555 GNAT does not require a separate pragma @code{Export}, but if none is
1556 present, @code{Convention Ada} is assumed, which is usually
1557 not what is wanted, so it is usually appropriate to use this
1558 pragma in conjunction with a @code{Export} or @code{Convention}
1559 pragma that specifies the desired foreign convention.
1560 Pragma @code{Export_Function}
1561 (and @code{Export}, if present) must appear in the same declarative
1562 region as the function to which they apply.
1564 @var{internal_name} must uniquely designate the function to which the
1565 pragma applies. If more than one function name exists of this name in
1566 the declarative part you must use the @code{Parameter_Types} and
1567 @code{Result_Type} parameters is mandatory to achieve the required
1568 unique designation. @var{subtype_ mark}s in these parameters must
1569 exactly match the subtypes in the corresponding function specification,
1570 using positional notation to match parameters with subtype marks.
1571 The form with an @code{'Access} attribute can be used to match an
1572 anonymous access parameter.
1575 @cindex Passing by descriptor
1576 Note that passing by descriptor is not supported, even on the OpenVMS
1579 @cindex Suppressing external name
1580 Special treatment is given if the EXTERNAL is an explicit null
1581 string or a static string expressions that evaluates to the null
1582 string. In this case, no external name is generated. This form
1583 still allows the specification of parameter mechanisms.
1585 @node Pragma Export_Object
1586 @unnumberedsec Pragma Export_Object
1587 @findex Export_Object
1591 @smallexample @c ada
1592 pragma Export_Object
1593 [Internal =>] LOCAL_NAME,
1594 [, [External =>] EXTERNAL_SYMBOL]
1595 [, [Size =>] EXTERNAL_SYMBOL]
1599 | static_string_EXPRESSION
1603 This pragma designates an object as exported, and apart from the
1604 extended rules for external symbols, is identical in effect to the use of
1605 the normal @code{Export} pragma applied to an object. You may use a
1606 separate Export pragma (and you probably should from the point of view
1607 of portability), but it is not required. @var{Size} is syntax checked,
1608 but otherwise ignored by GNAT@.
1610 @node Pragma Export_Procedure
1611 @unnumberedsec Pragma Export_Procedure
1612 @findex Export_Procedure
1616 @smallexample @c ada
1617 pragma Export_Procedure (
1618 [Internal =>] LOCAL_NAME
1619 [, [External =>] EXTERNAL_SYMBOL]
1620 [, [Parameter_Types =>] PARAMETER_TYPES]
1621 [, [Mechanism =>] MECHANISM]);
1625 | static_string_EXPRESSION
1630 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1634 | subtype_Name ' Access
1638 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1640 MECHANISM_ASSOCIATION ::=
1641 [formal_parameter_NAME =>] MECHANISM_NAME
1649 This pragma is identical to @code{Export_Function} except that it
1650 applies to a procedure rather than a function and the parameters
1651 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
1652 GNAT does not require a separate pragma @code{Export}, but if none is
1653 present, @code{Convention Ada} is assumed, which is usually
1654 not what is wanted, so it is usually appropriate to use this
1655 pragma in conjunction with a @code{Export} or @code{Convention}
1656 pragma that specifies the desired foreign convention.
1659 @cindex Passing by descriptor
1660 Note that passing by descriptor is not supported, even on the OpenVMS
1663 @cindex Suppressing external name
1664 Special treatment is given if the EXTERNAL is an explicit null
1665 string or a static string expressions that evaluates to the null
1666 string. In this case, no external name is generated. This form
1667 still allows the specification of parameter mechanisms.
1669 @node Pragma Export_Value
1670 @unnumberedsec Pragma Export_Value
1671 @findex Export_Value
1675 @smallexample @c ada
1676 pragma Export_Value (
1677 [Value =>] static_integer_EXPRESSION,
1678 [Link_Name =>] static_string_EXPRESSION);
1682 This pragma serves to export a static integer value for external use.
1683 The first argument specifies the value to be exported. The Link_Name
1684 argument specifies the symbolic name to be associated with the integer
1685 value. This pragma is useful for defining a named static value in Ada
1686 that can be referenced in assembly language units to be linked with
1687 the application. This pragma is currently supported only for the
1688 AAMP target and is ignored for other targets.
1690 @node Pragma Export_Valued_Procedure
1691 @unnumberedsec Pragma Export_Valued_Procedure
1692 @findex Export_Valued_Procedure
1696 @smallexample @c ada
1697 pragma Export_Valued_Procedure (
1698 [Internal =>] LOCAL_NAME
1699 [, [External =>] EXTERNAL_SYMBOL]
1700 [, [Parameter_Types =>] PARAMETER_TYPES]
1701 [, [Mechanism =>] MECHANISM]);
1705 | static_string_EXPRESSION
1710 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1714 | subtype_Name ' Access
1718 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1720 MECHANISM_ASSOCIATION ::=
1721 [formal_parameter_NAME =>] MECHANISM_NAME
1729 This pragma is identical to @code{Export_Procedure} except that the
1730 first parameter of @var{local_name}, which must be present, must be of
1731 mode @code{OUT}, and externally the subprogram is treated as a function
1732 with this parameter as the result of the function. GNAT provides for
1733 this capability to allow the use of @code{OUT} and @code{IN OUT}
1734 parameters in interfacing to external functions (which are not permitted
1736 GNAT does not require a separate pragma @code{Export}, but if none is
1737 present, @code{Convention Ada} is assumed, which is almost certainly
1738 not what is wanted since the whole point of this pragma is to interface
1739 with foreign language functions, so it is usually appropriate to use this
1740 pragma in conjunction with a @code{Export} or @code{Convention}
1741 pragma that specifies the desired foreign convention.
1744 @cindex Passing by descriptor
1745 Note that passing by descriptor is not supported, even on the OpenVMS
1748 @cindex Suppressing external name
1749 Special treatment is given if the EXTERNAL is an explicit null
1750 string or a static string expressions that evaluates to the null
1751 string. In this case, no external name is generated. This form
1752 still allows the specification of parameter mechanisms.
1754 @node Pragma Extend_System
1755 @unnumberedsec Pragma Extend_System
1756 @cindex @code{system}, extending
1758 @findex Extend_System
1762 @smallexample @c ada
1763 pragma Extend_System ([Name =>] IDENTIFIER);
1767 This pragma is used to provide backwards compatibility with other
1768 implementations that extend the facilities of package @code{System}. In
1769 GNAT, @code{System} contains only the definitions that are present in
1770 the Ada 95 RM@. However, other implementations, notably the DEC Ada 83
1771 implementation, provide many extensions to package @code{System}.
1773 For each such implementation accommodated by this pragma, GNAT provides a
1774 package @code{Aux_@var{xxx}}, e.g.@: @code{Aux_DEC} for the DEC Ada 83
1775 implementation, which provides the required additional definitions. You
1776 can use this package in two ways. You can @code{with} it in the normal
1777 way and access entities either by selection or using a @code{use}
1778 clause. In this case no special processing is required.
1780 However, if existing code contains references such as
1781 @code{System.@var{xxx}} where @var{xxx} is an entity in the extended
1782 definitions provided in package @code{System}, you may use this pragma
1783 to extend visibility in @code{System} in a non-standard way that
1784 provides greater compatibility with the existing code. Pragma
1785 @code{Extend_System} is a configuration pragma whose single argument is
1786 the name of the package containing the extended definition
1787 (e.g.@: @code{Aux_DEC} for the DEC Ada case). A unit compiled under
1788 control of this pragma will be processed using special visibility
1789 processing that looks in package @code{System.Aux_@var{xxx}} where
1790 @code{Aux_@var{xxx}} is the pragma argument for any entity referenced in
1791 package @code{System}, but not found in package @code{System}.
1793 You can use this pragma either to access a predefined @code{System}
1794 extension supplied with the compiler, for example @code{Aux_DEC} or
1795 you can construct your own extension unit following the above
1796 definition. Note that such a package is a child of @code{System}
1797 and thus is considered part of the implementation. To compile
1798 it you will have to use the appropriate switch for compiling
1799 system units. See the GNAT User's Guide for details.
1801 @node Pragma External
1802 @unnumberedsec Pragma External
1807 @smallexample @c ada
1809 [ Convention =>] convention_IDENTIFIER,
1810 [ Entity =>] local_NAME
1811 [, [External_Name =>] static_string_EXPRESSION ]
1812 [, [Link_Name =>] static_string_EXPRESSION ]);
1816 This pragma is identical in syntax and semantics to pragma
1817 @code{Export} as defined in the Ada Reference Manual. It is
1818 provided for compatibility with some Ada 83 compilers that
1819 used this pragma for exactly the same purposes as pragma
1820 @code{Export} before the latter was standardized.
1822 @node Pragma External_Name_Casing
1823 @unnumberedsec Pragma External_Name_Casing
1824 @cindex Dec Ada 83 casing compatibility
1825 @cindex External Names, casing
1826 @cindex Casing of External names
1827 @findex External_Name_Casing
1831 @smallexample @c ada
1832 pragma External_Name_Casing (
1833 Uppercase | Lowercase
1834 [, Uppercase | Lowercase | As_Is]);
1838 This pragma provides control over the casing of external names associated
1839 with Import and Export pragmas. There are two cases to consider:
1842 @item Implicit external names
1843 Implicit external names are derived from identifiers. The most common case
1844 arises when a standard Ada 95 Import or Export pragma is used with only two
1847 @smallexample @c ada
1848 pragma Import (C, C_Routine);
1852 Since Ada is a case insensitive language, the spelling of the identifier in
1853 the Ada source program does not provide any information on the desired
1854 casing of the external name, and so a convention is needed. In GNAT the
1855 default treatment is that such names are converted to all lower case
1856 letters. This corresponds to the normal C style in many environments.
1857 The first argument of pragma @code{External_Name_Casing} can be used to
1858 control this treatment. If @code{Uppercase} is specified, then the name
1859 will be forced to all uppercase letters. If @code{Lowercase} is specified,
1860 then the normal default of all lower case letters will be used.
1862 This same implicit treatment is also used in the case of extended DEC Ada 83
1863 compatible Import and Export pragmas where an external name is explicitly
1864 specified using an identifier rather than a string.
1866 @item Explicit external names
1867 Explicit external names are given as string literals. The most common case
1868 arises when a standard Ada 95 Import or Export pragma is used with three
1871 @smallexample @c ada
1872 pragma Import (C, C_Routine, "C_routine");
1876 In this case, the string literal normally provides the exact casing required
1877 for the external name. The second argument of pragma
1878 @code{External_Name_Casing} may be used to modify this behavior.
1879 If @code{Uppercase} is specified, then the name
1880 will be forced to all uppercase letters. If @code{Lowercase} is specified,
1881 then the name will be forced to all lowercase letters. A specification of
1882 @code{As_Is} provides the normal default behavior in which the casing is
1883 taken from the string provided.
1887 This pragma may appear anywhere that a pragma is valid. In particular, it
1888 can be used as a configuration pragma in the @file{gnat.adc} file, in which
1889 case it applies to all subsequent compilations, or it can be used as a program
1890 unit pragma, in which case it only applies to the current unit, or it can
1891 be used more locally to control individual Import/Export pragmas.
1893 It is primarily intended for use with OpenVMS systems, where many
1894 compilers convert all symbols to upper case by default. For interfacing to
1895 such compilers (e.g.@: the DEC C compiler), it may be convenient to use
1898 @smallexample @c ada
1899 pragma External_Name_Casing (Uppercase, Uppercase);
1903 to enforce the upper casing of all external symbols.
1905 @node Pragma Finalize_Storage_Only
1906 @unnumberedsec Pragma Finalize_Storage_Only
1907 @findex Finalize_Storage_Only
1911 @smallexample @c ada
1912 pragma Finalize_Storage_Only (first_subtype_LOCAL_NAME);
1916 This pragma allows the compiler not to emit a Finalize call for objects
1917 defined at the library level. This is mostly useful for types where
1918 finalization is only used to deal with storage reclamation since in most
1919 environments it is not necessary to reclaim memory just before terminating
1920 execution, hence the name.
1922 @node Pragma Float_Representation
1923 @unnumberedsec Pragma Float_Representation
1925 @findex Float_Representation
1929 @smallexample @c ada
1930 pragma Float_Representation (FLOAT_REP);
1932 FLOAT_REP ::= VAX_Float | IEEE_Float
1937 allows control over the internal representation chosen for the predefined
1938 floating point types declared in the packages @code{Standard} and
1939 @code{System}. On all systems other than OpenVMS, the argument must
1940 be @code{IEEE_Float} and the pragma has no effect. On OpenVMS, the
1941 argument may be @code{VAX_Float} to specify the use of the VAX float
1942 format for the floating-point types in Standard. This requires that
1943 the standard runtime libraries be recompiled. See the
1944 description of the @code{GNAT LIBRARY} command in the OpenVMS version
1945 of the GNAT Users Guide for details on the use of this command.
1948 @unnumberedsec Pragma Ident
1953 @smallexample @c ada
1954 pragma Ident (static_string_EXPRESSION);
1958 This pragma provides a string identification in the generated object file,
1959 if the system supports the concept of this kind of identification string.
1960 This pragma is allowed only in the outermost declarative part or
1961 declarative items of a compilation unit. If more than one @code{Ident}
1962 pragma is given, only the last one processed is effective.
1964 On OpenVMS systems, the effect of the pragma is identical to the effect of
1965 the DEC Ada 83 pragma of the same name. Note that in DEC Ada 83, the
1966 maximum allowed length is 31 characters, so if it is important to
1967 maintain compatibility with this compiler, you should obey this length
1970 @node Pragma Import_Exception
1971 @unnumberedsec Pragma Import_Exception
1973 @findex Import_Exception
1977 @smallexample @c ada
1978 pragma Import_Exception (
1979 [Internal =>] LOCAL_NAME,
1980 [, [External =>] EXTERNAL_SYMBOL,]
1981 [, [Form =>] Ada | VMS]
1982 [, [Code =>] static_integer_EXPRESSION]);
1986 | static_string_EXPRESSION
1990 This pragma is implemented only in the OpenVMS implementation of GNAT@.
1991 It allows OpenVMS conditions (for example, from OpenVMS system services or
1992 other OpenVMS languages) to be propagated to Ada programs as Ada exceptions.
1993 The pragma specifies that the exception associated with an exception
1994 declaration in an Ada program be defined externally (in non-Ada code).
1995 For further details on this pragma, see the
1996 DEC Ada Language Reference Manual, section 13.9a.3.1.
1998 @node Pragma Import_Function
1999 @unnumberedsec Pragma Import_Function
2000 @findex Import_Function
2004 @smallexample @c ada
2005 pragma Import_Function (
2006 [Internal =>] LOCAL_NAME,
2007 [, [External =>] EXTERNAL_SYMBOL]
2008 [, [Parameter_Types =>] PARAMETER_TYPES]
2009 [, [Result_Type =>] SUBTYPE_MARK]
2010 [, [Mechanism =>] MECHANISM]
2011 [, [Result_Mechanism =>] MECHANISM_NAME]
2012 [, [First_Optional_Parameter =>] IDENTIFIER]);
2016 | static_string_EXPRESSION
2020 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2024 | subtype_Name ' Access
2028 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2030 MECHANISM_ASSOCIATION ::=
2031 [formal_parameter_NAME =>] MECHANISM_NAME
2036 | Descriptor [([Class =>] CLASS_NAME)]
2038 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2042 This pragma is used in conjunction with a pragma @code{Import} to
2043 specify additional information for an imported function. The pragma
2044 @code{Import} (or equivalent pragma @code{Interface}) must precede the
2045 @code{Import_Function} pragma and both must appear in the same
2046 declarative part as the function specification.
2048 The @var{Internal} argument must uniquely designate
2049 the function to which the
2050 pragma applies. If more than one function name exists of this name in
2051 the declarative part you must use the @code{Parameter_Types} and
2052 @var{Result_Type} parameters to achieve the required unique
2053 designation. Subtype marks in these parameters must exactly match the
2054 subtypes in the corresponding function specification, using positional
2055 notation to match parameters with subtype marks.
2056 The form with an @code{'Access} attribute can be used to match an
2057 anonymous access parameter.
2059 You may optionally use the @var{Mechanism} and @var{Result_Mechanism}
2060 parameters to specify passing mechanisms for the
2061 parameters and result. If you specify a single mechanism name, it
2062 applies to all parameters. Otherwise you may specify a mechanism on a
2063 parameter by parameter basis using either positional or named
2064 notation. If the mechanism is not specified, the default mechanism
2068 @cindex Passing by descriptor
2069 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
2071 @code{First_Optional_Parameter} applies only to OpenVMS ports of GNAT@.
2072 It specifies that the designated parameter and all following parameters
2073 are optional, meaning that they are not passed at the generated code
2074 level (this is distinct from the notion of optional parameters in Ada
2075 where the parameters are passed anyway with the designated optional
2076 parameters). All optional parameters must be of mode @code{IN} and have
2077 default parameter values that are either known at compile time
2078 expressions, or uses of the @code{'Null_Parameter} attribute.
2080 @node Pragma Import_Object
2081 @unnumberedsec Pragma Import_Object
2082 @findex Import_Object
2086 @smallexample @c ada
2087 pragma Import_Object
2088 [Internal =>] LOCAL_NAME,
2089 [, [External =>] EXTERNAL_SYMBOL],
2090 [, [Size =>] EXTERNAL_SYMBOL]);
2094 | static_string_EXPRESSION
2098 This pragma designates an object as imported, and apart from the
2099 extended rules for external symbols, is identical in effect to the use of
2100 the normal @code{Import} pragma applied to an object. Unlike the
2101 subprogram case, you need not use a separate @code{Import} pragma,
2102 although you may do so (and probably should do so from a portability
2103 point of view). @var{size} is syntax checked, but otherwise ignored by
2106 @node Pragma Import_Procedure
2107 @unnumberedsec Pragma Import_Procedure
2108 @findex Import_Procedure
2112 @smallexample @c ada
2113 pragma Import_Procedure (
2114 [Internal =>] LOCAL_NAME,
2115 [, [External =>] EXTERNAL_SYMBOL]
2116 [, [Parameter_Types =>] PARAMETER_TYPES]
2117 [, [Mechanism =>] MECHANISM]
2118 [, [First_Optional_Parameter =>] IDENTIFIER]);
2122 | static_string_EXPRESSION
2126 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2130 | subtype_Name ' Access
2134 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2136 MECHANISM_ASSOCIATION ::=
2137 [formal_parameter_NAME =>] MECHANISM_NAME
2142 | Descriptor [([Class =>] CLASS_NAME)]
2144 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2148 This pragma is identical to @code{Import_Function} except that it
2149 applies to a procedure rather than a function and the parameters
2150 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
2152 @node Pragma Import_Valued_Procedure
2153 @unnumberedsec Pragma Import_Valued_Procedure
2154 @findex Import_Valued_Procedure
2158 @smallexample @c ada
2159 pragma Import_Valued_Procedure (
2160 [Internal =>] LOCAL_NAME,
2161 [, [External =>] EXTERNAL_SYMBOL]
2162 [, [Parameter_Types =>] PARAMETER_TYPES]
2163 [, [Mechanism =>] MECHANISM]
2164 [, [First_Optional_Parameter =>] IDENTIFIER]);
2168 | static_string_EXPRESSION
2172 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2176 | subtype_Name ' Access
2180 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2182 MECHANISM_ASSOCIATION ::=
2183 [formal_parameter_NAME =>] MECHANISM_NAME
2188 | Descriptor [([Class =>] CLASS_NAME)]
2190 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2194 This pragma is identical to @code{Import_Procedure} except that the
2195 first parameter of @var{local_name}, which must be present, must be of
2196 mode @code{OUT}, and externally the subprogram is treated as a function
2197 with this parameter as the result of the function. The purpose of this
2198 capability is to allow the use of @code{OUT} and @code{IN OUT}
2199 parameters in interfacing to external functions (which are not permitted
2200 in Ada functions). You may optionally use the @code{Mechanism}
2201 parameters to specify passing mechanisms for the parameters.
2202 If you specify a single mechanism name, it applies to all parameters.
2203 Otherwise you may specify a mechanism on a parameter by parameter
2204 basis using either positional or named notation. If the mechanism is not
2205 specified, the default mechanism is used.
2207 Note that it is important to use this pragma in conjunction with a separate
2208 pragma Import that specifies the desired convention, since otherwise the
2209 default convention is Ada, which is almost certainly not what is required.
2211 @node Pragma Initialize_Scalars
2212 @unnumberedsec Pragma Initialize_Scalars
2213 @findex Initialize_Scalars
2214 @cindex debugging with Initialize_Scalars
2218 @smallexample @c ada
2219 pragma Initialize_Scalars;
2223 This pragma is similar to @code{Normalize_Scalars} conceptually but has
2224 two important differences. First, there is no requirement for the pragma
2225 to be used uniformly in all units of a partition, in particular, it is fine
2226 to use this just for some or all of the application units of a partition,
2227 without needing to recompile the run-time library.
2229 In the case where some units are compiled with the pragma, and some without,
2230 then a declaration of a variable where the type is defined in package
2231 Standard or is locally declared will always be subject to initialization,
2232 as will any declaration of a scalar variable. For composite variables,
2233 whether the variable is initialized may also depend on whether the package
2234 in which the type of the variable is declared is compiled with the pragma.
2236 The other important difference is that there is control over the value used
2237 for initializing scalar objects. At bind time, you can select whether to
2238 initialize with invalid values (like Normalize_Scalars), or with high or
2239 low values, or with a specified bit pattern. See the users guide for binder
2240 options for specifying these cases.
2242 This means that you can compile a program, and then without having to
2243 recompile the program, you can run it with different values being used
2244 for initializing otherwise uninitialized values, to test if your program
2245 behavior depends on the choice. Of course the behavior should not change,
2246 and if it does, then most likely you have an erroneous reference to an
2247 uninitialized value.
2249 Note that pragma @code{Initialize_Scalars} is particularly useful in
2250 conjunction with the enhanced validity checking that is now provided
2251 in GNAT, which checks for invalid values under more conditions.
2252 Using this feature (see description of the @code{-gnatV} flag in the
2253 users guide) in conjunction with pragma @code{Initialize_Scalars}
2254 provides a powerful new tool to assist in the detection of problems
2255 caused by uninitialized variables.
2257 Note: the use of @code{Initialize_Scalars} has a fairly extensive
2258 effect on the generated code. This may cause your code to be
2259 substantially larger. It may also cause an increase in the amount
2260 of stack required, so it is probably a good idea to turn on stack
2261 checking (see description of stack checking in the GNAT users guide)
2262 when using this pragma.
2264 @node Pragma Inline_Always
2265 @unnumberedsec Pragma Inline_Always
2266 @findex Inline_Always
2270 @smallexample @c ada
2271 pragma Inline_Always (NAME [, NAME]);
2275 Similar to pragma @code{Inline} except that inlining is not subject to
2276 the use of option @code{-gnatn} and the inlining happens regardless of
2277 whether this option is used.
2279 @node Pragma Inline_Generic
2280 @unnumberedsec Pragma Inline_Generic
2281 @findex Inline_Generic
2285 @smallexample @c ada
2286 pragma Inline_Generic (generic_package_NAME);
2290 This is implemented for compatibility with DEC Ada 83 and is recognized,
2291 but otherwise ignored, by GNAT@. All generic instantiations are inlined
2292 by default when using GNAT@.
2294 @node Pragma Interface
2295 @unnumberedsec Pragma Interface
2300 @smallexample @c ada
2302 [Convention =>] convention_identifier,
2303 [Entity =>] local_name
2304 [, [External_Name =>] static_string_expression],
2305 [, [Link_Name =>] static_string_expression]);
2309 This pragma is identical in syntax and semantics to
2310 the standard Ada 95 pragma @code{Import}. It is provided for compatibility
2311 with Ada 83. The definition is upwards compatible both with pragma
2312 @code{Interface} as defined in the Ada 83 Reference Manual, and also
2313 with some extended implementations of this pragma in certain Ada 83
2316 @node Pragma Interface_Name
2317 @unnumberedsec Pragma Interface_Name
2318 @findex Interface_Name
2322 @smallexample @c ada
2323 pragma Interface_Name (
2324 [Entity =>] LOCAL_NAME
2325 [, [External_Name =>] static_string_EXPRESSION]
2326 [, [Link_Name =>] static_string_EXPRESSION]);
2330 This pragma provides an alternative way of specifying the interface name
2331 for an interfaced subprogram, and is provided for compatibility with Ada
2332 83 compilers that use the pragma for this purpose. You must provide at
2333 least one of @var{External_Name} or @var{Link_Name}.
2335 @node Pragma Interrupt_Handler
2336 @unnumberedsec Pragma Interrupt_Handler
2337 @findex Interrupt_Handler
2341 @smallexample @c ada
2342 pragma Interrupt_Handler (procedure_LOCAL_NAME);
2346 This program unit pragma is supported for parameterless protected procedures
2347 as described in Annex C of the Ada Reference Manual. On the AAMP target
2348 the pragma can also be specified for nonprotected parameterless procedures
2349 that are declared at the library level (which includes procedures
2350 declared at the top level of a library package). In the case of AAMP,
2351 when this pragma is applied to a nonprotected procedure, the instruction
2352 @code{IERET} is generated for returns from the procedure, enabling
2353 maskable interrupts, in place of the normal return instruction.
2355 @node Pragma Interrupt_State
2356 @unnumberedsec Pragma Interrupt_State
2357 @findex Interrupt_State
2361 @smallexample @c ada
2362 pragma Interrupt_State (Name => value, State => SYSTEM | RUNTIME | USER);
2366 Normally certain interrupts are reserved to the implementation. Any attempt
2367 to attach an interrupt causes Program_Error to be raised, as described in
2368 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
2369 many systems for an @kbd{Ctrl-C} interrupt. Normally this interrupt is
2370 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
2371 interrupt execution. Additionally, signals such as @code{SIGSEGV},
2372 @code{SIGABRT}, @code{SIGFPE} and @code{SIGILL} are often mapped to specific
2373 Ada exceptions, or used to implement run-time functions such as the
2374 @code{abort} statement and stack overflow checking.
2376 Pragma @code{Interrupt_State} provides a general mechanism for overriding
2377 such uses of interrupts. It subsumes the functionality of pragma
2378 @code{Unreserve_All_Interrupts}. Pragma @code{Interrupt_State} is not
2379 available on OS/2, Windows or VMS. On all other platforms than VxWorks,
2380 it applies to signals; on VxWorks, it applies to vectored hardware interrupts
2381 and may be used to mark interrupts required by the board support package
2384 Interrupts can be in one of three states:
2388 The interrupt is reserved (no Ada handler can be installed), and the
2389 Ada run-time may not install a handler. As a result you are guaranteed
2390 standard system default action if this interrupt is raised.
2394 The interrupt is reserved (no Ada handler can be installed). The run time
2395 is allowed to install a handler for internal control purposes, but is
2396 not required to do so.
2400 The interrupt is unreserved. The user may install a handler to provide
2405 These states are the allowed values of the @code{State} parameter of the
2406 pragma. The @code{Name} parameter is a value of the type
2407 @code{Ada.Interrupts.Interrupt_ID}. Typically, it is a name declared in
2408 @code{Ada.Interrupts.Names}.
2410 This is a configuration pragma, and the binder will check that there
2411 are no inconsistencies between different units in a partition in how a
2412 given interrupt is specified. It may appear anywhere a pragma is legal.
2414 The effect is to move the interrupt to the specified state.
2416 By declaring interrupts to be SYSTEM, you guarantee the standard system
2417 action, such as a core dump.
2419 By declaring interrupts to be USER, you guarantee that you can install
2422 Note that certain signals on many operating systems cannot be caught and
2423 handled by applications. In such cases, the pragma is ignored. See the
2424 operating system documentation, or the value of the array @code{Reserved}
2425 declared in the specification of package @code{System.OS_Interface}.
2427 Overriding the default state of signals used by the Ada runtime may interfere
2428 with an application's runtime behavior in the cases of the synchronous signals,
2429 and in the case of the signal used to implement the @code{abort} statement.
2431 @node Pragma Keep_Names
2432 @unnumberedsec Pragma Keep_Names
2437 @smallexample @c ada
2438 pragma Keep_Names ([On =>] enumeration_first_subtype_LOCAL_NAME);
2442 The @var{LOCAL_NAME} argument
2443 must refer to an enumeration first subtype
2444 in the current declarative part. The effect is to retain the enumeration
2445 literal names for use by @code{Image} and @code{Value} even if a global
2446 @code{Discard_Names} pragma applies. This is useful when you want to
2447 generally suppress enumeration literal names and for example you therefore
2448 use a @code{Discard_Names} pragma in the @file{gnat.adc} file, but you
2449 want to retain the names for specific enumeration types.
2451 @node Pragma License
2452 @unnumberedsec Pragma License
2454 @cindex License checking
2458 @smallexample @c ada
2459 pragma License (Unrestricted | GPL | Modified_GPL | Restricted);
2463 This pragma is provided to allow automated checking for appropriate license
2464 conditions with respect to the standard and modified GPL@. A pragma
2465 @code{License}, which is a configuration pragma that typically appears at
2466 the start of a source file or in a separate @file{gnat.adc} file, specifies
2467 the licensing conditions of a unit as follows:
2471 This is used for a unit that can be freely used with no license restrictions.
2472 Examples of such units are public domain units, and units from the Ada
2476 This is used for a unit that is licensed under the unmodified GPL, and which
2477 therefore cannot be @code{with}'ed by a restricted unit.
2480 This is used for a unit licensed under the GNAT modified GPL that includes
2481 a special exception paragraph that specifically permits the inclusion of
2482 the unit in programs without requiring the entire program to be released
2483 under the GPL@. This is the license used for the GNAT run-time which ensures
2484 that the run-time can be used freely in any program without GPL concerns.
2487 This is used for a unit that is restricted in that it is not permitted to
2488 depend on units that are licensed under the GPL@. Typical examples are
2489 proprietary code that is to be released under more restrictive license
2490 conditions. Note that restricted units are permitted to @code{with} units
2491 which are licensed under the modified GPL (this is the whole point of the
2497 Normally a unit with no @code{License} pragma is considered to have an
2498 unknown license, and no checking is done. However, standard GNAT headers
2499 are recognized, and license information is derived from them as follows.
2503 A GNAT license header starts with a line containing 78 hyphens. The following
2504 comment text is searched for the appearance of any of the following strings.
2506 If the string ``GNU General Public License'' is found, then the unit is assumed
2507 to have GPL license, unless the string ``As a special exception'' follows, in
2508 which case the license is assumed to be modified GPL@.
2510 If one of the strings
2511 ``This specification is adapted from the Ada Semantic Interface'' or
2512 ``This specification is derived from the Ada Reference Manual'' is found
2513 then the unit is assumed to be unrestricted.
2517 These default actions means that a program with a restricted license pragma
2518 will automatically get warnings if a GPL unit is inappropriately
2519 @code{with}'ed. For example, the program:
2521 @smallexample @c ada
2524 procedure Secret_Stuff is
2530 if compiled with pragma @code{License} (@code{Restricted}) in a
2531 @file{gnat.adc} file will generate the warning:
2536 >>> license of withed unit "Sem_Ch3" is incompatible
2538 2. with GNAT.Sockets;
2539 3. procedure Secret_Stuff is
2543 Here we get a warning on @code{Sem_Ch3} since it is part of the GNAT
2544 compiler and is licensed under the
2545 GPL, but no warning for @code{GNAT.Sockets} which is part of the GNAT
2546 run time, and is therefore licensed under the modified GPL@.
2548 @node Pragma Link_With
2549 @unnumberedsec Pragma Link_With
2554 @smallexample @c ada
2555 pragma Link_With (static_string_EXPRESSION @{,static_string_EXPRESSION@});
2559 This pragma is provided for compatibility with certain Ada 83 compilers.
2560 It has exactly the same effect as pragma @code{Linker_Options} except
2561 that spaces occurring within one of the string expressions are treated
2562 as separators. For example, in the following case:
2564 @smallexample @c ada
2565 pragma Link_With ("-labc -ldef");
2569 results in passing the strings @code{-labc} and @code{-ldef} as two
2570 separate arguments to the linker. In addition pragma Link_With allows
2571 multiple arguments, with the same effect as successive pragmas.
2573 @node Pragma Linker_Alias
2574 @unnumberedsec Pragma Linker_Alias
2575 @findex Linker_Alias
2579 @smallexample @c ada
2580 pragma Linker_Alias (
2581 [Entity =>] LOCAL_NAME
2582 [Alias =>] static_string_EXPRESSION);
2586 This pragma establishes a linker alias for the given named entity. For
2587 further details on the exact effect, consult the GCC manual.
2589 @node Pragma Linker_Section
2590 @unnumberedsec Pragma Linker_Section
2591 @findex Linker_Section
2595 @smallexample @c ada
2596 pragma Linker_Section (
2597 [Entity =>] LOCAL_NAME
2598 [Section =>] static_string_EXPRESSION);
2602 This pragma specifies the name of the linker section for the given entity.
2603 For further details on the exact effect, consult the GCC manual.
2605 @node Pragma Long_Float
2606 @unnumberedsec Pragma Long_Float
2612 @smallexample @c ada
2613 pragma Long_Float (FLOAT_FORMAT);
2615 FLOAT_FORMAT ::= D_Float | G_Float
2619 This pragma is implemented only in the OpenVMS implementation of GNAT@.
2620 It allows control over the internal representation chosen for the predefined
2621 type @code{Long_Float} and for floating point type representations with
2622 @code{digits} specified in the range 7 through 15.
2623 For further details on this pragma, see the
2624 @cite{DEC Ada Language Reference Manual}, section 3.5.7b. Note that to use
2625 this pragma, the standard runtime libraries must be recompiled. See the
2626 description of the @code{GNAT LIBRARY} command in the OpenVMS version
2627 of the GNAT User's Guide for details on the use of this command.
2629 @node Pragma Machine_Attribute
2630 @unnumberedsec Pragma Machine_Attribute
2631 @findex Machine_Attribute
2635 @smallexample @c ada
2636 pragma Machine_Attribute (
2637 [Attribute_Name =>] string_EXPRESSION,
2638 [Entity =>] LOCAL_NAME);
2642 Machine dependent attributes can be specified for types and/or
2643 declarations. Currently only subprogram entities are supported. This
2644 pragma is semantically equivalent to
2645 @code{__attribute__((@var{string_expression}))} in GNU C,
2646 where @code{@var{string_expression}} is
2647 recognized by the GNU C macros @code{VALID_MACHINE_TYPE_ATTRIBUTE} and
2648 @code{VALID_MACHINE_DECL_ATTRIBUTE} which are defined in the
2649 configuration header file @file{tm.h} for each machine. See the GCC
2650 manual for further information.
2652 @node Pragma Main_Storage
2653 @unnumberedsec Pragma Main_Storage
2655 @findex Main_Storage
2659 @smallexample @c ada
2661 (MAIN_STORAGE_OPTION [, MAIN_STORAGE_OPTION]);
2663 MAIN_STORAGE_OPTION ::=
2664 [WORKING_STORAGE =>] static_SIMPLE_EXPRESSION
2665 | [TOP_GUARD =>] static_SIMPLE_EXPRESSION
2670 This pragma is provided for compatibility with OpenVMS VAX Systems. It has
2671 no effect in GNAT, other than being syntax checked. Note that the pragma
2672 also has no effect in DEC Ada 83 for OpenVMS Alpha Systems.
2674 @node Pragma No_Return
2675 @unnumberedsec Pragma No_Return
2680 @smallexample @c ada
2681 pragma No_Return (procedure_LOCAL_NAME);
2685 @var{procedure_local_NAME} must refer to one or more procedure
2686 declarations in the current declarative part. A procedure to which this
2687 pragma is applied may not contain any explicit @code{return} statements,
2688 and also may not contain any implicit return statements from falling off
2689 the end of a statement sequence. One use of this pragma is to identify
2690 procedures whose only purpose is to raise an exception.
2692 Another use of this pragma is to suppress incorrect warnings about
2693 missing returns in functions, where the last statement of a function
2694 statement sequence is a call to such a procedure.
2696 @node Pragma Normalize_Scalars
2697 @unnumberedsec Pragma Normalize_Scalars
2698 @findex Normalize_Scalars
2702 @smallexample @c ada
2703 pragma Normalize_Scalars;
2707 This is a language defined pragma which is fully implemented in GNAT@. The
2708 effect is to cause all scalar objects that are not otherwise initialized
2709 to be initialized. The initial values are implementation dependent and
2713 @item Standard.Character
2715 Objects whose root type is Standard.Character are initialized to
2716 Character'Last. This will be out of range of the subtype only if
2717 the subtype range excludes this value.
2719 @item Standard.Wide_Character
2721 Objects whose root type is Standard.Wide_Character are initialized to
2722 Wide_Character'Last. This will be out of range of the subtype only if
2723 the subtype range excludes this value.
2727 Objects of an integer type are initialized to base_type'First, where
2728 base_type is the base type of the object type. This will be out of range
2729 of the subtype only if the subtype range excludes this value. For example,
2730 if you declare the subtype:
2732 @smallexample @c ada
2733 subtype Ityp is integer range 1 .. 10;
2737 then objects of type x will be initialized to Integer'First, a negative
2738 number that is certainly outside the range of subtype @code{Ityp}.
2741 Objects of all real types (fixed and floating) are initialized to
2742 base_type'First, where base_Type is the base type of the object type.
2743 This will be out of range of the subtype only if the subtype range
2744 excludes this value.
2747 Objects of a modular type are initialized to typ'Last. This will be out
2748 of range of the subtype only if the subtype excludes this value.
2750 @item Enumeration types
2751 Objects of an enumeration type are initialized to all one-bits, i.e.@: to
2752 the value @code{2 ** typ'Size - 1}. This will be out of range of the
2753 enumeration subtype in all cases except where the subtype contains
2754 exactly 2**8, 2**16, or 2**32 elements.
2758 @node Pragma Obsolescent
2759 @unnumberedsec Pragma Obsolescent
2764 @smallexample @c ada
2765 pragma Obsolescent [(static_string_EXPRESSION)];
2769 This pragma must occur immediately following a subprogram
2770 declaration. It indicates that the associated function or procedure
2771 is considered obsolescent and should not be used. Typically this is
2772 used when an API must be modified by eventually removing or modifying
2773 existing subprograms. The pragma can be used at an intermediate stage
2774 when the subprogram is still present, but will be removed later.
2776 The effect of this pragma is to output a warning message that the
2777 subprogram is obsolescent if the appropriate warning option in the
2778 compiler is activated. If a parameter is present, then a second
2779 warning message is given containing this text.
2781 @node Pragma Passive
2782 @unnumberedsec Pragma Passive
2787 @smallexample @c ada
2788 pragma Passive ([Semaphore | No]);
2792 Syntax checked, but otherwise ignored by GNAT@. This is recognized for
2793 compatibility with DEC Ada 83 implementations, where it is used within a
2794 task definition to request that a task be made passive. If the argument
2795 @code{Semaphore} is present, or the argument is omitted, then DEC Ada 83
2796 treats the pragma as an assertion that the containing task is passive
2797 and that optimization of context switch with this task is permitted and
2798 desired. If the argument @code{No} is present, the task must not be
2799 optimized. GNAT does not attempt to optimize any tasks in this manner
2800 (since protected objects are available in place of passive tasks).
2802 @node Pragma Polling
2803 @unnumberedsec Pragma Polling
2808 @smallexample @c ada
2809 pragma Polling (ON | OFF);
2813 This pragma controls the generation of polling code. This is normally off.
2814 If @code{pragma Polling (ON)} is used then periodic calls are generated to
2815 the routine @code{Ada.Exceptions.Poll}. This routine is a separate unit in the
2816 runtime library, and can be found in file @file{a-excpol.adb}.
2818 Pragma @code{Polling} can appear as a configuration pragma (for example it
2819 can be placed in the @file{gnat.adc} file) to enable polling globally, or it
2820 can be used in the statement or declaration sequence to control polling
2823 A call to the polling routine is generated at the start of every loop and
2824 at the start of every subprogram call. This guarantees that the @code{Poll}
2825 routine is called frequently, and places an upper bound (determined by
2826 the complexity of the code) on the period between two @code{Poll} calls.
2828 The primary purpose of the polling interface is to enable asynchronous
2829 aborts on targets that cannot otherwise support it (for example Windows
2830 NT), but it may be used for any other purpose requiring periodic polling.
2831 The standard version is null, and can be replaced by a user program. This
2832 will require re-compilation of the @code{Ada.Exceptions} package that can
2833 be found in files @file{a-except.ads} and @file{a-except.adb}.
2835 A standard alternative unit (in file @file{4wexcpol.adb} in the standard GNAT
2836 distribution) is used to enable the asynchronous abort capability on
2837 targets that do not normally support the capability. The version of
2838 @code{Poll} in this file makes a call to the appropriate runtime routine
2839 to test for an abort condition.
2841 Note that polling can also be enabled by use of the @code{-gnatP} switch. See
2842 the @cite{GNAT User's Guide} for details.
2844 @node Pragma Profile (Ravenscar)
2845 @unnumberedsec Pragma Profile (Ravenscar)
2850 @smallexample @c ada
2851 pragma Profile (Ravenscar);
2855 A configuration pragma that establishes the following set of configuration
2859 @item Task_Dispatching_Policy (FIFO_Within_Priorities)
2860 [RM D.2.2] Tasks are dispatched following a preemptive
2861 priority-ordered scheduling policy.
2863 @item Locking_Policy (Ceiling_Locking)
2864 [RM D.3] While tasks and interrupts execute a protected action, they inherit
2865 the ceiling priority of the corresponding protected object.
2867 @c @item Detect_Blocking
2868 @c This pragma forces the detection of potentially blocking operations within a
2869 @c protected operation, and to raise Program_Error if that happens.
2873 plus the following set of restrictions:
2876 @item Max_Entry_Queue_Length = 1
2877 Defines the maximum number of calls that are queued on a (protected) entry.
2878 Note that this restrictions is checked at run time. Violation of this
2879 restriction results in the raising of Program_Error exception at the point of
2880 the call. For the Profile (Ravenscar) the value of Max_Entry_Queue_Length is
2881 always 1 and hence no task can be queued on a protected entry.
2883 @item Max_Protected_Entries = 1
2884 [RM D.7] Specifies the maximum number of entries per protected type. The
2885 bounds of every entry family of a protected unit shall be static, or shall be
2886 defined by a discriminant of a subtype whose corresponding bound is static.
2887 For the Profile (Ravenscar) the value of Max_Protected_Entries is always 1.
2889 @item Max_Task_Entries = 0
2890 [RM D.7] Specifies the maximum number of entries
2891 per task. The bounds of every entry family
2892 of a task unit shall be static, or shall be
2893 defined by a discriminant of a subtype whose
2894 corresponding bound is static. A value of zero
2895 indicates that no rendezvous are possible. For
2896 the Profile (Ravenscar), the value of Max_Task_Entries is always
2899 @item No_Abort_Statements
2900 [RM D.7] There are no abort_statements, and there are
2901 no calls to Task_Identification.Abort_Task.
2903 @item No_Asynchronous_Control
2904 [RM D.7] There are no semantic dependences on the package
2905 Asynchronous_Task_Control.
2908 There are no semantic dependencies on the package Ada.Calendar.
2910 @item No_Dynamic_Attachment
2911 There is no call to any of the operations defined in package Ada.Interrupts
2912 (Is_Reserved, Is_Attached, Current_Handler, Attach_Handler, Exchange_Handler,
2913 Detach_Handler, and Reference).
2915 @item No_Dynamic_Priorities
2916 [RM D.7] There are no semantic dependencies on the package Dynamic_Priorities.
2918 @item No_Implicit_Heap_Allocations
2919 [RM D.7] No constructs are allowed to cause implicit heap allocation.
2921 @item No_Local_Protected_Objects
2922 Protected objects and access types that designate
2923 such objects shall be declared only at library level.
2925 @item No_Protected_Type_Allocators
2926 There are no allocators for protected types or
2927 types containing protected subcomponents.
2929 @item No_Relative_Delay
2930 There are no delay_relative statements.
2932 @item No_Requeue_Statements
2933 Requeue statements are not allowed.
2935 @item No_Select_Statements
2936 There are no select_statements.
2938 @item No_Task_Allocators
2939 [RM D.7] There are no allocators for task types
2940 or types containing task subcomponents.
2942 @item No_Task_Attributes_Package
2943 There are no semantic dependencies on the Ada.Task_Attributes package.
2945 @item No_Task_Hierarchy
2946 [RM D.7] All (non-environment) tasks depend
2947 directly on the environment task of the partition.
2949 @item No_Task_Termination
2950 Tasks which terminate are erroneous.
2952 @item Simple_Barriers
2953 Entry barrier condition expressions shall be either static
2954 boolean expressions or boolean objects which are declared in
2955 the protected type which contains the entry.
2959 This set of configuration pragmas and restrictions correspond to the
2960 definition of the ``Ravenscar Profile'' for limited tasking, devised and
2961 published by the @cite{International Real-Time Ada Workshop}, 1997,
2962 and whose most recent description is available at
2963 @url{ftp://ftp.openravenscar.org/openravenscar/ravenscar00.pdf}.
2965 The original definition of the profile was revised at subsequent IRTAW
2966 meetings. It has been included in the ISO
2967 @cite{Guide for the Use of the Ada Programming Language in High
2968 Integrity Systems}, and has been approved by ISO/IEC/SC22/WG9 for inclusion in
2969 the next revision of the standard. The formal definition given by
2970 the Ada Rapporteur Group (ARG) can be found in two Ada Issues (AI-249 and
2971 AI-305) available at
2972 @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs/AI-00249.TXT} and
2973 @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/AIs/AI-00305.TXT}
2976 The above set is a superset of the restrictions provided by pragma
2977 @code{Profile (Restricted)}, it includes six additional restrictions
2978 (@code{Simple_Barriers}, @code{No_Select_Statements},
2979 @code{No_Calendar}, @code{No_Implicit_Heap_Allocations},
2980 @code{No_Relative_Delay} and @code{No_Task_Termination}). This means
2981 that pragma @code{Profile (Ravenscar)}, like the pragma
2982 @code{Profile (Restricted)},
2983 automatically causes the use of a simplified,
2984 more efficient version of the tasking run-time system.
2986 @node Pragma Profile (Restricted)
2987 @unnumberedsec Pragma Profile (Restricted)
2988 @findex Restricted Run Time
2992 @smallexample @c ada
2993 pragma Profile (Restricted);
2997 A configuration pragma that establishes the following set of restrictions:
3000 @item No_Abort_Statements
3001 @item No_Entry_Queue
3002 @item No_Task_Hierarchy
3003 @item No_Task_Allocators
3004 @item No_Dynamic_Priorities
3005 @item No_Terminate_Alternatives
3006 @item No_Dynamic_Attachment
3007 @item No_Protected_Type_Allocators
3008 @item No_Local_Protected_Objects
3009 @item No_Requeue_Statements
3010 @item No_Task_Attributes_Package
3011 @item Max_Asynchronous_Select_Nesting = 0
3012 @item Max_Task_Entries = 0
3013 @item Max_Protected_Entries = 1
3014 @item Max_Select_Alternatives = 0
3018 This set of restrictions causes the automatic selection of a simplified
3019 version of the run time that provides improved performance for the
3020 limited set of tasking functionality permitted by this set of restrictions.
3022 @node Pragma Propagate_Exceptions
3023 @unnumberedsec Pragma Propagate_Exceptions
3024 @findex Propagate_Exceptions
3025 @cindex Zero Cost Exceptions
3029 @smallexample @c ada
3030 pragma Propagate_Exceptions (subprogram_LOCAL_NAME);
3034 This pragma indicates that the given entity, which is the name of an
3035 imported foreign-language subprogram may receive an Ada exception,
3036 and that the exception should be propagated. It is relevant only if
3037 zero cost exception handling is in use, and is thus never needed if
3038 the alternative @code{longjmp} / @code{setjmp} implementation of
3039 exceptions is used (although it is harmless to use it in such cases).
3041 The implementation of fast exceptions always properly propagates
3042 exceptions through Ada code, as described in the Ada Reference Manual.
3043 However, this manual is silent about the propagation of exceptions
3044 through foreign code. For example, consider the
3045 situation where @code{P1} calls
3046 @code{P2}, and @code{P2} calls @code{P3}, where
3047 @code{P1} and @code{P3} are in Ada, but @code{P2} is in C@.
3048 @code{P3} raises an Ada exception. The question is whether or not
3049 it will be propagated through @code{P2} and can be handled in
3052 For the @code{longjmp} / @code{setjmp} implementation of exceptions,
3053 the answer is always yes. For some targets on which zero cost exception
3054 handling is implemented, the answer is also always yes. However, there
3055 are some targets, notably in the current version all x86 architecture
3056 targets, in which the answer is that such propagation does not
3057 happen automatically. If such propagation is required on these
3058 targets, it is mandatory to use @code{Propagate_Exceptions} to
3059 name all foreign language routines through which Ada exceptions
3062 @node Pragma Psect_Object
3063 @unnumberedsec Pragma Psect_Object
3064 @findex Psect_Object
3068 @smallexample @c ada
3069 pragma Psect_Object (
3070 [Internal =>] LOCAL_NAME,
3071 [, [External =>] EXTERNAL_SYMBOL]
3072 [, [Size =>] EXTERNAL_SYMBOL]);
3076 | static_string_EXPRESSION
3080 This pragma is identical in effect to pragma @code{Common_Object}.
3082 @node Pragma Pure_Function
3083 @unnumberedsec Pragma Pure_Function
3084 @findex Pure_Function
3088 @smallexample @c ada
3089 pragma Pure_Function ([Entity =>] function_LOCAL_NAME);
3093 This pragma appears in the same declarative part as a function
3094 declaration (or a set of function declarations if more than one
3095 overloaded declaration exists, in which case the pragma applies
3096 to all entities). It specifies that the function @code{Entity} is
3097 to be considered pure for the purposes of code generation. This means
3098 that the compiler can assume that there are no side effects, and
3099 in particular that two calls with identical arguments produce the
3100 same result. It also means that the function can be used in an
3103 Note that, quite deliberately, there are no static checks to try
3104 to ensure that this promise is met, so @code{Pure_Function} can be used
3105 with functions that are conceptually pure, even if they do modify
3106 global variables. For example, a square root function that is
3107 instrumented to count the number of times it is called is still
3108 conceptually pure, and can still be optimized, even though it
3109 modifies a global variable (the count). Memo functions are another
3110 example (where a table of previous calls is kept and consulted to
3111 avoid re-computation).
3114 Note: Most functions in a @code{Pure} package are automatically pure, and
3115 there is no need to use pragma @code{Pure_Function} for such functions. One
3116 exception is any function that has at least one formal of type
3117 @code{System.Address} or a type derived from it. Such functions are not
3118 considered pure by default, since the compiler assumes that the
3119 @code{Address} parameter may be functioning as a pointer and that the
3120 referenced data may change even if the address value does not.
3121 Similarly, imported functions are not considered to be pure by default,
3122 since there is no way of checking that they are in fact pure. The use
3123 of pragma @code{Pure_Function} for such a function will override these default
3124 assumption, and cause the compiler to treat a designated subprogram as pure
3127 Note: If pragma @code{Pure_Function} is applied to a renamed function, it
3128 applies to the underlying renamed function. This can be used to
3129 disambiguate cases of overloading where some but not all functions
3130 in a set of overloaded functions are to be designated as pure.
3132 @node Pragma Restriction_Warnings
3133 @unnumberedsec Pragma Restriction_Warnings
3134 @findex Restriction_Warnings
3138 @smallexample @c ada
3139 pragma Restriction_Warnings
3140 (restriction_IDENTIFIER @{, restriction_IDENTIFIER@});
3144 This pragma allows a series of restriction identifiers to be
3145 specified (the list of allowed identifiers is the same as for
3146 pragma @code{Restrictions}). For each of these identifiers
3147 the compiler checks for violations of the restriction, but
3148 generates a warning message rather than an error message
3149 if the restriction is violated.
3151 @node Pragma Source_File_Name
3152 @unnumberedsec Pragma Source_File_Name
3153 @findex Source_File_Name
3157 @smallexample @c ada
3158 pragma Source_File_Name (
3159 [Unit_Name =>] unit_NAME,
3160 Spec_File_Name => STRING_LITERAL);
3162 pragma Source_File_Name (
3163 [Unit_Name =>] unit_NAME,
3164 Body_File_Name => STRING_LITERAL);
3168 Use this to override the normal naming convention. It is a configuration
3169 pragma, and so has the usual applicability of configuration pragmas
3170 (i.e.@: it applies to either an entire partition, or to all units in a
3171 compilation, or to a single unit, depending on how it is used.
3172 @var{unit_name} is mapped to @var{file_name_literal}. The identifier for
3173 the second argument is required, and indicates whether this is the file
3174 name for the spec or for the body.
3176 Another form of the @code{Source_File_Name} pragma allows
3177 the specification of patterns defining alternative file naming schemes
3178 to apply to all files.
3180 @smallexample @c ada
3181 pragma Source_File_Name
3182 (Spec_File_Name => STRING_LITERAL
3183 [,Casing => CASING_SPEC]
3184 [,Dot_Replacement => STRING_LITERAL]);
3186 pragma Source_File_Name
3187 (Body_File_Name => STRING_LITERAL
3188 [,Casing => CASING_SPEC]
3189 [,Dot_Replacement => STRING_LITERAL]);
3191 pragma Source_File_Name
3192 (Subunit_File_Name => STRING_LITERAL
3193 [,Casing => CASING_SPEC]
3194 [,Dot_Replacement => STRING_LITERAL]);
3196 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
3200 The first argument is a pattern that contains a single asterisk indicating
3201 the point at which the unit name is to be inserted in the pattern string
3202 to form the file name. The second argument is optional. If present it
3203 specifies the casing of the unit name in the resulting file name string.
3204 The default is lower case. Finally the third argument allows for systematic
3205 replacement of any dots in the unit name by the specified string literal.
3207 A pragma Source_File_Name cannot appear after a
3208 @ref{Pragma Source_File_Name_Project}.
3210 For more details on the use of the @code{Source_File_Name} pragma,
3211 see the sections ``Using Other File Names'' and
3212 ``Alternative File Naming Schemes'' in the @cite{GNAT User's Guide}.
3214 @node Pragma Source_File_Name_Project
3215 @unnumberedsec Pragma Source_File_Name_Project
3216 @findex Source_File_Name_Project
3219 This pragma has the same syntax and semantics as pragma Source_File_Name.
3220 It is only allowed as a stand alone configuration pragma.
3221 It cannot appear after a @ref{Pragma Source_File_Name}, and
3222 most importantly, once pragma Source_File_Name_Project appears,
3223 no further Source_File_Name pragmas are allowed.
3225 The intention is that Source_File_Name_Project pragmas are always
3226 generated by the Project Manager in a manner consistent with the naming
3227 specified in a project file, and when naming is controlled in this manner,
3228 it is not permissible to attempt to modify this naming scheme using
3229 Source_File_Name pragmas (which would not be known to the project manager).
3231 @node Pragma Source_Reference
3232 @unnumberedsec Pragma Source_Reference
3233 @findex Source_Reference
3237 @smallexample @c ada
3238 pragma Source_Reference (INTEGER_LITERAL, STRING_LITERAL);
3242 This pragma must appear as the first line of a source file.
3243 @var{integer_literal} is the logical line number of the line following
3244 the pragma line (for use in error messages and debugging
3245 information). @var{string_literal} is a static string constant that
3246 specifies the file name to be used in error messages and debugging
3247 information. This is most notably used for the output of @code{gnatchop}
3248 with the @code{-r} switch, to make sure that the original unchopped
3249 source file is the one referred to.
3251 The second argument must be a string literal, it cannot be a static
3252 string expression other than a string literal. This is because its value
3253 is needed for error messages issued by all phases of the compiler.
3255 @node Pragma Stream_Convert
3256 @unnumberedsec Pragma Stream_Convert
3257 @findex Stream_Convert
3261 @smallexample @c ada
3262 pragma Stream_Convert (
3263 [Entity =>] type_LOCAL_NAME,
3264 [Read =>] function_NAME,
3265 [Write =>] function_NAME);
3269 This pragma provides an efficient way of providing stream functions for
3270 types defined in packages. Not only is it simpler to use than declaring
3271 the necessary functions with attribute representation clauses, but more
3272 significantly, it allows the declaration to made in such a way that the
3273 stream packages are not loaded unless they are needed. The use of
3274 the Stream_Convert pragma adds no overhead at all, unless the stream
3275 attributes are actually used on the designated type.
3277 The first argument specifies the type for which stream functions are
3278 provided. The second parameter provides a function used to read values
3279 of this type. It must name a function whose argument type may be any
3280 subtype, and whose returned type must be the type given as the first
3281 argument to the pragma.
3283 The meaning of the @var{Read}
3284 parameter is that if a stream attribute directly
3285 or indirectly specifies reading of the type given as the first parameter,
3286 then a value of the type given as the argument to the Read function is
3287 read from the stream, and then the Read function is used to convert this
3288 to the required target type.
3290 Similarly the @var{Write} parameter specifies how to treat write attributes
3291 that directly or indirectly apply to the type given as the first parameter.
3292 It must have an input parameter of the type specified by the first parameter,
3293 and the return type must be the same as the input type of the Read function.
3294 The effect is to first call the Write function to convert to the given stream
3295 type, and then write the result type to the stream.
3297 The Read and Write functions must not be overloaded subprograms. If necessary
3298 renamings can be supplied to meet this requirement.
3299 The usage of this attribute is best illustrated by a simple example, taken
3300 from the GNAT implementation of package Ada.Strings.Unbounded:
3302 @smallexample @c ada
3303 function To_Unbounded (S : String)
3304 return Unbounded_String
3305 renames To_Unbounded_String;
3307 pragma Stream_Convert
3308 (Unbounded_String, To_Unbounded, To_String);
3312 The specifications of the referenced functions, as given in the Ada 95
3313 Reference Manual are:
3315 @smallexample @c ada
3316 function To_Unbounded_String (Source : String)
3317 return Unbounded_String;
3319 function To_String (Source : Unbounded_String)
3324 The effect is that if the value of an unbounded string is written to a
3325 stream, then the representation of the item in the stream is in the same
3326 format used for @code{Standard.String}, and this same representation is
3327 expected when a value of this type is read from the stream.
3329 @node Pragma Style_Checks
3330 @unnumberedsec Pragma Style_Checks
3331 @findex Style_Checks
3335 @smallexample @c ada
3336 pragma Style_Checks (string_LITERAL | ALL_CHECKS |
3337 On | Off [, LOCAL_NAME]);
3341 This pragma is used in conjunction with compiler switches to control the
3342 built in style checking provided by GNAT@. The compiler switches, if set,
3343 provide an initial setting for the switches, and this pragma may be used
3344 to modify these settings, or the settings may be provided entirely by
3345 the use of the pragma. This pragma can be used anywhere that a pragma
3346 is legal, including use as a configuration pragma (including use in
3347 the @file{gnat.adc} file).
3349 The form with a string literal specifies which style options are to be
3350 activated. These are additive, so they apply in addition to any previously
3351 set style check options. The codes for the options are the same as those
3352 used in the @code{-gnaty} switch to @code{gcc} or @code{gnatmake}.
3353 For example the following two methods can be used to enable
3358 @smallexample @c ada
3359 pragma Style_Checks ("l");
3364 gcc -c -gnatyl @dots{}
3369 The form ALL_CHECKS activates all standard checks (its use is equivalent
3370 to the use of the @code{gnaty} switch with no options. See GNAT User's
3373 The forms with @code{Off} and @code{On}
3374 can be used to temporarily disable style checks
3375 as shown in the following example:
3377 @smallexample @c ada
3381 pragma Style_Checks ("k"); -- requires keywords in lower case
3382 pragma Style_Checks (Off); -- turn off style checks
3383 NULL; -- this will not generate an error message
3384 pragma Style_Checks (On); -- turn style checks back on
3385 NULL; -- this will generate an error message
3389 Finally the two argument form is allowed only if the first argument is
3390 @code{On} or @code{Off}. The effect is to turn of semantic style checks
3391 for the specified entity, as shown in the following example:
3393 @smallexample @c ada
3397 pragma Style_Checks ("r"); -- require consistency of identifier casing
3399 Rf1 : Integer := ARG; -- incorrect, wrong case
3400 pragma Style_Checks (Off, Arg);
3401 Rf2 : Integer := ARG; -- OK, no error
3404 @node Pragma Subtitle
3405 @unnumberedsec Pragma Subtitle
3410 @smallexample @c ada
3411 pragma Subtitle ([Subtitle =>] STRING_LITERAL);
3415 This pragma is recognized for compatibility with other Ada compilers
3416 but is ignored by GNAT@.
3418 @node Pragma Suppress_All
3419 @unnumberedsec Pragma Suppress_All
3420 @findex Suppress_All
3424 @smallexample @c ada
3425 pragma Suppress_All;
3429 This pragma can only appear immediately following a compilation
3430 unit. The effect is to apply @code{Suppress (All_Checks)} to the unit
3431 which it follows. This pragma is implemented for compatibility with DEC
3432 Ada 83 usage. The use of pragma @code{Suppress (All_Checks)} as a normal
3433 configuration pragma is the preferred usage in GNAT@.
3435 @node Pragma Suppress_Exception_Locations
3436 @unnumberedsec Pragma Suppress_Exception_Locations
3437 @findex Suppress_Exception_Locations
3441 @smallexample @c ada
3442 pragma Suppress_Exception_Locations;
3446 In normal mode, a raise statement for an exception by default generates
3447 an exception message giving the file name and line number for the location
3448 of the raise. This is useful for debugging and logging purposes, but this
3449 entails extra space for the strings for the messages. The configuration
3450 pragma @code{Suppress_Exception_Locations} can be used to suppress the
3451 generation of these strings, with the result that space is saved, but the
3452 exception message for such raises is null. This configuration pragma may
3453 appear in a global configuration pragma file, or in a specific unit as
3454 usual. It is not required that this pragma be used consistently within
3455 a partition, so it is fine to have some units within a partition compiled
3456 with this pragma and others compiled in normal mode without it.
3458 @node Pragma Suppress_Initialization
3459 @unnumberedsec Pragma Suppress_Initialization
3460 @findex Suppress_Initialization
3461 @cindex Suppressing initialization
3462 @cindex Initialization, suppression of
3466 @smallexample @c ada
3467 pragma Suppress_Initialization ([Entity =>] type_Name);
3471 This pragma suppresses any implicit or explicit initialization
3472 associated with the given type name for all variables of this type.
3474 @node Pragma Task_Info
3475 @unnumberedsec Pragma Task_Info
3480 @smallexample @c ada
3481 pragma Task_Info (EXPRESSION);
3485 This pragma appears within a task definition (like pragma
3486 @code{Priority}) and applies to the task in which it appears. The
3487 argument must be of type @code{System.Task_Info.Task_Info_Type}.
3488 The @code{Task_Info} pragma provides system dependent control over
3489 aspects of tasking implementation, for example, the ability to map
3490 tasks to specific processors. For details on the facilities available
3491 for the version of GNAT that you are using, see the documentation
3492 in the specification of package System.Task_Info in the runtime
3495 @node Pragma Task_Name
3496 @unnumberedsec Pragma Task_Name
3501 @smallexample @c ada
3502 pragma Task_Name (string_EXPRESSION);
3506 This pragma appears within a task definition (like pragma
3507 @code{Priority}) and applies to the task in which it appears. The
3508 argument must be of type String, and provides a name to be used for
3509 the task instance when the task is created. Note that this expression
3510 is not required to be static, and in particular, it can contain
3511 references to task discriminants. This facility can be used to
3512 provide different names for different tasks as they are created,
3513 as illustrated in the example below.
3515 The task name is recorded internally in the run-time structures
3516 and is accessible to tools like the debugger. In addition the
3517 routine @code{Ada.Task_Identification.Image} will return this
3518 string, with a unique task address appended.
3520 @smallexample @c ada
3521 -- Example of the use of pragma Task_Name
3523 with Ada.Task_Identification;
3524 use Ada.Task_Identification;
3525 with Text_IO; use Text_IO;
3528 type Astring is access String;
3530 task type Task_Typ (Name : access String) is
3531 pragma Task_Name (Name.all);
3534 task body Task_Typ is
3535 Nam : constant String := Image (Current_Task);
3537 Put_Line ("-->" & Nam (1 .. 14) & "<--");
3540 type Ptr_Task is access Task_Typ;
3541 Task_Var : Ptr_Task;
3545 new Task_Typ (new String'("This is task 1"));
3547 new Task_Typ (new String'("This is task 2"));
3551 @node Pragma Task_Storage
3552 @unnumberedsec Pragma Task_Storage
3553 @findex Task_Storage
3556 @smallexample @c ada
3557 pragma Task_Storage (
3558 [Task_Type =>] LOCAL_NAME,
3559 [Top_Guard =>] static_integer_EXPRESSION);
3563 This pragma specifies the length of the guard area for tasks. The guard
3564 area is an additional storage area allocated to a task. A value of zero
3565 means that either no guard area is created or a minimal guard area is
3566 created, depending on the target. This pragma can appear anywhere a
3567 @code{Storage_Size} attribute definition clause is allowed for a task
3570 @node Pragma Thread_Body
3571 @unnumberedsec Pragma Thread_Body
3575 @smallexample @c ada
3576 pragma Thread_Body (
3577 [Entity =>] LOCAL_NAME,
3578 [[Secondary_Stack_Size =>] static_integer_EXPRESSION)];
3582 This pragma specifies that the subprogram whose name is given as the
3583 @code{Entity} argument is a thread body, which will be activated
3584 by being called via its Address from foreign code. The purpose is
3585 to allow execution and registration of the foreign thread within the
3586 Ada run-time system.
3588 See the library unit @code{System.Threads} for details on the expansion of
3589 a thread body subprogram, including the calls made to subprograms
3590 within System.Threads to register the task. This unit also lists the
3591 targets and runtime systems for which this pragma is supported.
3593 A thread body subprogram may not be called directly from Ada code, and
3594 it is not permitted to apply the Access (or Unrestricted_Access) attributes
3595 to such a subprogram. The only legitimate way of calling such a subprogram
3596 is to pass its Address to foreign code and then make the call from the
3599 A thread body subprogram may have any parameters, and it may be a function
3600 returning a result. The convention of the thread body subprogram may be
3601 set in the usual manner using @code{pragma Convention}.
3603 The secondary stack size parameter, if given, is used to set the size
3604 of secondary stack for the thread. The secondary stack is allocated as
3605 a local variable of the expanded thread body subprogram, and thus is
3606 allocated out of the main thread stack size. If no secondary stack
3607 size parameter is present, the default size (from the declaration in
3608 @code{System.Secondary_Stack} is used.
3610 @node Pragma Time_Slice
3611 @unnumberedsec Pragma Time_Slice
3616 @smallexample @c ada
3617 pragma Time_Slice (static_duration_EXPRESSION);
3621 For implementations of GNAT on operating systems where it is possible
3622 to supply a time slice value, this pragma may be used for this purpose.
3623 It is ignored if it is used in a system that does not allow this control,
3624 or if it appears in other than the main program unit.
3626 Note that the effect of this pragma is identical to the effect of the
3627 DEC Ada 83 pragma of the same name when operating under OpenVMS systems.
3630 @unnumberedsec Pragma Title
3635 @smallexample @c ada
3636 pragma Title (TITLING_OPTION [, TITLING OPTION]);
3639 [Title =>] STRING_LITERAL,
3640 | [Subtitle =>] STRING_LITERAL
3644 Syntax checked but otherwise ignored by GNAT@. This is a listing control
3645 pragma used in DEC Ada 83 implementations to provide a title and/or
3646 subtitle for the program listing. The program listing generated by GNAT
3647 does not have titles or subtitles.
3649 Unlike other pragmas, the full flexibility of named notation is allowed
3650 for this pragma, i.e.@: the parameters may be given in any order if named
3651 notation is used, and named and positional notation can be mixed
3652 following the normal rules for procedure calls in Ada.
3654 @node Pragma Unchecked_Union
3655 @unnumberedsec Pragma Unchecked_Union
3657 @findex Unchecked_Union
3661 @smallexample @c ada
3662 pragma Unchecked_Union (first_subtype_LOCAL_NAME);
3666 This pragma is used to declare that the specified type should be represented
3668 equivalent to a C union type, and is intended only for use in
3669 interfacing with C code that uses union types. In Ada terms, the named
3670 type must obey the following rules:
3674 It is a non-tagged non-limited record type.
3676 It has a single discrete discriminant with a default value.
3678 The component list consists of a single variant part.
3680 Each variant has a component list with a single component.
3682 No nested variants are allowed.
3684 No component has an explicit default value.
3686 No component has a non-static constraint.
3690 In addition, given a type that meets the above requirements, the
3691 following restrictions apply to its use throughout the program:
3695 The discriminant name can be mentioned only in an aggregate.
3697 No subtypes may be created of this type.
3699 The type may not be constrained by giving a discriminant value.
3701 The type cannot be passed as the actual for a generic formal with a
3706 Equality and inequality operations on @code{unchecked_unions} are not
3707 available, since there is no discriminant to compare and the compiler
3708 does not even know how many bits to compare. It is implementation
3709 dependent whether this is detected at compile time as an illegality or
3710 whether it is undetected and considered to be an erroneous construct. In
3711 GNAT, a direct comparison is illegal, but GNAT does not attempt to catch
3712 the composite case (where two composites are compared that contain an
3713 unchecked union component), so such comparisons are simply considered
3716 The layout of the resulting type corresponds exactly to a C union, where
3717 each branch of the union corresponds to a single variant in the Ada
3718 record. The semantics of the Ada program is not changed in any way by
3719 the pragma, i.e.@: provided the above restrictions are followed, and no
3720 erroneous incorrect references to fields or erroneous comparisons occur,
3721 the semantics is exactly as described by the Ada reference manual.
3722 Pragma @code{Suppress (Discriminant_Check)} applies implicitly to the
3723 type and the default convention is C.
3725 @node Pragma Unimplemented_Unit
3726 @unnumberedsec Pragma Unimplemented_Unit
3727 @findex Unimplemented_Unit
3731 @smallexample @c ada
3732 pragma Unimplemented_Unit;
3736 If this pragma occurs in a unit that is processed by the compiler, GNAT
3737 aborts with the message @samp{@var{xxx} not implemented}, where
3738 @var{xxx} is the name of the current compilation unit. This pragma is
3739 intended to allow the compiler to handle unimplemented library units in
3742 The abort only happens if code is being generated. Thus you can use
3743 specs of unimplemented packages in syntax or semantic checking mode.
3745 @node Pragma Universal_Data
3746 @unnumberedsec Pragma Universal_Data
3747 @findex Universal_Data
3751 @smallexample @c ada
3752 pragma Universal_Data [(library_unit_Name)];
3756 This pragma is supported only for the AAMP target and is ignored for
3757 other targets. The pragma specifies that all library-level objects
3758 (Counter 0 data) associated with the library unit are to be accessed
3759 and updated using universal addressing (24-bit addresses for AAMP5)
3760 rather than the default of 16-bit Data Environment (DENV) addressing.
3761 Use of this pragma will generally result in less efficient code for
3762 references to global data associated with the library unit, but
3763 allows such data to be located anywhere in memory. This pragma is
3764 a library unit pragma, but can also be used as a configuration pragma
3765 (including use in the @file{gnat.adc} file). The functionality
3766 of this pragma is also available by applying the -univ switch on the
3767 compilations of units where universal addressing of the data is desired.
3769 @node Pragma Unreferenced
3770 @unnumberedsec Pragma Unreferenced
3771 @findex Unreferenced
3772 @cindex Warnings, unreferenced
3776 @smallexample @c ada
3777 pragma Unreferenced (local_Name @{, local_Name@});
3781 This pragma signals that the entities whose names are listed are
3782 deliberately not referenced in the current source unit. This
3783 suppresses warnings about the
3784 entities being unreferenced, and in addition a warning will be
3785 generated if one of these entities is in fact referenced in the
3786 same unit as the pragma (or in the corresponding body, or one
3789 This is particularly useful for clearly signaling that a particular
3790 parameter is not referenced in some particular subprogram implementation
3791 and that this is deliberate. It can also be useful in the case of
3792 objects declared only for their initialization or finalization side
3795 If @code{local_Name} identifies more than one matching homonym in the
3796 current scope, then the entity most recently declared is the one to which
3799 The left hand side of an assignment does not count as a reference for the
3800 purpose of this pragma. Thus it is fine to assign to an entity for which
3801 pragma Unreferenced is given.
3803 @node Pragma Unreserve_All_Interrupts
3804 @unnumberedsec Pragma Unreserve_All_Interrupts
3805 @findex Unreserve_All_Interrupts
3809 @smallexample @c ada
3810 pragma Unreserve_All_Interrupts;
3814 Normally certain interrupts are reserved to the implementation. Any attempt
3815 to attach an interrupt causes Program_Error to be raised, as described in
3816 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
3817 many systems for a @kbd{Ctrl-C} interrupt. Normally this interrupt is
3818 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
3819 interrupt execution.
3821 If the pragma @code{Unreserve_All_Interrupts} appears anywhere in any unit in
3822 a program, then all such interrupts are unreserved. This allows the
3823 program to handle these interrupts, but disables their standard
3824 functions. For example, if this pragma is used, then pressing
3825 @kbd{Ctrl-C} will not automatically interrupt execution. However,
3826 a program can then handle the @code{SIGINT} interrupt as it chooses.
3828 For a full list of the interrupts handled in a specific implementation,
3829 see the source code for the specification of @code{Ada.Interrupts.Names} in
3830 file @file{a-intnam.ads}. This is a target dependent file that contains the
3831 list of interrupts recognized for a given target. The documentation in
3832 this file also specifies what interrupts are affected by the use of
3833 the @code{Unreserve_All_Interrupts} pragma.
3835 For a more general facility for controlling what interrupts can be
3836 handled, see pragma @code{Interrupt_State}, which subsumes the functionality
3837 of the @code{Unreserve_All_Interrupts} pragma.
3839 @node Pragma Unsuppress
3840 @unnumberedsec Pragma Unsuppress
3845 @smallexample @c ada
3846 pragma Unsuppress (IDENTIFIER [, [On =>] NAME]);
3850 This pragma undoes the effect of a previous pragma @code{Suppress}. If
3851 there is no corresponding pragma @code{Suppress} in effect, it has no
3852 effect. The range of the effect is the same as for pragma
3853 @code{Suppress}. The meaning of the arguments is identical to that used
3854 in pragma @code{Suppress}.
3856 One important application is to ensure that checks are on in cases where
3857 code depends on the checks for its correct functioning, so that the code
3858 will compile correctly even if the compiler switches are set to suppress
3861 @node Pragma Use_VADS_Size
3862 @unnumberedsec Pragma Use_VADS_Size
3863 @cindex @code{Size}, VADS compatibility
3864 @findex Use_VADS_Size
3868 @smallexample @c ada
3869 pragma Use_VADS_Size;
3873 This is a configuration pragma. In a unit to which it applies, any use
3874 of the 'Size attribute is automatically interpreted as a use of the
3875 'VADS_Size attribute. Note that this may result in incorrect semantic
3876 processing of valid Ada 95 programs. This is intended to aid in the
3877 handling of legacy code which depends on the interpretation of Size
3878 as implemented in the VADS compiler. See description of the VADS_Size
3879 attribute for further details.
3881 @node Pragma Validity_Checks
3882 @unnumberedsec Pragma Validity_Checks
3883 @findex Validity_Checks
3887 @smallexample @c ada
3888 pragma Validity_Checks (string_LITERAL | ALL_CHECKS | On | Off);
3892 This pragma is used in conjunction with compiler switches to control the
3893 built-in validity checking provided by GNAT@. The compiler switches, if set
3894 provide an initial setting for the switches, and this pragma may be used
3895 to modify these settings, or the settings may be provided entirely by
3896 the use of the pragma. This pragma can be used anywhere that a pragma
3897 is legal, including use as a configuration pragma (including use in
3898 the @file{gnat.adc} file).
3900 The form with a string literal specifies which validity options are to be
3901 activated. The validity checks are first set to include only the default
3902 reference manual settings, and then a string of letters in the string
3903 specifies the exact set of options required. The form of this string
3904 is exactly as described for the @code{-gnatVx} compiler switch (see the
3905 GNAT users guide for details). For example the following two methods
3906 can be used to enable validity checking for mode @code{in} and
3907 @code{in out} subprogram parameters:
3911 @smallexample @c ada
3912 pragma Validity_Checks ("im");
3917 gcc -c -gnatVim @dots{}
3922 The form ALL_CHECKS activates all standard checks (its use is equivalent
3923 to the use of the @code{gnatva} switch.
3925 The forms with @code{Off} and @code{On}
3926 can be used to temporarily disable validity checks
3927 as shown in the following example:
3929 @smallexample @c ada
3933 pragma Validity_Checks ("c"); -- validity checks for copies
3934 pragma Validity_Checks (Off); -- turn off validity checks
3935 A := B; -- B will not be validity checked
3936 pragma Validity_Checks (On); -- turn validity checks back on
3937 A := C; -- C will be validity checked
3940 @node Pragma Volatile
3941 @unnumberedsec Pragma Volatile
3946 @smallexample @c ada
3947 pragma Volatile (local_NAME);
3951 This pragma is defined by the Ada 95 Reference Manual, and the GNAT
3952 implementation is fully conformant with this definition. The reason it
3953 is mentioned in this section is that a pragma of the same name was supplied
3954 in some Ada 83 compilers, including DEC Ada 83. The Ada 95 implementation
3955 of pragma Volatile is upwards compatible with the implementation in
3958 @node Pragma Warnings
3959 @unnumberedsec Pragma Warnings
3964 @smallexample @c ada
3965 pragma Warnings (On | Off [, LOCAL_NAME]);
3969 Normally warnings are enabled, with the output being controlled by
3970 the command line switch. Warnings (@code{Off}) turns off generation of
3971 warnings until a Warnings (@code{On}) is encountered or the end of the
3972 current unit. If generation of warnings is turned off using this
3973 pragma, then no warning messages are output, regardless of the
3974 setting of the command line switches.
3976 The form with a single argument is a configuration pragma.
3978 If the @var{local_name} parameter is present, warnings are suppressed for
3979 the specified entity. This suppression is effective from the point where
3980 it occurs till the end of the extended scope of the variable (similar to
3981 the scope of @code{Suppress}).
3983 @node Pragma Weak_External
3984 @unnumberedsec Pragma Weak_External
3985 @findex Weak_External
3989 @smallexample @c ada
3990 pragma Weak_External ([Entity =>] LOCAL_NAME);
3994 This pragma specifies that the given entity should be marked as a weak
3995 external (one that does not have to be resolved) for the linker. For
3996 further details, consult the GCC manual.
3998 @node Implementation Defined Attributes
3999 @chapter Implementation Defined Attributes
4000 Ada 95 defines (throughout the Ada 95 reference manual,
4001 summarized in annex K),
4002 a set of attributes that provide useful additional functionality in all
4003 areas of the language. These language defined attributes are implemented
4004 in GNAT and work as described in the Ada 95 Reference Manual.
4006 In addition, Ada 95 allows implementations to define additional
4007 attributes whose meaning is defined by the implementation. GNAT provides
4008 a number of these implementation-dependent attributes which can be used
4009 to extend and enhance the functionality of the compiler. This section of
4010 the GNAT reference manual describes these additional attributes.
4012 Note that any program using these attributes may not be portable to
4013 other compilers (although GNAT implements this set of attributes on all
4014 platforms). Therefore if portability to other compilers is an important
4015 consideration, you should minimize the use of these attributes.
4026 * Default_Bit_Order::
4034 * Has_Access_Values::
4035 * Has_Discriminants::
4041 * Max_Interrupt_Priority::
4043 * Maximum_Alignment::
4047 * Passed_By_Reference::
4058 * Unconstrained_Array::
4059 * Universal_Literal_String::
4060 * Unrestricted_Access::
4068 @unnumberedsec Abort_Signal
4069 @findex Abort_Signal
4071 @code{Standard'Abort_Signal} (@code{Standard} is the only allowed
4072 prefix) provides the entity for the special exception used to signal
4073 task abort or asynchronous transfer of control. Normally this attribute
4074 should only be used in the tasking runtime (it is highly peculiar, and
4075 completely outside the normal semantics of Ada, for a user program to
4076 intercept the abort exception).
4079 @unnumberedsec Address_Size
4080 @cindex Size of @code{Address}
4081 @findex Address_Size
4083 @code{Standard'Address_Size} (@code{Standard} is the only allowed
4084 prefix) is a static constant giving the number of bits in an
4085 @code{Address}. It is the same value as System.Address'Size,
4086 but has the advantage of being static, while a direct
4087 reference to System.Address'Size is non-static because Address
4091 @unnumberedsec Asm_Input
4094 The @code{Asm_Input} attribute denotes a function that takes two
4095 parameters. The first is a string, the second is an expression of the
4096 type designated by the prefix. The first (string) argument is required
4097 to be a static expression, and is the constraint for the parameter,
4098 (e.g.@: what kind of register is required). The second argument is the
4099 value to be used as the input argument. The possible values for the
4100 constant are the same as those used in the RTL, and are dependent on
4101 the configuration file used to built the GCC back end.
4102 @ref{Machine Code Insertions}
4105 @unnumberedsec Asm_Output
4108 The @code{Asm_Output} attribute denotes a function that takes two
4109 parameters. The first is a string, the second is the name of a variable
4110 of the type designated by the attribute prefix. The first (string)
4111 argument is required to be a static expression and designates the
4112 constraint for the parameter (e.g.@: what kind of register is
4113 required). The second argument is the variable to be updated with the
4114 result. The possible values for constraint are the same as those used in
4115 the RTL, and are dependent on the configuration file used to build the
4116 GCC back end. If there are no output operands, then this argument may
4117 either be omitted, or explicitly given as @code{No_Output_Operands}.
4118 @ref{Machine Code Insertions}
4121 @unnumberedsec AST_Entry
4125 This attribute is implemented only in OpenVMS versions of GNAT@. Applied to
4126 the name of an entry, it yields a value of the predefined type AST_Handler
4127 (declared in the predefined package System, as extended by the use of
4128 pragma @code{Extend_System (Aux_DEC)}). This value enables the given entry to
4129 be called when an AST occurs. For further details, refer to the @cite{DEC Ada
4130 Language Reference Manual}, section 9.12a.
4135 @code{@var{obj}'Bit}, where @var{obj} is any object, yields the bit
4136 offset within the storage unit (byte) that contains the first bit of
4137 storage allocated for the object. The value of this attribute is of the
4138 type @code{Universal_Integer}, and is always a non-negative number not
4139 exceeding the value of @code{System.Storage_Unit}.
4141 For an object that is a variable or a constant allocated in a register,
4142 the value is zero. (The use of this attribute does not force the
4143 allocation of a variable to memory).
4145 For an object that is a formal parameter, this attribute applies
4146 to either the matching actual parameter or to a copy of the
4147 matching actual parameter.
4149 For an access object the value is zero. Note that
4150 @code{@var{obj}.all'Bit} is subject to an @code{Access_Check} for the
4151 designated object. Similarly for a record component
4152 @code{@var{X}.@var{C}'Bit} is subject to a discriminant check and
4153 @code{@var{X}(@var{I}).Bit} and @code{@var{X}(@var{I1}..@var{I2})'Bit}
4154 are subject to index checks.
4156 This attribute is designed to be compatible with the DEC Ada 83 definition
4157 and implementation of the @code{Bit} attribute.
4160 @unnumberedsec Bit_Position
4161 @findex Bit_Position
4163 @code{@var{R.C}'Bit}, where @var{R} is a record object and C is one
4164 of the fields of the record type, yields the bit
4165 offset within the record contains the first bit of
4166 storage allocated for the object. The value of this attribute is of the
4167 type @code{Universal_Integer}. The value depends only on the field
4168 @var{C} and is independent of the alignment of
4169 the containing record @var{R}.
4172 @unnumberedsec Code_Address
4173 @findex Code_Address
4174 @cindex Subprogram address
4175 @cindex Address of subprogram code
4178 attribute may be applied to subprograms in Ada 95, but the
4179 intended effect from the Ada 95 reference manual seems to be to provide
4180 an address value which can be used to call the subprogram by means of
4181 an address clause as in the following example:
4183 @smallexample @c ada
4184 procedure K is @dots{}
4187 for L'Address use K'Address;
4188 pragma Import (Ada, L);
4192 A call to @code{L} is then expected to result in a call to @code{K}@.
4193 In Ada 83, where there were no access-to-subprogram values, this was
4194 a common work around for getting the effect of an indirect call.
4195 GNAT implements the above use of @code{Address} and the technique
4196 illustrated by the example code works correctly.
4198 However, for some purposes, it is useful to have the address of the start
4199 of the generated code for the subprogram. On some architectures, this is
4200 not necessarily the same as the @code{Address} value described above.
4201 For example, the @code{Address} value may reference a subprogram
4202 descriptor rather than the subprogram itself.
4204 The @code{'Code_Address} attribute, which can only be applied to
4205 subprogram entities, always returns the address of the start of the
4206 generated code of the specified subprogram, which may or may not be
4207 the same value as is returned by the corresponding @code{'Address}
4210 @node Default_Bit_Order
4211 @unnumberedsec Default_Bit_Order
4213 @cindex Little endian
4214 @findex Default_Bit_Order
4216 @code{Standard'Default_Bit_Order} (@code{Standard} is the only
4217 permissible prefix), provides the value @code{System.Default_Bit_Order}
4218 as a @code{Pos} value (0 for @code{High_Order_First}, 1 for
4219 @code{Low_Order_First}). This is used to construct the definition of
4220 @code{Default_Bit_Order} in package @code{System}.
4223 @unnumberedsec Elaborated
4226 The prefix of the @code{'Elaborated} attribute must be a unit name. The
4227 value is a Boolean which indicates whether or not the given unit has been
4228 elaborated. This attribute is primarily intended for internal use by the
4229 generated code for dynamic elaboration checking, but it can also be used
4230 in user programs. The value will always be True once elaboration of all
4231 units has been completed. An exception is for units which need no
4232 elaboration, the value is always False for such units.
4235 @unnumberedsec Elab_Body
4238 This attribute can only be applied to a program unit name. It returns
4239 the entity for the corresponding elaboration procedure for elaborating
4240 the body of the referenced unit. This is used in the main generated
4241 elaboration procedure by the binder and is not normally used in any
4242 other context. However, there may be specialized situations in which it
4243 is useful to be able to call this elaboration procedure from Ada code,
4244 e.g.@: if it is necessary to do selective re-elaboration to fix some
4248 @unnumberedsec Elab_Spec
4251 This attribute can only be applied to a program unit name. It returns
4252 the entity for the corresponding elaboration procedure for elaborating
4253 the specification of the referenced unit. This is used in the main
4254 generated elaboration procedure by the binder and is not normally used
4255 in any other context. However, there may be specialized situations in
4256 which it is useful to be able to call this elaboration procedure from
4257 Ada code, e.g.@: if it is necessary to do selective re-elaboration to fix
4262 @cindex Ada 83 attributes
4265 The @code{Emax} attribute is provided for compatibility with Ada 83. See
4266 the Ada 83 reference manual for an exact description of the semantics of
4270 @unnumberedsec Enum_Rep
4271 @cindex Representation of enums
4274 For every enumeration subtype @var{S}, @code{@var{S}'Enum_Rep} denotes a
4275 function with the following spec:
4277 @smallexample @c ada
4278 function @var{S}'Enum_Rep (Arg : @var{S}'Base)
4279 return @i{Universal_Integer};
4283 It is also allowable to apply @code{Enum_Rep} directly to an object of an
4284 enumeration type or to a non-overloaded enumeration
4285 literal. In this case @code{@var{S}'Enum_Rep} is equivalent to
4286 @code{@var{typ}'Enum_Rep(@var{S})} where @var{typ} is the type of the
4287 enumeration literal or object.
4289 The function returns the representation value for the given enumeration
4290 value. This will be equal to value of the @code{Pos} attribute in the
4291 absence of an enumeration representation clause. This is a static
4292 attribute (i.e.@: the result is static if the argument is static).
4294 @code{@var{S}'Enum_Rep} can also be used with integer types and objects,
4295 in which case it simply returns the integer value. The reason for this
4296 is to allow it to be used for @code{(<>)} discrete formal arguments in
4297 a generic unit that can be instantiated with either enumeration types
4298 or integer types. Note that if @code{Enum_Rep} is used on a modular
4299 type whose upper bound exceeds the upper bound of the largest signed
4300 integer type, and the argument is a variable, so that the universal
4301 integer calculation is done at run-time, then the call to @code{Enum_Rep}
4302 may raise @code{Constraint_Error}.
4305 @unnumberedsec Epsilon
4306 @cindex Ada 83 attributes
4309 The @code{Epsilon} attribute is provided for compatibility with Ada 83. See
4310 the Ada 83 reference manual for an exact description of the semantics of
4314 @unnumberedsec Fixed_Value
4317 For every fixed-point type @var{S}, @code{@var{S}'Fixed_Value} denotes a
4318 function with the following specification:
4320 @smallexample @c ada
4321 function @var{S}'Fixed_Value (Arg : @i{Universal_Integer})
4326 The value returned is the fixed-point value @var{V} such that
4328 @smallexample @c ada
4329 @var{V} = Arg * @var{S}'Small
4333 The effect is thus similar to first converting the argument to the
4334 integer type used to represent @var{S}, and then doing an unchecked
4335 conversion to the fixed-point type. The difference is
4336 that there are full range checks, to ensure that the result is in range.
4337 This attribute is primarily intended for use in implementation of the
4338 input-output functions for fixed-point values.
4340 @node Has_Access_Values
4341 @unnumberedsec Has_Access_Values
4342 @cindex Access values, testing for
4343 @findex Has_Access_Values
4345 The prefix of the @code{Has_Access_Values} attribute is a type. The result
4346 is a Boolean value which is True if the is an access type, or is a composite
4347 type with a component (at any nesting depth) that is an access type, and is
4349 The intended use of this attribute is in conjunction with generic
4350 definitions. If the attribute is applied to a generic private type, it
4351 indicates whether or not the corresponding actual type has access values.
4353 @node Has_Discriminants
4354 @unnumberedsec Has_Discriminants
4355 @cindex Discriminants, testing for
4356 @findex Has_Discriminants
4358 The prefix of the @code{Has_Discriminants} attribute is a type. The result
4359 is a Boolean value which is True if the type has discriminants, and False
4360 otherwise. The intended use of this attribute is in conjunction with generic
4361 definitions. If the attribute is applied to a generic private type, it
4362 indicates whether or not the corresponding actual type has discriminants.
4368 The @code{Img} attribute differs from @code{Image} in that it may be
4369 applied to objects as well as types, in which case it gives the
4370 @code{Image} for the subtype of the object. This is convenient for
4373 @smallexample @c ada
4374 Put_Line ("X = " & X'Img);
4378 has the same meaning as the more verbose:
4380 @smallexample @c ada
4381 Put_Line ("X = " & @var{T}'Image (X));
4385 where @var{T} is the (sub)type of the object @code{X}.
4388 @unnumberedsec Integer_Value
4389 @findex Integer_Value
4391 For every integer type @var{S}, @code{@var{S}'Integer_Value} denotes a
4392 function with the following spec:
4394 @smallexample @c ada
4395 function @var{S}'Integer_Value (Arg : @i{Universal_Fixed})
4400 The value returned is the integer value @var{V}, such that
4402 @smallexample @c ada
4403 Arg = @var{V} * @var{T}'Small
4407 where @var{T} is the type of @code{Arg}.
4408 The effect is thus similar to first doing an unchecked conversion from
4409 the fixed-point type to its corresponding implementation type, and then
4410 converting the result to the target integer type. The difference is
4411 that there are full range checks, to ensure that the result is in range.
4412 This attribute is primarily intended for use in implementation of the
4413 standard input-output functions for fixed-point values.
4416 @unnumberedsec Large
4417 @cindex Ada 83 attributes
4420 The @code{Large} attribute is provided for compatibility with Ada 83. See
4421 the Ada 83 reference manual for an exact description of the semantics of
4425 @unnumberedsec Machine_Size
4426 @findex Machine_Size
4428 This attribute is identical to the @code{Object_Size} attribute. It is
4429 provided for compatibility with the DEC Ada 83 attribute of this name.
4432 @unnumberedsec Mantissa
4433 @cindex Ada 83 attributes
4436 The @code{Mantissa} attribute is provided for compatibility with Ada 83. See
4437 the Ada 83 reference manual for an exact description of the semantics of
4440 @node Max_Interrupt_Priority
4441 @unnumberedsec Max_Interrupt_Priority
4442 @cindex Interrupt priority, maximum
4443 @findex Max_Interrupt_Priority
4445 @code{Standard'Max_Interrupt_Priority} (@code{Standard} is the only
4446 permissible prefix), provides the same value as
4447 @code{System.Max_Interrupt_Priority}.
4450 @unnumberedsec Max_Priority
4451 @cindex Priority, maximum
4452 @findex Max_Priority
4454 @code{Standard'Max_Priority} (@code{Standard} is the only permissible
4455 prefix) provides the same value as @code{System.Max_Priority}.
4457 @node Maximum_Alignment
4458 @unnumberedsec Maximum_Alignment
4459 @cindex Alignment, maximum
4460 @findex Maximum_Alignment
4462 @code{Standard'Maximum_Alignment} (@code{Standard} is the only
4463 permissible prefix) provides the maximum useful alignment value for the
4464 target. This is a static value that can be used to specify the alignment
4465 for an object, guaranteeing that it is properly aligned in all
4468 @node Mechanism_Code
4469 @unnumberedsec Mechanism_Code
4470 @cindex Return values, passing mechanism
4471 @cindex Parameters, passing mechanism
4472 @findex Mechanism_Code
4474 @code{@var{function}'Mechanism_Code} yields an integer code for the
4475 mechanism used for the result of function, and
4476 @code{@var{subprogram}'Mechanism_Code (@var{n})} yields the mechanism
4477 used for formal parameter number @var{n} (a static integer value with 1
4478 meaning the first parameter) of @var{subprogram}. The code returned is:
4486 by descriptor (default descriptor class)
4488 by descriptor (UBS: unaligned bit string)
4490 by descriptor (UBSB: aligned bit string with arbitrary bounds)
4492 by descriptor (UBA: unaligned bit array)
4494 by descriptor (S: string, also scalar access type parameter)
4496 by descriptor (SB: string with arbitrary bounds)
4498 by descriptor (A: contiguous array)
4500 by descriptor (NCA: non-contiguous array)
4504 Values from 3 through 10 are only relevant to Digital OpenVMS implementations.
4507 @node Null_Parameter
4508 @unnumberedsec Null_Parameter
4509 @cindex Zero address, passing
4510 @findex Null_Parameter
4512 A reference @code{@var{T}'Null_Parameter} denotes an imaginary object of
4513 type or subtype @var{T} allocated at machine address zero. The attribute
4514 is allowed only as the default expression of a formal parameter, or as
4515 an actual expression of a subprogram call. In either case, the
4516 subprogram must be imported.
4518 The identity of the object is represented by the address zero in the
4519 argument list, independent of the passing mechanism (explicit or
4522 This capability is needed to specify that a zero address should be
4523 passed for a record or other composite object passed by reference.
4524 There is no way of indicating this without the @code{Null_Parameter}
4528 @unnumberedsec Object_Size
4529 @cindex Size, used for objects
4532 The size of an object is not necessarily the same as the size of the type
4533 of an object. This is because by default object sizes are increased to be
4534 a multiple of the alignment of the object. For example,
4535 @code{Natural'Size} is
4536 31, but by default objects of type @code{Natural} will have a size of 32 bits.
4537 Similarly, a record containing an integer and a character:
4539 @smallexample @c ada
4547 will have a size of 40 (that is @code{Rec'Size} will be 40. The
4548 alignment will be 4, because of the
4549 integer field, and so the default size of record objects for this type
4550 will be 64 (8 bytes).
4552 The @code{@var{type}'Object_Size} attribute
4553 has been added to GNAT to allow the
4554 default object size of a type to be easily determined. For example,
4555 @code{Natural'Object_Size} is 32, and
4556 @code{Rec'Object_Size} (for the record type in the above example) will be
4557 64. Note also that, unlike the situation with the
4558 @code{Size} attribute as defined in the Ada RM, the
4559 @code{Object_Size} attribute can be specified individually
4560 for different subtypes. For example:
4562 @smallexample @c ada
4563 type R is new Integer;
4564 subtype R1 is R range 1 .. 10;
4565 subtype R2 is R range 1 .. 10;
4566 for R2'Object_Size use 8;
4570 In this example, @code{R'Object_Size} and @code{R1'Object_Size} are both
4571 32 since the default object size for a subtype is the same as the object size
4572 for the parent subtype. This means that objects of type @code{R}
4574 by default be 32 bits (four bytes). But objects of type
4575 @code{R2} will be only
4576 8 bits (one byte), since @code{R2'Object_Size} has been set to 8.
4578 @node Passed_By_Reference
4579 @unnumberedsec Passed_By_Reference
4580 @cindex Parameters, when passed by reference
4581 @findex Passed_By_Reference
4583 @code{@var{type}'Passed_By_Reference} for any subtype @var{type} returns
4584 a value of type @code{Boolean} value that is @code{True} if the type is
4585 normally passed by reference and @code{False} if the type is normally
4586 passed by copy in calls. For scalar types, the result is always @code{False}
4587 and is static. For non-scalar types, the result is non-static.
4590 @unnumberedsec Range_Length
4591 @findex Range_Length
4593 @code{@var{type}'Range_Length} for any discrete type @var{type} yields
4594 the number of values represented by the subtype (zero for a null
4595 range). The result is static for static subtypes. @code{Range_Length}
4596 applied to the index subtype of a one dimensional array always gives the
4597 same result as @code{Range} applied to the array itself.
4600 @unnumberedsec Safe_Emax
4601 @cindex Ada 83 attributes
4604 The @code{Safe_Emax} attribute is provided for compatibility with Ada 83. See
4605 the Ada 83 reference manual for an exact description of the semantics of
4609 @unnumberedsec Safe_Large
4610 @cindex Ada 83 attributes
4613 The @code{Safe_Large} attribute is provided for compatibility with Ada 83. See
4614 the Ada 83 reference manual for an exact description of the semantics of
4618 @unnumberedsec Small
4619 @cindex Ada 83 attributes
4622 The @code{Small} attribute is defined in Ada 95 only for fixed-point types.
4623 GNAT also allows this attribute to be applied to floating-point types
4624 for compatibility with Ada 83. See
4625 the Ada 83 reference manual for an exact description of the semantics of
4626 this attribute when applied to floating-point types.
4629 @unnumberedsec Storage_Unit
4630 @findex Storage_Unit
4632 @code{Standard'Storage_Unit} (@code{Standard} is the only permissible
4633 prefix) provides the same value as @code{System.Storage_Unit}.
4636 @unnumberedsec Target_Name
4639 @code{Standard'Target_Name} (@code{Standard} is the only permissible
4640 prefix) provides a static string value that identifies the target
4641 for the current compilation. For GCC implementations, this is the
4642 standard gcc target name without the terminating slash (for
4643 example, GNAT 5.0 on windows yields "i586-pc-mingw32msv").
4649 @code{Standard'Tick} (@code{Standard} is the only permissible prefix)
4650 provides the same value as @code{System.Tick},
4653 @unnumberedsec To_Address
4656 The @code{System'To_Address}
4657 (@code{System} is the only permissible prefix)
4658 denotes a function identical to
4659 @code{System.Storage_Elements.To_Address} except that
4660 it is a static attribute. This means that if its argument is
4661 a static expression, then the result of the attribute is a
4662 static expression. The result is that such an expression can be
4663 used in contexts (e.g.@: preelaborable packages) which require a
4664 static expression and where the function call could not be used
4665 (since the function call is always non-static, even if its
4666 argument is static).
4669 @unnumberedsec Type_Class
4672 @code{@var{type}'Type_Class} for any type or subtype @var{type} yields
4673 the value of the type class for the full type of @var{type}. If
4674 @var{type} is a generic formal type, the value is the value for the
4675 corresponding actual subtype. The value of this attribute is of type
4676 @code{System.Aux_DEC.Type_Class}, which has the following definition:
4678 @smallexample @c ada
4680 (Type_Class_Enumeration,
4682 Type_Class_Fixed_Point,
4683 Type_Class_Floating_Point,
4688 Type_Class_Address);
4692 Protected types yield the value @code{Type_Class_Task}, which thus
4693 applies to all concurrent types. This attribute is designed to
4694 be compatible with the DEC Ada 83 attribute of the same name.
4697 @unnumberedsec UET_Address
4700 The @code{UET_Address} attribute can only be used for a prefix which
4701 denotes a library package. It yields the address of the unit exception
4702 table when zero cost exception handling is used. This attribute is
4703 intended only for use within the GNAT implementation. See the unit
4704 @code{Ada.Exceptions} in files @file{a-except.ads} and @file{a-except.adb}
4705 for details on how this attribute is used in the implementation.
4707 @node Unconstrained_Array
4708 @unnumberedsec Unconstrained_Array
4709 @findex Unconstrained_Array
4711 The @code{Unconstrained_Array} attribute can be used with a prefix that
4712 denotes any type or subtype. It is a static attribute that yields
4713 @code{True} if the prefix designates an unconstrained array,
4714 and @code{False} otherwise. In a generic instance, the result is
4715 still static, and yields the result of applying this test to the
4718 @node Universal_Literal_String
4719 @unnumberedsec Universal_Literal_String
4720 @cindex Named numbers, representation of
4721 @findex Universal_Literal_String
4723 The prefix of @code{Universal_Literal_String} must be a named
4724 number. The static result is the string consisting of the characters of
4725 the number as defined in the original source. This allows the user
4726 program to access the actual text of named numbers without intermediate
4727 conversions and without the need to enclose the strings in quotes (which
4728 would preclude their use as numbers). This is used internally for the
4729 construction of values of the floating-point attributes from the file
4730 @file{ttypef.ads}, but may also be used by user programs.
4732 @node Unrestricted_Access
4733 @unnumberedsec Unrestricted_Access
4734 @cindex @code{Access}, unrestricted
4735 @findex Unrestricted_Access
4737 The @code{Unrestricted_Access} attribute is similar to @code{Access}
4738 except that all accessibility and aliased view checks are omitted. This
4739 is a user-beware attribute. It is similar to
4740 @code{Address}, for which it is a desirable replacement where the value
4741 desired is an access type. In other words, its effect is identical to
4742 first applying the @code{Address} attribute and then doing an unchecked
4743 conversion to a desired access type. In GNAT, but not necessarily in
4744 other implementations, the use of static chains for inner level
4745 subprograms means that @code{Unrestricted_Access} applied to a
4746 subprogram yields a value that can be called as long as the subprogram
4747 is in scope (normal Ada 95 accessibility rules restrict this usage).
4749 It is possible to use @code{Unrestricted_Access} for any type, but care
4750 must be excercised if it is used to create pointers to unconstrained
4751 objects. In this case, the resulting pointer has the same scope as the
4752 context of the attribute, and may not be returned to some enclosing
4753 scope. For instance, a function cannot use @code{Unrestricted_Access}
4754 to create a unconstrained pointer and then return that value to the
4758 @unnumberedsec VADS_Size
4759 @cindex @code{Size}, VADS compatibility
4762 The @code{'VADS_Size} attribute is intended to make it easier to port
4763 legacy code which relies on the semantics of @code{'Size} as implemented
4764 by the VADS Ada 83 compiler. GNAT makes a best effort at duplicating the
4765 same semantic interpretation. In particular, @code{'VADS_Size} applied
4766 to a predefined or other primitive type with no Size clause yields the
4767 Object_Size (for example, @code{Natural'Size} is 32 rather than 31 on
4768 typical machines). In addition @code{'VADS_Size} applied to an object
4769 gives the result that would be obtained by applying the attribute to
4770 the corresponding type.
4773 @unnumberedsec Value_Size
4774 @cindex @code{Size}, setting for not-first subtype
4776 @code{@var{type}'Value_Size} is the number of bits required to represent
4777 a value of the given subtype. It is the same as @code{@var{type}'Size},
4778 but, unlike @code{Size}, may be set for non-first subtypes.
4781 @unnumberedsec Wchar_T_Size
4782 @findex Wchar_T_Size
4783 @code{Standard'Wchar_T_Size} (@code{Standard} is the only permissible
4784 prefix) provides the size in bits of the C @code{wchar_t} type
4785 primarily for constructing the definition of this type in
4786 package @code{Interfaces.C}.
4789 @unnumberedsec Word_Size
4791 @code{Standard'Word_Size} (@code{Standard} is the only permissible
4792 prefix) provides the value @code{System.Word_Size}.
4794 @c ------------------------
4795 @node Implementation Advice
4796 @chapter Implementation Advice
4798 The main text of the Ada 95 Reference Manual describes the required
4799 behavior of all Ada 95 compilers, and the GNAT compiler conforms to
4802 In addition, there are sections throughout the Ada 95
4803 reference manual headed
4804 by the phrase ``implementation advice''. These sections are not normative,
4805 i.e.@: they do not specify requirements that all compilers must
4806 follow. Rather they provide advice on generally desirable behavior. You
4807 may wonder why they are not requirements. The most typical answer is
4808 that they describe behavior that seems generally desirable, but cannot
4809 be provided on all systems, or which may be undesirable on some systems.
4811 As far as practical, GNAT follows the implementation advice sections in
4812 the Ada 95 Reference Manual. This chapter contains a table giving the
4813 reference manual section number, paragraph number and several keywords
4814 for each advice. Each entry consists of the text of the advice followed
4815 by the GNAT interpretation of this advice. Most often, this simply says
4816 ``followed'', which means that GNAT follows the advice. However, in a
4817 number of cases, GNAT deliberately deviates from this advice, in which
4818 case the text describes what GNAT does and why.
4820 @cindex Error detection
4821 @unnumberedsec 1.1.3(20): Error Detection
4824 If an implementation detects the use of an unsupported Specialized Needs
4825 Annex feature at run time, it should raise @code{Program_Error} if
4828 Not relevant. All specialized needs annex features are either supported,
4829 or diagnosed at compile time.
4832 @unnumberedsec 1.1.3(31): Child Units
4835 If an implementation wishes to provide implementation-defined
4836 extensions to the functionality of a language-defined library unit, it
4837 should normally do so by adding children to the library unit.
4841 @cindex Bounded errors
4842 @unnumberedsec 1.1.5(12): Bounded Errors
4845 If an implementation detects a bounded error or erroneous
4846 execution, it should raise @code{Program_Error}.
4848 Followed in all cases in which the implementation detects a bounded
4849 error or erroneous execution. Not all such situations are detected at
4853 @unnumberedsec 2.8(16): Pragmas
4856 Normally, implementation-defined pragmas should have no semantic effect
4857 for error-free programs; that is, if the implementation-defined pragmas
4858 are removed from a working program, the program should still be legal,
4859 and should still have the same semantics.
4861 The following implementation defined pragmas are exceptions to this
4873 @item CPP_Constructor
4881 @item Interface_Name
4883 @item Machine_Attribute
4885 @item Unimplemented_Unit
4887 @item Unchecked_Union
4892 In each of the above cases, it is essential to the purpose of the pragma
4893 that this advice not be followed. For details see the separate section
4894 on implementation defined pragmas.
4896 @unnumberedsec 2.8(17-19): Pragmas
4899 Normally, an implementation should not define pragmas that can
4900 make an illegal program legal, except as follows:
4904 A pragma used to complete a declaration, such as a pragma @code{Import};
4908 A pragma used to configure the environment by adding, removing, or
4909 replacing @code{library_items}.
4911 See response to paragraph 16 of this same section.
4913 @cindex Character Sets
4914 @cindex Alternative Character Sets
4915 @unnumberedsec 3.5.2(5): Alternative Character Sets
4918 If an implementation supports a mode with alternative interpretations
4919 for @code{Character} and @code{Wide_Character}, the set of graphic
4920 characters of @code{Character} should nevertheless remain a proper
4921 subset of the set of graphic characters of @code{Wide_Character}. Any
4922 character set ``localizations'' should be reflected in the results of
4923 the subprograms defined in the language-defined package
4924 @code{Characters.Handling} (see A.3) available in such a mode. In a mode with
4925 an alternative interpretation of @code{Character}, the implementation should
4926 also support a corresponding change in what is a legal
4927 @code{identifier_letter}.
4929 Not all wide character modes follow this advice, in particular the JIS
4930 and IEC modes reflect standard usage in Japan, and in these encoding,
4931 the upper half of the Latin-1 set is not part of the wide-character
4932 subset, since the most significant bit is used for wide character
4933 encoding. However, this only applies to the external forms. Internally
4934 there is no such restriction.
4936 @cindex Integer types
4937 @unnumberedsec 3.5.4(28): Integer Types
4941 An implementation should support @code{Long_Integer} in addition to
4942 @code{Integer} if the target machine supports 32-bit (or longer)
4943 arithmetic. No other named integer subtypes are recommended for package
4944 @code{Standard}. Instead, appropriate named integer subtypes should be
4945 provided in the library package @code{Interfaces} (see B.2).
4947 @code{Long_Integer} is supported. Other standard integer types are supported
4948 so this advice is not fully followed. These types
4949 are supported for convenient interface to C, and so that all hardware
4950 types of the machine are easily available.
4951 @unnumberedsec 3.5.4(29): Integer Types
4955 An implementation for a two's complement machine should support
4956 modular types with a binary modulus up to @code{System.Max_Int*2+2}. An
4957 implementation should support a non-binary modules up to @code{Integer'Last}.
4961 @cindex Enumeration values
4962 @unnumberedsec 3.5.5(8): Enumeration Values
4965 For the evaluation of a call on @code{@var{S}'Pos} for an enumeration
4966 subtype, if the value of the operand does not correspond to the internal
4967 code for any enumeration literal of its type (perhaps due to an
4968 un-initialized variable), then the implementation should raise
4969 @code{Program_Error}. This is particularly important for enumeration
4970 types with noncontiguous internal codes specified by an
4971 enumeration_representation_clause.
4976 @unnumberedsec 3.5.7(17): Float Types
4979 An implementation should support @code{Long_Float} in addition to
4980 @code{Float} if the target machine supports 11 or more digits of
4981 precision. No other named floating point subtypes are recommended for
4982 package @code{Standard}. Instead, appropriate named floating point subtypes
4983 should be provided in the library package @code{Interfaces} (see B.2).
4985 @code{Short_Float} and @code{Long_Long_Float} are also provided. The
4986 former provides improved compatibility with other implementations
4987 supporting this type. The latter corresponds to the highest precision
4988 floating-point type supported by the hardware. On most machines, this
4989 will be the same as @code{Long_Float}, but on some machines, it will
4990 correspond to the IEEE extended form. The notable case is all ia32
4991 (x86) implementations, where @code{Long_Long_Float} corresponds to
4992 the 80-bit extended precision format supported in hardware on this
4993 processor. Note that the 128-bit format on SPARC is not supported,
4994 since this is a software rather than a hardware format.
4996 @cindex Multidimensional arrays
4997 @cindex Arrays, multidimensional
4998 @unnumberedsec 3.6.2(11): Multidimensional Arrays
5001 An implementation should normally represent multidimensional arrays in
5002 row-major order, consistent with the notation used for multidimensional
5003 array aggregates (see 4.3.3). However, if a pragma @code{Convention}
5004 (@code{Fortran}, @dots{}) applies to a multidimensional array type, then
5005 column-major order should be used instead (see B.5, ``Interfacing with
5010 @findex Duration'Small
5011 @unnumberedsec 9.6(30-31): Duration'Small
5014 Whenever possible in an implementation, the value of @code{Duration'Small}
5015 should be no greater than 100 microseconds.
5017 Followed. (@code{Duration'Small} = 10**(@minus{}9)).
5021 The time base for @code{delay_relative_statements} should be monotonic;
5022 it need not be the same time base as used for @code{Calendar.Clock}.
5026 @unnumberedsec 10.2.1(12): Consistent Representation
5029 In an implementation, a type declared in a pre-elaborated package should
5030 have the same representation in every elaboration of a given version of
5031 the package, whether the elaborations occur in distinct executions of
5032 the same program, or in executions of distinct programs or partitions
5033 that include the given version.
5035 Followed, except in the case of tagged types. Tagged types involve
5036 implicit pointers to a local copy of a dispatch table, and these pointers
5037 have representations which thus depend on a particular elaboration of the
5038 package. It is not easy to see how it would be possible to follow this
5039 advice without severely impacting efficiency of execution.
5041 @cindex Exception information
5042 @unnumberedsec 11.4.1(19): Exception Information
5045 @code{Exception_Message} by default and @code{Exception_Information}
5046 should produce information useful for
5047 debugging. @code{Exception_Message} should be short, about one
5048 line. @code{Exception_Information} can be long. @code{Exception_Message}
5049 should not include the
5050 @code{Exception_Name}. @code{Exception_Information} should include both
5051 the @code{Exception_Name} and the @code{Exception_Message}.
5053 Followed. For each exception that doesn't have a specified
5054 @code{Exception_Message}, the compiler generates one containing the location
5055 of the raise statement. This location has the form ``file:line'', where
5056 file is the short file name (without path information) and line is the line
5057 number in the file. Note that in the case of the Zero Cost Exception
5058 mechanism, these messages become redundant with the Exception_Information that
5059 contains a full backtrace of the calling sequence, so they are disabled.
5060 To disable explicitly the generation of the source location message, use the
5061 Pragma @code{Discard_Names}.
5063 @cindex Suppression of checks
5064 @cindex Checks, suppression of
5065 @unnumberedsec 11.5(28): Suppression of Checks
5068 The implementation should minimize the code executed for checks that
5069 have been suppressed.
5073 @cindex Representation clauses
5074 @unnumberedsec 13.1 (21-24): Representation Clauses
5077 The recommended level of support for all representation items is
5078 qualified as follows:
5082 An implementation need not support representation items containing
5083 non-static expressions, except that an implementation should support a
5084 representation item for a given entity if each non-static expression in
5085 the representation item is a name that statically denotes a constant
5086 declared before the entity.
5088 Followed. GNAT does not support non-static expressions in representation
5089 clauses unless they are constants declared before the entity. For
5092 @smallexample @c ada
5094 for X'Address use To_address (16#2000#);
5098 will be rejected, since the To_Address expression is non-static. Instead
5101 @smallexample @c ada
5102 X_Address : constant Address : = To_Address (16#2000#);
5104 for X'Address use X_Address;
5109 An implementation need not support a specification for the @code{Size}
5110 for a given composite subtype, nor the size or storage place for an
5111 object (including a component) of a given composite subtype, unless the
5112 constraints on the subtype and its composite subcomponents (if any) are
5113 all static constraints.
5115 Followed. Size Clauses are not permitted on non-static components, as
5120 An aliased component, or a component whose type is by-reference, should
5121 always be allocated at an addressable location.
5125 @cindex Packed types
5126 @unnumberedsec 13.2(6-8): Packed Types
5129 If a type is packed, then the implementation should try to minimize
5130 storage allocated to objects of the type, possibly at the expense of
5131 speed of accessing components, subject to reasonable complexity in
5132 addressing calculations.
5136 The recommended level of support pragma @code{Pack} is:
5138 For a packed record type, the components should be packed as tightly as
5139 possible subject to the Sizes of the component subtypes, and subject to
5140 any @code{record_representation_clause} that applies to the type; the
5141 implementation may, but need not, reorder components or cross aligned
5142 word boundaries to improve the packing. A component whose @code{Size} is
5143 greater than the word size may be allocated an integral number of words.
5145 Followed. Tight packing of arrays is supported for all component sizes
5146 up to 64-bits. If the array component size is 1 (that is to say, if
5147 the component is a boolean type or an enumeration type with two values)
5148 then values of the type are implicitly initialized to zero. This
5149 happens both for objects of the packed type, and for objects that have a
5150 subcomponent of the packed type.
5154 An implementation should support Address clauses for imported
5158 @cindex @code{Address} clauses
5159 @unnumberedsec 13.3(14-19): Address Clauses
5163 For an array @var{X}, @code{@var{X}'Address} should point at the first
5164 component of the array, and not at the array bounds.
5170 The recommended level of support for the @code{Address} attribute is:
5172 @code{@var{X}'Address} should produce a useful result if @var{X} is an
5173 object that is aliased or of a by-reference type, or is an entity whose
5174 @code{Address} has been specified.
5176 Followed. A valid address will be produced even if none of those
5177 conditions have been met. If necessary, the object is forced into
5178 memory to ensure the address is valid.
5182 An implementation should support @code{Address} clauses for imported
5189 Objects (including subcomponents) that are aliased or of a by-reference
5190 type should be allocated on storage element boundaries.
5196 If the @code{Address} of an object is specified, or it is imported or exported,
5197 then the implementation should not perform optimizations based on
5198 assumptions of no aliases.
5202 @cindex @code{Alignment} clauses
5203 @unnumberedsec 13.3(29-35): Alignment Clauses
5206 The recommended level of support for the @code{Alignment} attribute for
5209 An implementation should support specified Alignments that are factors
5210 and multiples of the number of storage elements per word, subject to the
5217 An implementation need not support specified @code{Alignment}s for
5218 combinations of @code{Size}s and @code{Alignment}s that cannot be easily
5219 loaded and stored by available machine instructions.
5225 An implementation need not support specified @code{Alignment}s that are
5226 greater than the maximum @code{Alignment} the implementation ever returns by
5233 The recommended level of support for the @code{Alignment} attribute for
5236 Same as above, for subtypes, but in addition:
5242 For stand-alone library-level objects of statically constrained
5243 subtypes, the implementation should support all @code{Alignment}s
5244 supported by the target linker. For example, page alignment is likely to
5245 be supported for such objects, but not for subtypes.
5249 @cindex @code{Size} clauses
5250 @unnumberedsec 13.3(42-43): Size Clauses
5253 The recommended level of support for the @code{Size} attribute of
5256 A @code{Size} clause should be supported for an object if the specified
5257 @code{Size} is at least as large as its subtype's @code{Size}, and
5258 corresponds to a size in storage elements that is a multiple of the
5259 object's @code{Alignment} (if the @code{Alignment} is nonzero).
5263 @unnumberedsec 13.3(50-56): Size Clauses
5266 If the @code{Size} of a subtype is specified, and allows for efficient
5267 independent addressability (see 9.10) on the target architecture, then
5268 the @code{Size} of the following objects of the subtype should equal the
5269 @code{Size} of the subtype:
5271 Aliased objects (including components).
5277 @code{Size} clause on a composite subtype should not affect the
5278 internal layout of components.
5284 The recommended level of support for the @code{Size} attribute of subtypes is:
5288 The @code{Size} (if not specified) of a static discrete or fixed point
5289 subtype should be the number of bits needed to represent each value
5290 belonging to the subtype using an unbiased representation, leaving space
5291 for a sign bit only if the subtype contains negative values. If such a
5292 subtype is a first subtype, then an implementation should support a
5293 specified @code{Size} for it that reflects this representation.
5299 For a subtype implemented with levels of indirection, the @code{Size}
5300 should include the size of the pointers, but not the size of what they
5305 @cindex @code{Component_Size} clauses
5306 @unnumberedsec 13.3(71-73): Component Size Clauses
5309 The recommended level of support for the @code{Component_Size}
5314 An implementation need not support specified @code{Component_Sizes} that are
5315 less than the @code{Size} of the component subtype.
5321 An implementation should support specified @code{Component_Size}s that
5322 are factors and multiples of the word size. For such
5323 @code{Component_Size}s, the array should contain no gaps between
5324 components. For other @code{Component_Size}s (if supported), the array
5325 should contain no gaps between components when packing is also
5326 specified; the implementation should forbid this combination in cases
5327 where it cannot support a no-gaps representation.
5331 @cindex Enumeration representation clauses
5332 @cindex Representation clauses, enumeration
5333 @unnumberedsec 13.4(9-10): Enumeration Representation Clauses
5336 The recommended level of support for enumeration representation clauses
5339 An implementation need not support enumeration representation clauses
5340 for boolean types, but should at minimum support the internal codes in
5341 the range @code{System.Min_Int.System.Max_Int}.
5345 @cindex Record representation clauses
5346 @cindex Representation clauses, records
5347 @unnumberedsec 13.5.1(17-22): Record Representation Clauses
5350 The recommended level of support for
5351 @*@code{record_representation_clauses} is:
5353 An implementation should support storage places that can be extracted
5354 with a load, mask, shift sequence of machine code, and set with a load,
5355 shift, mask, store sequence, given the available machine instructions
5362 A storage place should be supported if its size is equal to the
5363 @code{Size} of the component subtype, and it starts and ends on a
5364 boundary that obeys the @code{Alignment} of the component subtype.
5370 If the default bit ordering applies to the declaration of a given type,
5371 then for a component whose subtype's @code{Size} is less than the word
5372 size, any storage place that does not cross an aligned word boundary
5373 should be supported.
5379 An implementation may reserve a storage place for the tag field of a
5380 tagged type, and disallow other components from overlapping that place.
5382 Followed. The storage place for the tag field is the beginning of the tagged
5383 record, and its size is Address'Size. GNAT will reject an explicit component
5384 clause for the tag field.
5388 An implementation need not support a @code{component_clause} for a
5389 component of an extension part if the storage place is not after the
5390 storage places of all components of the parent type, whether or not
5391 those storage places had been specified.
5393 Followed. The above advice on record representation clauses is followed,
5394 and all mentioned features are implemented.
5396 @cindex Storage place attributes
5397 @unnumberedsec 13.5.2(5): Storage Place Attributes
5400 If a component is represented using some form of pointer (such as an
5401 offset) to the actual data of the component, and this data is contiguous
5402 with the rest of the object, then the storage place attributes should
5403 reflect the place of the actual data, not the pointer. If a component is
5404 allocated discontinuously from the rest of the object, then a warning
5405 should be generated upon reference to one of its storage place
5408 Followed. There are no such components in GNAT@.
5410 @cindex Bit ordering
5411 @unnumberedsec 13.5.3(7-8): Bit Ordering
5414 The recommended level of support for the non-default bit ordering is:
5418 If @code{Word_Size} = @code{Storage_Unit}, then the implementation
5419 should support the non-default bit ordering in addition to the default
5422 Followed. Word size does not equal storage size in this implementation.
5423 Thus non-default bit ordering is not supported.
5425 @cindex @code{Address}, as private type
5426 @unnumberedsec 13.7(37): Address as Private
5429 @code{Address} should be of a private type.
5433 @cindex Operations, on @code{Address}
5434 @cindex @code{Address}, operations of
5435 @unnumberedsec 13.7.1(16): Address Operations
5438 Operations in @code{System} and its children should reflect the target
5439 environment semantics as closely as is reasonable. For example, on most
5440 machines, it makes sense for address arithmetic to ``wrap around''.
5441 Operations that do not make sense should raise @code{Program_Error}.
5443 Followed. Address arithmetic is modular arithmetic that wraps around. No
5444 operation raises @code{Program_Error}, since all operations make sense.
5446 @cindex Unchecked conversion
5447 @unnumberedsec 13.9(14-17): Unchecked Conversion
5450 The @code{Size} of an array object should not include its bounds; hence,
5451 the bounds should not be part of the converted data.
5457 The implementation should not generate unnecessary run-time checks to
5458 ensure that the representation of @var{S} is a representation of the
5459 target type. It should take advantage of the permission to return by
5460 reference when possible. Restrictions on unchecked conversions should be
5461 avoided unless required by the target environment.
5463 Followed. There are no restrictions on unchecked conversion. A warning is
5464 generated if the source and target types do not have the same size since
5465 the semantics in this case may be target dependent.
5469 The recommended level of support for unchecked conversions is:
5473 Unchecked conversions should be supported and should be reversible in
5474 the cases where this clause defines the result. To enable meaningful use
5475 of unchecked conversion, a contiguous representation should be used for
5476 elementary subtypes, for statically constrained array subtypes whose
5477 component subtype is one of the subtypes described in this paragraph,
5478 and for record subtypes without discriminants whose component subtypes
5479 are described in this paragraph.
5483 @cindex Heap usage, implicit
5484 @unnumberedsec 13.11(23-25): Implicit Heap Usage
5487 An implementation should document any cases in which it dynamically
5488 allocates heap storage for a purpose other than the evaluation of an
5491 Followed, the only other points at which heap storage is dynamically
5492 allocated are as follows:
5496 At initial elaboration time, to allocate dynamically sized global
5500 To allocate space for a task when a task is created.
5503 To extend the secondary stack dynamically when needed. The secondary
5504 stack is used for returning variable length results.
5509 A default (implementation-provided) storage pool for an
5510 access-to-constant type should not have overhead to support deallocation of
5517 A storage pool for an anonymous access type should be created at the
5518 point of an allocator for the type, and be reclaimed when the designated
5519 object becomes inaccessible.
5523 @cindex Unchecked deallocation
5524 @unnumberedsec 13.11.2(17): Unchecked De-allocation
5527 For a standard storage pool, @code{Free} should actually reclaim the
5532 @cindex Stream oriented attributes
5533 @unnumberedsec 13.13.2(17): Stream Oriented Attributes
5536 If a stream element is the same size as a storage element, then the
5537 normal in-memory representation should be used by @code{Read} and
5538 @code{Write} for scalar objects. Otherwise, @code{Read} and @code{Write}
5539 should use the smallest number of stream elements needed to represent
5540 all values in the base range of the scalar type.
5543 Followed. By default, GNAT uses the interpretation suggested by AI-195,
5544 which specifies using the size of the first subtype.
5545 However, such an implementation is based on direct binary
5546 representations and is therefore target- and endianness-dependent.
5547 To address this issue, GNAT also supplies an alternate implementation
5548 of the stream attributes @code{Read} and @code{Write},
5549 which uses the target-independent XDR standard representation
5551 @cindex XDR representation
5552 @cindex @code{Read} attribute
5553 @cindex @code{Write} attribute
5554 @cindex Stream oriented attributes
5555 The XDR implementation is provided as an alternative body of the
5556 @code{System.Stream_Attributes} package, in the file
5557 @file{s-strxdr.adb} in the GNAT library.
5558 There is no @file{s-strxdr.ads} file.
5559 In order to install the XDR implementation, do the following:
5561 @item Replace the default implementation of the
5562 @code{System.Stream_Attributes} package with the XDR implementation.
5563 For example on a Unix platform issue the commands:
5565 $ mv s-stratt.adb s-strold.adb
5566 $ mv s-strxdr.adb s-stratt.adb
5570 Rebuild the GNAT run-time library as documented in the
5571 @cite{GNAT User's Guide}
5574 @unnumberedsec A.1(52): Names of Predefined Numeric Types
5577 If an implementation provides additional named predefined integer types,
5578 then the names should end with @samp{Integer} as in
5579 @samp{Long_Integer}. If an implementation provides additional named
5580 predefined floating point types, then the names should end with
5581 @samp{Float} as in @samp{Long_Float}.
5585 @findex Ada.Characters.Handling
5586 @unnumberedsec A.3.2(49): @code{Ada.Characters.Handling}
5589 If an implementation provides a localized definition of @code{Character}
5590 or @code{Wide_Character}, then the effects of the subprograms in
5591 @code{Characters.Handling} should reflect the localizations. See also
5594 Followed. GNAT provides no such localized definitions.
5596 @cindex Bounded-length strings
5597 @unnumberedsec A.4.4(106): Bounded-Length String Handling
5600 Bounded string objects should not be implemented by implicit pointers
5601 and dynamic allocation.
5603 Followed. No implicit pointers or dynamic allocation are used.
5605 @cindex Random number generation
5606 @unnumberedsec A.5.2(46-47): Random Number Generation
5609 Any storage associated with an object of type @code{Generator} should be
5610 reclaimed on exit from the scope of the object.
5616 If the generator period is sufficiently long in relation to the number
5617 of distinct initiator values, then each possible value of
5618 @code{Initiator} passed to @code{Reset} should initiate a sequence of
5619 random numbers that does not, in a practical sense, overlap the sequence
5620 initiated by any other value. If this is not possible, then the mapping
5621 between initiator values and generator states should be a rapidly
5622 varying function of the initiator value.
5624 Followed. The generator period is sufficiently long for the first
5625 condition here to hold true.
5627 @findex Get_Immediate
5628 @unnumberedsec A.10.7(23): @code{Get_Immediate}
5631 The @code{Get_Immediate} procedures should be implemented with
5632 unbuffered input. For a device such as a keyboard, input should be
5633 @dfn{available} if a key has already been typed, whereas for a disk
5634 file, input should always be available except at end of file. For a file
5635 associated with a keyboard-like device, any line-editing features of the
5636 underlying operating system should be disabled during the execution of
5637 @code{Get_Immediate}.
5639 Followed on all targets except VxWorks. For VxWorks, there is no way to
5640 provide this functionality that does not result in the input buffer being
5641 flushed before the @code{Get_Immediate} call. A special unit
5642 @code{Interfaces.Vxworks.IO} is provided that contains routines to enable
5646 @unnumberedsec B.1(39-41): Pragma @code{Export}
5649 If an implementation supports pragma @code{Export} to a given language,
5650 then it should also allow the main subprogram to be written in that
5651 language. It should support some mechanism for invoking the elaboration
5652 of the Ada library units included in the system, and for invoking the
5653 finalization of the environment task. On typical systems, the
5654 recommended mechanism is to provide two subprograms whose link names are
5655 @code{adainit} and @code{adafinal}. @code{adainit} should contain the
5656 elaboration code for library units. @code{adafinal} should contain the
5657 finalization code. These subprograms should have no effect the second
5658 and subsequent time they are called.
5664 Automatic elaboration of pre-elaborated packages should be
5665 provided when pragma @code{Export} is supported.
5667 Followed when the main program is in Ada. If the main program is in a
5668 foreign language, then
5669 @code{adainit} must be called to elaborate pre-elaborated
5674 For each supported convention @var{L} other than @code{Intrinsic}, an
5675 implementation should support @code{Import} and @code{Export} pragmas
5676 for objects of @var{L}-compatible types and for subprograms, and pragma
5677 @code{Convention} for @var{L}-eligible types and for subprograms,
5678 presuming the other language has corresponding features. Pragma
5679 @code{Convention} need not be supported for scalar types.
5683 @cindex Package @code{Interfaces}
5685 @unnumberedsec B.2(12-13): Package @code{Interfaces}
5688 For each implementation-defined convention identifier, there should be a
5689 child package of package Interfaces with the corresponding name. This
5690 package should contain any declarations that would be useful for
5691 interfacing to the language (implementation) represented by the
5692 convention. Any declarations useful for interfacing to any language on
5693 the given hardware architecture should be provided directly in
5696 Followed. An additional package not defined
5697 in the Ada 95 Reference Manual is @code{Interfaces.CPP}, used
5698 for interfacing to C++.
5702 An implementation supporting an interface to C, COBOL, or Fortran should
5703 provide the corresponding package or packages described in the following
5706 Followed. GNAT provides all the packages described in this section.
5708 @cindex C, interfacing with
5709 @unnumberedsec B.3(63-71): Interfacing with C
5712 An implementation should support the following interface correspondences
5719 An Ada procedure corresponds to a void-returning C function.
5725 An Ada function corresponds to a non-void C function.
5731 An Ada @code{in} scalar parameter is passed as a scalar argument to a C
5738 An Ada @code{in} parameter of an access-to-object type with designated
5739 type @var{T} is passed as a @code{@var{t}*} argument to a C function,
5740 where @var{t} is the C type corresponding to the Ada type @var{T}.
5746 An Ada access @var{T} parameter, or an Ada @code{out} or @code{in out}
5747 parameter of an elementary type @var{T}, is passed as a @code{@var{t}*}
5748 argument to a C function, where @var{t} is the C type corresponding to
5749 the Ada type @var{T}. In the case of an elementary @code{out} or
5750 @code{in out} parameter, a pointer to a temporary copy is used to
5751 preserve by-copy semantics.
5757 An Ada parameter of a record type @var{T}, of any mode, is passed as a
5758 @code{@var{t}*} argument to a C function, where @var{t} is the C
5759 structure corresponding to the Ada type @var{T}.
5761 Followed. This convention may be overridden by the use of the C_Pass_By_Copy
5762 pragma, or Convention, or by explicitly specifying the mechanism for a given
5763 call using an extended import or export pragma.
5767 An Ada parameter of an array type with component type @var{T}, of any
5768 mode, is passed as a @code{@var{t}*} argument to a C function, where
5769 @var{t} is the C type corresponding to the Ada type @var{T}.
5775 An Ada parameter of an access-to-subprogram type is passed as a pointer
5776 to a C function whose prototype corresponds to the designated
5777 subprogram's specification.
5781 @cindex COBOL, interfacing with
5782 @unnumberedsec B.4(95-98): Interfacing with COBOL
5785 An Ada implementation should support the following interface
5786 correspondences between Ada and COBOL@.
5792 An Ada access @var{T} parameter is passed as a @samp{BY REFERENCE} data item of
5793 the COBOL type corresponding to @var{T}.
5799 An Ada in scalar parameter is passed as a @samp{BY CONTENT} data item of
5800 the corresponding COBOL type.
5806 Any other Ada parameter is passed as a @samp{BY REFERENCE} data item of the
5807 COBOL type corresponding to the Ada parameter type; for scalars, a local
5808 copy is used if necessary to ensure by-copy semantics.
5812 @cindex Fortran, interfacing with
5813 @unnumberedsec B.5(22-26): Interfacing with Fortran
5816 An Ada implementation should support the following interface
5817 correspondences between Ada and Fortran:
5823 An Ada procedure corresponds to a Fortran subroutine.
5829 An Ada function corresponds to a Fortran function.
5835 An Ada parameter of an elementary, array, or record type @var{T} is
5836 passed as a @var{T} argument to a Fortran procedure, where @var{T} is
5837 the Fortran type corresponding to the Ada type @var{T}, and where the
5838 INTENT attribute of the corresponding dummy argument matches the Ada
5839 formal parameter mode; the Fortran implementation's parameter passing
5840 conventions are used. For elementary types, a local copy is used if
5841 necessary to ensure by-copy semantics.
5847 An Ada parameter of an access-to-subprogram type is passed as a
5848 reference to a Fortran procedure whose interface corresponds to the
5849 designated subprogram's specification.
5853 @cindex Machine operations
5854 @unnumberedsec C.1(3-5): Access to Machine Operations
5857 The machine code or intrinsic support should allow access to all
5858 operations normally available to assembly language programmers for the
5859 target environment, including privileged instructions, if any.
5865 The interfacing pragmas (see Annex B) should support interface to
5866 assembler; the default assembler should be associated with the
5867 convention identifier @code{Assembler}.
5873 If an entity is exported to assembly language, then the implementation
5874 should allocate it at an addressable location, and should ensure that it
5875 is retained by the linking process, even if not otherwise referenced
5876 from the Ada code. The implementation should assume that any call to a
5877 machine code or assembler subprogram is allowed to read or update every
5878 object that is specified as exported.
5882 @unnumberedsec C.1(10-16): Access to Machine Operations
5885 The implementation should ensure that little or no overhead is
5886 associated with calling intrinsic and machine-code subprograms.
5888 Followed for both intrinsics and machine-code subprograms.
5892 It is recommended that intrinsic subprograms be provided for convenient
5893 access to any machine operations that provide special capabilities or
5894 efficiency and that are not otherwise available through the language
5897 Followed. A full set of machine operation intrinsic subprograms is provided.
5901 Atomic read-modify-write operations---e.g.@:, test and set, compare and
5902 swap, decrement and test, enqueue/dequeue.
5904 Followed on any target supporting such operations.
5908 Standard numeric functions---e.g.@:, sin, log.
5910 Followed on any target supporting such operations.
5914 String manipulation operations---e.g.@:, translate and test.
5916 Followed on any target supporting such operations.
5920 Vector operations---e.g.@:, compare vector against thresholds.
5922 Followed on any target supporting such operations.
5926 Direct operations on I/O ports.
5928 Followed on any target supporting such operations.
5930 @cindex Interrupt support
5931 @unnumberedsec C.3(28): Interrupt Support
5934 If the @code{Ceiling_Locking} policy is not in effect, the
5935 implementation should provide means for the application to specify which
5936 interrupts are to be blocked during protected actions, if the underlying
5937 system allows for a finer-grain control of interrupt blocking.
5939 Followed. The underlying system does not allow for finer-grain control
5940 of interrupt blocking.
5942 @cindex Protected procedure handlers
5943 @unnumberedsec C.3.1(20-21): Protected Procedure Handlers
5946 Whenever possible, the implementation should allow interrupt handlers to
5947 be called directly by the hardware.
5951 This is never possible under IRIX, so this is followed by default.
5953 Followed on any target where the underlying operating system permits
5958 Whenever practical, violations of any
5959 implementation-defined restrictions should be detected before run time.
5961 Followed. Compile time warnings are given when possible.
5963 @cindex Package @code{Interrupts}
5965 @unnumberedsec C.3.2(25): Package @code{Interrupts}
5969 If implementation-defined forms of interrupt handler procedures are
5970 supported, such as protected procedures with parameters, then for each
5971 such form of a handler, a type analogous to @code{Parameterless_Handler}
5972 should be specified in a child package of @code{Interrupts}, with the
5973 same operations as in the predefined package Interrupts.
5977 @cindex Pre-elaboration requirements
5978 @unnumberedsec C.4(14): Pre-elaboration Requirements
5981 It is recommended that pre-elaborated packages be implemented in such a
5982 way that there should be little or no code executed at run time for the
5983 elaboration of entities not already covered by the Implementation
5986 Followed. Executable code is generated in some cases, e.g.@: loops
5987 to initialize large arrays.
5989 @unnumberedsec C.5(8): Pragma @code{Discard_Names}
5993 If the pragma applies to an entity, then the implementation should
5994 reduce the amount of storage used for storing names associated with that
5999 @cindex Package @code{Task_Attributes}
6000 @findex Task_Attributes
6001 @unnumberedsec C.7.2(30): The Package Task_Attributes
6004 Some implementations are targeted to domains in which memory use at run
6005 time must be completely deterministic. For such implementations, it is
6006 recommended that the storage for task attributes will be pre-allocated
6007 statically and not from the heap. This can be accomplished by either
6008 placing restrictions on the number and the size of the task's
6009 attributes, or by using the pre-allocated storage for the first @var{N}
6010 attribute objects, and the heap for the others. In the latter case,
6011 @var{N} should be documented.
6013 Not followed. This implementation is not targeted to such a domain.
6015 @cindex Locking Policies
6016 @unnumberedsec D.3(17): Locking Policies
6020 The implementation should use names that end with @samp{_Locking} for
6021 locking policies defined by the implementation.
6023 Followed. A single implementation-defined locking policy is defined,
6024 whose name (@code{Inheritance_Locking}) follows this suggestion.
6026 @cindex Entry queuing policies
6027 @unnumberedsec D.4(16): Entry Queuing Policies
6030 Names that end with @samp{_Queuing} should be used
6031 for all implementation-defined queuing policies.
6033 Followed. No such implementation-defined queuing policies exist.
6035 @cindex Preemptive abort
6036 @unnumberedsec D.6(9-10): Preemptive Abort
6039 Even though the @code{abort_statement} is included in the list of
6040 potentially blocking operations (see 9.5.1), it is recommended that this
6041 statement be implemented in a way that never requires the task executing
6042 the @code{abort_statement} to block.
6048 On a multi-processor, the delay associated with aborting a task on
6049 another processor should be bounded; the implementation should use
6050 periodic polling, if necessary, to achieve this.
6054 @cindex Tasking restrictions
6055 @unnumberedsec D.7(21): Tasking Restrictions
6058 When feasible, the implementation should take advantage of the specified
6059 restrictions to produce a more efficient implementation.
6061 GNAT currently takes advantage of these restrictions by providing an optimized
6062 run time when the Ravenscar profile and the GNAT restricted run time set
6063 of restrictions are specified. See pragma @code{Profile (Ravenscar)} and
6064 pragma @code{Profile (Restricted)} for more details.
6066 @cindex Time, monotonic
6067 @unnumberedsec D.8(47-49): Monotonic Time
6070 When appropriate, implementations should provide configuration
6071 mechanisms to change the value of @code{Tick}.
6073 Such configuration mechanisms are not appropriate to this implementation
6074 and are thus not supported.
6078 It is recommended that @code{Calendar.Clock} and @code{Real_Time.Clock}
6079 be implemented as transformations of the same time base.
6085 It is recommended that the @dfn{best} time base which exists in
6086 the underlying system be available to the application through
6087 @code{Clock}. @dfn{Best} may mean highest accuracy or largest range.
6091 @cindex Partition communication subsystem
6093 @unnumberedsec E.5(28-29): Partition Communication Subsystem
6096 Whenever possible, the PCS on the called partition should allow for
6097 multiple tasks to call the RPC-receiver with different messages and
6098 should allow them to block until the corresponding subprogram body
6101 Followed by GLADE, a separately supplied PCS that can be used with
6106 The @code{Write} operation on a stream of type @code{Params_Stream_Type}
6107 should raise @code{Storage_Error} if it runs out of space trying to
6108 write the @code{Item} into the stream.
6110 Followed by GLADE, a separately supplied PCS that can be used with
6113 @cindex COBOL support
6114 @unnumberedsec F(7): COBOL Support
6117 If COBOL (respectively, C) is widely supported in the target
6118 environment, implementations supporting the Information Systems Annex
6119 should provide the child package @code{Interfaces.COBOL} (respectively,
6120 @code{Interfaces.C}) specified in Annex B and should support a
6121 @code{convention_identifier} of COBOL (respectively, C) in the interfacing
6122 pragmas (see Annex B), thus allowing Ada programs to interface with
6123 programs written in that language.
6127 @cindex Decimal radix support
6128 @unnumberedsec F.1(2): Decimal Radix Support
6131 Packed decimal should be used as the internal representation for objects
6132 of subtype @var{S} when @var{S}'Machine_Radix = 10.
6134 Not followed. GNAT ignores @var{S}'Machine_Radix and always uses binary
6138 @unnumberedsec G: Numerics
6141 If Fortran (respectively, C) is widely supported in the target
6142 environment, implementations supporting the Numerics Annex
6143 should provide the child package @code{Interfaces.Fortran} (respectively,
6144 @code{Interfaces.C}) specified in Annex B and should support a
6145 @code{convention_identifier} of Fortran (respectively, C) in the interfacing
6146 pragmas (see Annex B), thus allowing Ada programs to interface with
6147 programs written in that language.
6151 @cindex Complex types
6152 @unnumberedsec G.1.1(56-58): Complex Types
6155 Because the usual mathematical meaning of multiplication of a complex
6156 operand and a real operand is that of the scaling of both components of
6157 the former by the latter, an implementation should not perform this
6158 operation by first promoting the real operand to complex type and then
6159 performing a full complex multiplication. In systems that, in the
6160 future, support an Ada binding to IEC 559:1989, the latter technique
6161 will not generate the required result when one of the components of the
6162 complex operand is infinite. (Explicit multiplication of the infinite
6163 component by the zero component obtained during promotion yields a NaN
6164 that propagates into the final result.) Analogous advice applies in the
6165 case of multiplication of a complex operand and a pure-imaginary
6166 operand, and in the case of division of a complex operand by a real or
6167 pure-imaginary operand.
6173 Similarly, because the usual mathematical meaning of addition of a
6174 complex operand and a real operand is that the imaginary operand remains
6175 unchanged, an implementation should not perform this operation by first
6176 promoting the real operand to complex type and then performing a full
6177 complex addition. In implementations in which the @code{Signed_Zeros}
6178 attribute of the component type is @code{True} (and which therefore
6179 conform to IEC 559:1989 in regard to the handling of the sign of zero in
6180 predefined arithmetic operations), the latter technique will not
6181 generate the required result when the imaginary component of the complex
6182 operand is a negatively signed zero. (Explicit addition of the negative
6183 zero to the zero obtained during promotion yields a positive zero.)
6184 Analogous advice applies in the case of addition of a complex operand
6185 and a pure-imaginary operand, and in the case of subtraction of a
6186 complex operand and a real or pure-imaginary operand.
6192 Implementations in which @code{Real'Signed_Zeros} is @code{True} should
6193 attempt to provide a rational treatment of the signs of zero results and
6194 result components. As one example, the result of the @code{Argument}
6195 function should have the sign of the imaginary component of the
6196 parameter @code{X} when the point represented by that parameter lies on
6197 the positive real axis; as another, the sign of the imaginary component
6198 of the @code{Compose_From_Polar} function should be the same as
6199 (respectively, the opposite of) that of the @code{Argument} parameter when that
6200 parameter has a value of zero and the @code{Modulus} parameter has a
6201 nonnegative (respectively, negative) value.
6205 @cindex Complex elementary functions
6206 @unnumberedsec G.1.2(49): Complex Elementary Functions
6209 Implementations in which @code{Complex_Types.Real'Signed_Zeros} is
6210 @code{True} should attempt to provide a rational treatment of the signs
6211 of zero results and result components. For example, many of the complex
6212 elementary functions have components that are odd functions of one of
6213 the parameter components; in these cases, the result component should
6214 have the sign of the parameter component at the origin. Other complex
6215 elementary functions have zero components whose sign is opposite that of
6216 a parameter component at the origin, or is always positive or always
6221 @cindex Accuracy requirements
6222 @unnumberedsec G.2.4(19): Accuracy Requirements
6225 The versions of the forward trigonometric functions without a
6226 @code{Cycle} parameter should not be implemented by calling the
6227 corresponding version with a @code{Cycle} parameter of
6228 @code{2.0*Numerics.Pi}, since this will not provide the required
6229 accuracy in some portions of the domain. For the same reason, the
6230 version of @code{Log} without a @code{Base} parameter should not be
6231 implemented by calling the corresponding version with a @code{Base}
6232 parameter of @code{Numerics.e}.
6236 @cindex Complex arithmetic accuracy
6237 @cindex Accuracy, complex arithmetic
6238 @unnumberedsec G.2.6(15): Complex Arithmetic Accuracy
6242 The version of the @code{Compose_From_Polar} function without a
6243 @code{Cycle} parameter should not be implemented by calling the
6244 corresponding version with a @code{Cycle} parameter of
6245 @code{2.0*Numerics.Pi}, since this will not provide the required
6246 accuracy in some portions of the domain.
6250 @c -----------------------------------------
6251 @node Implementation Defined Characteristics
6252 @chapter Implementation Defined Characteristics
6255 In addition to the implementation dependent pragmas and attributes, and
6256 the implementation advice, there are a number of other features of Ada
6257 95 that are potentially implementation dependent. These are mentioned
6258 throughout the Ada 95 Reference Manual, and are summarized in annex M@.
6260 A requirement for conforming Ada compilers is that they provide
6261 documentation describing how the implementation deals with each of these
6262 issues. In this chapter, you will find each point in annex M listed
6263 followed by a description in italic font of how GNAT
6267 implementation on IRIX 5.3 operating system or greater
6269 handles the implementation dependence.
6271 You can use this chapter as a guide to minimizing implementation
6272 dependent features in your programs if portability to other compilers
6273 and other operating systems is an important consideration. The numbers
6274 in each section below correspond to the paragraph number in the Ada 95
6280 @strong{2}. Whether or not each recommendation given in Implementation
6281 Advice is followed. See 1.1.2(37).
6284 @xref{Implementation Advice}.
6289 @strong{3}. Capacity limitations of the implementation. See 1.1.3(3).
6292 The complexity of programs that can be processed is limited only by the
6293 total amount of available virtual memory, and disk space for the
6294 generated object files.
6299 @strong{4}. Variations from the standard that are impractical to avoid
6300 given the implementation's execution environment. See 1.1.3(6).
6303 There are no variations from the standard.
6308 @strong{5}. Which @code{code_statement}s cause external
6309 interactions. See 1.1.3(10).
6312 Any @code{code_statement} can potentially cause external interactions.
6317 @strong{6}. The coded representation for the text of an Ada
6318 program. See 2.1(4).
6321 See separate section on source representation.
6326 @strong{7}. The control functions allowed in comments. See 2.1(14).
6329 See separate section on source representation.
6334 @strong{8}. The representation for an end of line. See 2.2(2).
6337 See separate section on source representation.
6342 @strong{9}. Maximum supported line length and lexical element
6343 length. See 2.2(15).
6346 The maximum line length is 255 characters an the maximum length of a
6347 lexical element is also 255 characters.
6352 @strong{10}. Implementation defined pragmas. See 2.8(14).
6356 @xref{Implementation Defined Pragmas}.
6361 @strong{11}. Effect of pragma @code{Optimize}. See 2.8(27).
6364 Pragma @code{Optimize}, if given with a @code{Time} or @code{Space}
6365 parameter, checks that the optimization flag is set, and aborts if it is
6371 @strong{12}. The sequence of characters of the value returned by
6372 @code{@var{S}'Image} when some of the graphic characters of
6373 @code{@var{S}'Wide_Image} are not defined in @code{Character}. See
6377 The sequence of characters is as defined by the wide character encoding
6378 method used for the source. See section on source representation for
6384 @strong{13}. The predefined integer types declared in
6385 @code{Standard}. See 3.5.4(25).
6389 @item Short_Short_Integer
6392 (Short) 16 bit signed
6396 64 bit signed (Alpha OpenVMS only)
6397 32 bit signed (all other targets)
6398 @item Long_Long_Integer
6405 @strong{14}. Any nonstandard integer types and the operators defined
6406 for them. See 3.5.4(26).
6409 There are no nonstandard integer types.
6414 @strong{15}. Any nonstandard real types and the operators defined for
6418 There are no nonstandard real types.
6423 @strong{16}. What combinations of requested decimal precision and range
6424 are supported for floating point types. See 3.5.7(7).
6427 The precision and range is as defined by the IEEE standard.
6432 @strong{17}. The predefined floating point types declared in
6433 @code{Standard}. See 3.5.7(16).
6440 (Short) 32 bit IEEE short
6443 @item Long_Long_Float
6444 64 bit IEEE long (80 bit IEEE long on x86 processors)
6450 @strong{18}. The small of an ordinary fixed point type. See 3.5.9(8).
6453 @code{Fine_Delta} is 2**(@minus{}63)
6458 @strong{19}. What combinations of small, range, and digits are
6459 supported for fixed point types. See 3.5.9(10).
6462 Any combinations are permitted that do not result in a small less than
6463 @code{Fine_Delta} and do not result in a mantissa larger than 63 bits.
6464 If the mantissa is larger than 53 bits on machines where Long_Long_Float
6465 is 64 bits (true of all architectures except ia32), then the output from
6466 Text_IO is accurate to only 53 bits, rather than the full mantissa. This
6467 is because floating-point conversions are used to convert fixed point.
6472 @strong{20}. The result of @code{Tags.Expanded_Name} for types declared
6473 within an unnamed @code{block_statement}. See 3.9(10).
6476 Block numbers of the form @code{B@var{nnn}}, where @var{nnn} is a
6477 decimal integer are allocated.
6482 @strong{21}. Implementation-defined attributes. See 4.1.4(12).
6485 @xref{Implementation Defined Attributes}.
6490 @strong{22}. Any implementation-defined time types. See 9.6(6).
6493 There are no implementation-defined time types.
6498 @strong{23}. The time base associated with relative delays.
6501 See 9.6(20). The time base used is that provided by the C library
6502 function @code{gettimeofday}.
6507 @strong{24}. The time base of the type @code{Calendar.Time}. See
6511 The time base used is that provided by the C library function
6512 @code{gettimeofday}.
6517 @strong{25}. The time zone used for package @code{Calendar}
6518 operations. See 9.6(24).
6521 The time zone used by package @code{Calendar} is the current system time zone
6522 setting for local time, as accessed by the C library function
6528 @strong{26}. Any limit on @code{delay_until_statements} of
6529 @code{select_statements}. See 9.6(29).
6532 There are no such limits.
6537 @strong{27}. Whether or not two non overlapping parts of a composite
6538 object are independently addressable, in the case where packing, record
6539 layout, or @code{Component_Size} is specified for the object. See
6543 Separate components are independently addressable if they do not share
6544 overlapping storage units.
6549 @strong{28}. The representation for a compilation. See 10.1(2).
6552 A compilation is represented by a sequence of files presented to the
6553 compiler in a single invocation of the @code{gcc} command.
6558 @strong{29}. Any restrictions on compilations that contain multiple
6559 compilation_units. See 10.1(4).
6562 No single file can contain more than one compilation unit, but any
6563 sequence of files can be presented to the compiler as a single
6569 @strong{30}. The mechanisms for creating an environment and for adding
6570 and replacing compilation units. See 10.1.4(3).
6573 See separate section on compilation model.
6578 @strong{31}. The manner of explicitly assigning library units to a
6579 partition. See 10.2(2).
6582 If a unit contains an Ada main program, then the Ada units for the partition
6583 are determined by recursive application of the rules in the Ada Reference
6584 Manual section 10.2(2-6). In other words, the Ada units will be those that
6585 are needed by the main program, and then this definition of need is applied
6586 recursively to those units, and the partition contains the transitive
6587 closure determined by this relationship. In short, all the necessary units
6588 are included, with no need to explicitly specify the list. If additional
6589 units are required, e.g.@: by foreign language units, then all units must be
6590 mentioned in the context clause of one of the needed Ada units.
6592 If the partition contains no main program, or if the main program is in
6593 a language other than Ada, then GNAT
6594 provides the binder options @code{-z} and @code{-n} respectively, and in
6595 this case a list of units can be explicitly supplied to the binder for
6596 inclusion in the partition (all units needed by these units will also
6597 be included automatically). For full details on the use of these
6598 options, refer to the @cite{GNAT User's Guide} sections on Binding
6604 @strong{32}. The implementation-defined means, if any, of specifying
6605 which compilation units are needed by a given compilation unit. See
6609 The units needed by a given compilation unit are as defined in
6610 the Ada Reference Manual section 10.2(2-6). There are no
6611 implementation-defined pragmas or other implementation-defined
6612 means for specifying needed units.
6617 @strong{33}. The manner of designating the main subprogram of a
6618 partition. See 10.2(7).
6621 The main program is designated by providing the name of the
6622 corresponding @file{ALI} file as the input parameter to the binder.
6627 @strong{34}. The order of elaboration of @code{library_items}. See
6631 The first constraint on ordering is that it meets the requirements of
6632 chapter 10 of the Ada 95 Reference Manual. This still leaves some
6633 implementation dependent choices, which are resolved by first
6634 elaborating bodies as early as possible (i.e.@: in preference to specs
6635 where there is a choice), and second by evaluating the immediate with
6636 clauses of a unit to determine the probably best choice, and
6637 third by elaborating in alphabetical order of unit names
6638 where a choice still remains.
6643 @strong{35}. Parameter passing and function return for the main
6644 subprogram. See 10.2(21).
6647 The main program has no parameters. It may be a procedure, or a function
6648 returning an integer type. In the latter case, the returned integer
6649 value is the return code of the program (overriding any value that
6650 may have been set by a call to @code{Ada.Command_Line.Set_Exit_Status}).
6655 @strong{36}. The mechanisms for building and running partitions. See
6659 GNAT itself supports programs with only a single partition. The GNATDIST
6660 tool provided with the GLADE package (which also includes an implementation
6661 of the PCS) provides a completely flexible method for building and running
6662 programs consisting of multiple partitions. See the separate GLADE manual
6668 @strong{37}. The details of program execution, including program
6669 termination. See 10.2(25).
6672 See separate section on compilation model.
6677 @strong{38}. The semantics of any non-active partitions supported by the
6678 implementation. See 10.2(28).
6681 Passive partitions are supported on targets where shared memory is
6682 provided by the operating system. See the GLADE reference manual for
6688 @strong{39}. The information returned by @code{Exception_Message}. See
6692 Exception message returns the null string unless a specific message has
6693 been passed by the program.
6698 @strong{40}. The result of @code{Exceptions.Exception_Name} for types
6699 declared within an unnamed @code{block_statement}. See 11.4.1(12).
6702 Blocks have implementation defined names of the form @code{B@var{nnn}}
6703 where @var{nnn} is an integer.
6708 @strong{41}. The information returned by
6709 @code{Exception_Information}. See 11.4.1(13).
6712 @code{Exception_Information} returns a string in the following format:
6715 @emph{Exception_Name:} nnnnn
6716 @emph{Message:} mmmmm
6718 @emph{Call stack traceback locations:}
6719 0xhhhh 0xhhhh 0xhhhh ... 0xhhh
6727 @code{nnnn} is the fully qualified name of the exception in all upper
6728 case letters. This line is always present.
6731 @code{mmmm} is the message (this line present only if message is non-null)
6734 @code{ppp} is the Process Id value as a decimal integer (this line is
6735 present only if the Process Id is non-zero). Currently we are
6736 not making use of this field.
6739 The Call stack traceback locations line and the following values
6740 are present only if at least one traceback location was recorded.
6741 The values are given in C style format, with lower case letters
6742 for a-f, and only as many digits present as are necessary.
6746 The line terminator sequence at the end of each line, including
6747 the last line is a single @code{LF} character (@code{16#0A#}).
6752 @strong{42}. Implementation-defined check names. See 11.5(27).
6755 No implementation-defined check names are supported.
6760 @strong{43}. The interpretation of each aspect of representation. See
6764 See separate section on data representations.
6769 @strong{44}. Any restrictions placed upon representation items. See
6773 See separate section on data representations.
6778 @strong{45}. The meaning of @code{Size} for indefinite subtypes. See
6782 Size for an indefinite subtype is the maximum possible size, except that
6783 for the case of a subprogram parameter, the size of the parameter object
6789 @strong{46}. The default external representation for a type tag. See
6793 The default external representation for a type tag is the fully expanded
6794 name of the type in upper case letters.
6799 @strong{47}. What determines whether a compilation unit is the same in
6800 two different partitions. See 13.3(76).
6803 A compilation unit is the same in two different partitions if and only
6804 if it derives from the same source file.
6809 @strong{48}. Implementation-defined components. See 13.5.1(15).
6812 The only implementation defined component is the tag for a tagged type,
6813 which contains a pointer to the dispatching table.
6818 @strong{49}. If @code{Word_Size} = @code{Storage_Unit}, the default bit
6819 ordering. See 13.5.3(5).
6822 @code{Word_Size} (32) is not the same as @code{Storage_Unit} (8) for this
6823 implementation, so no non-default bit ordering is supported. The default
6824 bit ordering corresponds to the natural endianness of the target architecture.
6829 @strong{50}. The contents of the visible part of package @code{System}
6830 and its language-defined children. See 13.7(2).
6833 See the definition of these packages in files @file{system.ads} and
6834 @file{s-stoele.ads}.
6839 @strong{51}. The contents of the visible part of package
6840 @code{System.Machine_Code}, and the meaning of
6841 @code{code_statements}. See 13.8(7).
6844 See the definition and documentation in file @file{s-maccod.ads}.
6849 @strong{52}. The effect of unchecked conversion. See 13.9(11).
6852 Unchecked conversion between types of the same size
6853 and results in an uninterpreted transmission of the bits from one type
6854 to the other. If the types are of unequal sizes, then in the case of
6855 discrete types, a shorter source is first zero or sign extended as
6856 necessary, and a shorter target is simply truncated on the left.
6857 For all non-discrete types, the source is first copied if necessary
6858 to ensure that the alignment requirements of the target are met, then
6859 a pointer is constructed to the source value, and the result is obtained
6860 by dereferencing this pointer after converting it to be a pointer to the
6866 @strong{53}. The manner of choosing a storage pool for an access type
6867 when @code{Storage_Pool} is not specified for the type. See 13.11(17).
6870 There are 3 different standard pools used by the compiler when
6871 @code{Storage_Pool} is not specified depending whether the type is local
6872 to a subprogram or defined at the library level and whether
6873 @code{Storage_Size}is specified or not. See documentation in the runtime
6874 library units @code{System.Pool_Global}, @code{System.Pool_Size} and
6875 @code{System.Pool_Local} in files @file{s-poosiz.ads},
6876 @file{s-pooglo.ads} and @file{s-pooloc.ads} for full details on the
6882 @strong{54}. Whether or not the implementation provides user-accessible
6883 names for the standard pool type(s). See 13.11(17).
6887 See documentation in the sources of the run time mentioned in paragraph
6888 @strong{53} . All these pools are accessible by means of @code{with}'ing
6894 @strong{55}. The meaning of @code{Storage_Size}. See 13.11(18).
6897 @code{Storage_Size} is measured in storage units, and refers to the
6898 total space available for an access type collection, or to the primary
6899 stack space for a task.
6904 @strong{56}. Implementation-defined aspects of storage pools. See
6908 See documentation in the sources of the run time mentioned in paragraph
6909 @strong{53} for details on GNAT-defined aspects of storage pools.
6914 @strong{57}. The set of restrictions allowed in a pragma
6915 @code{Restrictions}. See 13.12(7).
6918 All RM defined Restriction identifiers are implemented. The following
6919 additional restriction identifiers are provided. There are two separate
6920 lists of implementation dependent restriction identifiers. The first
6921 set requires consistency throughout a partition (in other words, if the
6922 restriction identifier is used for any compilation unit in the partition,
6923 then all compilation units in the partition must obey the restriction.
6927 @item Simple_Barriers
6928 @findex Simple_Barriers
6929 This restriction ensures at compile time that barriers in entry declarations
6930 for protected types are restricted to either static boolean expressions or
6931 references to simple boolean variables defined in the private part of the
6932 protected type. No other form of entry barriers is permitted. This is one
6933 of the restrictions of the Ravenscar profile for limited tasking (see also
6934 pragma @code{Profile (Ravenscar)}).
6936 @item Max_Entry_Queue_Length => Expr
6937 @findex Max_Entry_Queue_Length
6938 This restriction is a declaration that any protected entry compiled in
6939 the scope of the restriction has at most the specified number of
6940 tasks waiting on the entry
6941 at any one time, and so no queue is required. This restriction is not
6942 checked at compile time. A program execution is erroneous if an attempt
6943 is made to queue more than the specified number of tasks on such an entry.
6947 This restriction ensures at compile time that there is no implicit or
6948 explicit dependence on the package @code{Ada.Calendar}.
6950 @item No_Direct_Boolean_Operators
6951 @findex No_Direct_Boolean_Operators
6952 This restriction ensures that no logical (and/or/xor) or comparison
6953 operators are used on operands of type Boolean (or any type derived
6954 from Boolean). This is intended for use in safety critical programs
6955 where the certification protocol requires the use of short-circuit
6956 (and then, or else) forms for all composite boolean operations.
6958 @item No_Dynamic_Attachment
6959 @findex No_Dynamic_Attachment
6960 This restriction ensures that there is no call to any of the operations
6961 defined in package Ada.Interrupts.
6963 @item No_Enumeration_Maps
6964 @findex No_Enumeration_Maps
6965 This restriction ensures at compile time that no operations requiring
6966 enumeration maps are used (that is Image and Value attributes applied
6967 to enumeration types).
6969 @item No_Entry_Calls_In_Elaboration_Code
6970 @findex No_Entry_Calls_In_Elaboration_Code
6971 This restriction ensures at compile time that no task or protected entry
6972 calls are made during elaboration code. As a result of the use of this
6973 restriction, the compiler can assume that no code past an accept statement
6974 in a task can be executed at elaboration time.
6976 @item No_Exception_Handlers
6977 @findex No_Exception_Handlers
6978 This restriction ensures at compile time that there are no explicit
6979 exception handlers. It also indicates that no exception propagation will
6980 be provided. In this mode, exceptions may be raised but will result in
6981 an immediate call to the last chance handler, a routine that the user
6982 must define with the following profile:
6984 procedure Last_Chance_Handler
6985 (Source_Location : System.Address; Line : Integer);
6986 pragma Export (C, Last_Chance_Handler,
6987 "__gnat_last_chance_handler");
6989 The parameter is a C null-terminated string representing a message to be
6990 associated with the exception (typically the source location of the raise
6991 statement generated by the compiler). The Line parameter when non-zero
6992 represents the line number in the source program where the raise occurs.
6994 @item No_Exception_Streams
6995 @findex No_Exception_Streams
6996 This restriction ensures at compile time that no stream operations for
6997 types Exception_Id or Exception_Occurrence are used. This also makes it
6998 impossible to pass exceptions to or from a partition with this restriction
6999 in a distributed environment. If this exception is active, then the generated
7000 code is simplified by omitting the otherwise-required global registration
7001 of exceptions when they are declared.
7003 @item No_Implicit_Conditionals
7004 @findex No_Implicit_Conditionals
7005 This restriction ensures that the generated code does not contain any
7006 implicit conditionals, either by modifying the generated code where possible,
7007 or by rejecting any construct that would otherwise generate an implicit
7008 conditional. Note that this check does not include run time constraint
7009 checks, which on some targets may generate implicit conditionals as
7010 well. To control the latter, constraint checks can be suppressed in the
7013 @item No_Implicit_Dynamic_Code
7014 @findex No_Implicit_Dynamic_Code
7015 This restriction prevents the compiler from building ``trampolines''.
7016 This is a structure that is built on the stack and contains dynamic
7017 code to be executed at run time. A trampoline is needed to indirectly
7018 address a nested subprogram (that is a subprogram that is not at the
7019 library level). The restriction prevents the use of any of the
7020 attributes @code{Address}, @code{Access} or @code{Unrestricted_Access}
7021 being applied to a subprogram that is not at the library level.
7023 @item No_Implicit_Loops
7024 @findex No_Implicit_Loops
7025 This restriction ensures that the generated code does not contain any
7026 implicit @code{for} loops, either by modifying
7027 the generated code where possible,
7028 or by rejecting any construct that would otherwise generate an implicit
7031 @item No_Initialize_Scalars
7032 @findex No_Initialize_Scalars
7033 This restriction ensures that no unit in the partition is compiled with
7034 pragma Initialize_Scalars. This allows the generation of more efficient
7035 code, and in particular eliminates dummy null initialization routines that
7036 are otherwise generated for some record and array types.
7038 @item No_Local_Protected_Objects
7039 @findex No_Local_Protected_Objects
7040 This restriction ensures at compile time that protected objects are
7041 only declared at the library level.
7043 @item No_Protected_Type_Allocators
7044 @findex No_Protected_Type_Allocators
7045 This restriction ensures at compile time that there are no allocator
7046 expressions that attempt to allocate protected objects.
7048 @item No_Secondary_Stack
7049 @findex No_Secondary_Stack
7050 This restriction ensures at compile time that the generated code does not
7051 contain any reference to the secondary stack. The secondary stack is used
7052 to implement functions returning unconstrained objects (arrays or records)
7055 @item No_Select_Statements
7056 @findex No_Select_Statements
7057 This restriction ensures at compile time no select statements of any kind
7058 are permitted, that is the keyword @code{select} may not appear.
7059 This is one of the restrictions of the Ravenscar
7060 profile for limited tasking (see also pragma @code{Profile (Ravenscar)}).
7062 @item No_Standard_Storage_Pools
7063 @findex No_Standard_Storage_Pools
7064 This restriction ensures at compile time that no access types
7065 use the standard default storage pool. Any access type declared must
7066 have an explicit Storage_Pool attribute defined specifying a
7067 user-defined storage pool.
7071 This restriction ensures at compile time that there are no implicit or
7072 explicit dependencies on the package @code{Ada.Streams}.
7074 @item No_Task_Attributes_Package
7075 @findex No_Task_Attributes_Package
7076 This restriction ensures at compile time that there are no implicit or
7077 explicit dependencies on the package @code{Ada.Task_Attributes}.
7079 @item No_Task_Termination
7080 @findex No_Task_Termination
7081 This restriction ensures at compile time that no terminate alternatives
7082 appear in any task body.
7086 This restriction prevents the declaration of tasks or task types throughout
7087 the partition. It is similar in effect to the use of @code{Max_Tasks => 0}
7088 except that violations are caught at compile time and cause an error message
7089 to be output either by the compiler or binder.
7091 @item No_Wide_Characters
7092 @findex No_Wide_Characters
7093 This restriction ensures at compile time that no uses of the types
7094 @code{Wide_Character} or @code{Wide_String}
7095 appear, and that no wide character literals
7096 appear in the program (that is literals representing characters not in
7097 type @code{Character}.
7099 @item Static_Priorities
7100 @findex Static_Priorities
7101 This restriction ensures at compile time that all priority expressions
7102 are static, and that there are no dependencies on the package
7103 @code{Ada.Dynamic_Priorities}.
7105 @item Static_Storage_Size
7106 @findex Static_Storage_Size
7107 This restriction ensures at compile time that any expression appearing
7108 in a Storage_Size pragma or attribute definition clause is static.
7113 The second set of implementation dependent restriction identifiers
7114 does not require partition-wide consistency.
7115 The restriction may be enforced for a single
7116 compilation unit without any effect on any of the
7117 other compilation units in the partition.
7121 @item No_Elaboration_Code
7122 @findex No_Elaboration_Code
7123 This restriction ensures at compile time that no elaboration code is
7124 generated. Note that this is not the same condition as is enforced
7125 by pragma @code{Preelaborate}. There are cases in which pragma
7126 @code{Preelaborate} still permits code to be generated (e.g.@: code
7127 to initialize a large array to all zeroes), and there are cases of units
7128 which do not meet the requirements for pragma @code{Preelaborate},
7129 but for which no elaboration code is generated. Generally, it is
7130 the case that preelaborable units will meet the restrictions, with
7131 the exception of large aggregates initialized with an others_clause,
7132 and exception declarations (which generate calls to a run-time
7133 registry procedure). Note that this restriction is enforced on
7134 a unit by unit basis, it need not be obeyed consistently
7135 throughout a partition.
7137 @item No_Entry_Queue
7138 @findex No_Entry_Queue
7139 This restriction is a declaration that any protected entry compiled in
7140 the scope of the restriction has at most one task waiting on the entry
7141 at any one time, and so no queue is required. This restriction is not
7142 checked at compile time. A program execution is erroneous if an attempt
7143 is made to queue a second task on such an entry.
7145 @item No_Implementation_Attributes
7146 @findex No_Implementation_Attributes
7147 This restriction checks at compile time that no GNAT-defined attributes
7148 are present. With this restriction, the only attributes that can be used
7149 are those defined in the Ada 95 Reference Manual.
7151 @item No_Implementation_Pragmas
7152 @findex No_Implementation_Pragmas
7153 This restriction checks at compile time that no GNAT-defined pragmas
7154 are present. With this restriction, the only pragmas that can be used
7155 are those defined in the Ada 95 Reference Manual.
7157 @item No_Implementation_Restrictions
7158 @findex No_Implementation_Restrictions
7159 This restriction checks at compile time that no GNAT-defined restriction
7160 identifiers (other than @code{No_Implementation_Restrictions} itself)
7161 are present. With this restriction, the only other restriction identifiers
7162 that can be used are those defined in the Ada 95 Reference Manual.
7169 @strong{58}. The consequences of violating limitations on
7170 @code{Restrictions} pragmas. See 13.12(9).
7173 Restrictions that can be checked at compile time result in illegalities
7174 if violated. Currently there are no other consequences of violating
7180 @strong{59}. The representation used by the @code{Read} and
7181 @code{Write} attributes of elementary types in terms of stream
7182 elements. See 13.13.2(9).
7185 The representation is the in-memory representation of the base type of
7186 the type, using the number of bits corresponding to the
7187 @code{@var{type}'Size} value, and the natural ordering of the machine.
7192 @strong{60}. The names and characteristics of the numeric subtypes
7193 declared in the visible part of package @code{Standard}. See A.1(3).
7196 See items describing the integer and floating-point types supported.
7201 @strong{61}. The accuracy actually achieved by the elementary
7202 functions. See A.5.1(1).
7205 The elementary functions correspond to the functions available in the C
7206 library. Only fast math mode is implemented.
7211 @strong{62}. The sign of a zero result from some of the operators or
7212 functions in @code{Numerics.Generic_Elementary_Functions}, when
7213 @code{Float_Type'Signed_Zeros} is @code{True}. See A.5.1(46).
7216 The sign of zeroes follows the requirements of the IEEE 754 standard on
7222 @strong{63}. The value of
7223 @code{Numerics.Float_Random.Max_Image_Width}. See A.5.2(27).
7226 Maximum image width is 649, see library file @file{a-numran.ads}.
7231 @strong{64}. The value of
7232 @code{Numerics.Discrete_Random.Max_Image_Width}. See A.5.2(27).
7235 Maximum image width is 80, see library file @file{a-nudira.ads}.
7240 @strong{65}. The algorithms for random number generation. See
7244 The algorithm is documented in the source files @file{a-numran.ads} and
7245 @file{a-numran.adb}.
7250 @strong{66}. The string representation of a random number generator's
7251 state. See A.5.2(38).
7254 See the documentation contained in the file @file{a-numran.adb}.
7259 @strong{67}. The minimum time interval between calls to the
7260 time-dependent Reset procedure that are guaranteed to initiate different
7261 random number sequences. See A.5.2(45).
7264 The minimum period between reset calls to guarantee distinct series of
7265 random numbers is one microsecond.
7270 @strong{68}. The values of the @code{Model_Mantissa},
7271 @code{Model_Emin}, @code{Model_Epsilon}, @code{Model},
7272 @code{Safe_First}, and @code{Safe_Last} attributes, if the Numerics
7273 Annex is not supported. See A.5.3(72).
7276 See the source file @file{ttypef.ads} for the values of all numeric
7282 @strong{69}. Any implementation-defined characteristics of the
7283 input-output packages. See A.7(14).
7286 There are no special implementation defined characteristics for these
7292 @strong{70}. The value of @code{Buffer_Size} in @code{Storage_IO}. See
7296 All type representations are contiguous, and the @code{Buffer_Size} is
7297 the value of @code{@var{type}'Size} rounded up to the next storage unit
7303 @strong{71}. External files for standard input, standard output, and
7304 standard error See A.10(5).
7307 These files are mapped onto the files provided by the C streams
7308 libraries. See source file @file{i-cstrea.ads} for further details.
7313 @strong{72}. The accuracy of the value produced by @code{Put}. See
7317 If more digits are requested in the output than are represented by the
7318 precision of the value, zeroes are output in the corresponding least
7319 significant digit positions.
7324 @strong{73}. The meaning of @code{Argument_Count}, @code{Argument}, and
7325 @code{Command_Name}. See A.15(1).
7328 These are mapped onto the @code{argv} and @code{argc} parameters of the
7329 main program in the natural manner.
7334 @strong{74}. Implementation-defined convention names. See B.1(11).
7337 The following convention names are supported
7345 Synonym for Assembler
7347 Synonym for Assembler
7350 @item C_Pass_By_Copy
7351 Allowed only for record types, like C, but also notes that record
7352 is to be passed by copy rather than reference.
7358 Treated the same as C
7360 Treated the same as C
7364 For support of pragma @code{Import} with convention Intrinsic, see
7365 separate section on Intrinsic Subprograms.
7367 Stdcall (used for Windows implementations only). This convention correspond
7368 to the WINAPI (previously called Pascal convention) C/C++ convention under
7369 Windows. A function with this convention cleans the stack before exit.
7375 Stubbed is a special convention used to indicate that the body of the
7376 subprogram will be entirely ignored. Any call to the subprogram
7377 is converted into a raise of the @code{Program_Error} exception. If a
7378 pragma @code{Import} specifies convention @code{stubbed} then no body need
7379 be present at all. This convention is useful during development for the
7380 inclusion of subprograms whose body has not yet been written.
7384 In addition, all otherwise unrecognized convention names are also
7385 treated as being synonymous with convention C@. In all implementations
7386 except for VMS, use of such other names results in a warning. In VMS
7387 implementations, these names are accepted silently.
7392 @strong{75}. The meaning of link names. See B.1(36).
7395 Link names are the actual names used by the linker.
7400 @strong{76}. The manner of choosing link names when neither the link
7401 name nor the address of an imported or exported entity is specified. See
7405 The default linker name is that which would be assigned by the relevant
7406 external language, interpreting the Ada name as being in all lower case
7412 @strong{77}. The effect of pragma @code{Linker_Options}. See B.1(37).
7415 The string passed to @code{Linker_Options} is presented uninterpreted as
7416 an argument to the link command, unless it contains Ascii.NUL characters.
7417 NUL characters if they appear act as argument separators, so for example
7419 @smallexample @c ada
7420 pragma Linker_Options ("-labc" & ASCII.Nul & "-ldef");
7424 causes two separate arguments @code{-labc} and @code{-ldef} to be passed to the
7425 linker. The order of linker options is preserved for a given unit. The final
7426 list of options passed to the linker is in reverse order of the elaboration
7427 order. For example, linker options fo a body always appear before the options
7428 from the corresponding package spec.
7433 @strong{78}. The contents of the visible part of package
7434 @code{Interfaces} and its language-defined descendants. See B.2(1).
7437 See files with prefix @file{i-} in the distributed library.
7442 @strong{79}. Implementation-defined children of package
7443 @code{Interfaces}. The contents of the visible part of package
7444 @code{Interfaces}. See B.2(11).
7447 See files with prefix @file{i-} in the distributed library.
7452 @strong{80}. The types @code{Floating}, @code{Long_Floating},
7453 @code{Binary}, @code{Long_Binary}, @code{Decimal_ Element}, and
7454 @code{COBOL_Character}; and the initialization of the variables
7455 @code{Ada_To_COBOL} and @code{COBOL_To_Ada}, in
7456 @code{Interfaces.COBOL}. See B.4(50).
7463 (Floating) Long_Float
7468 @item Decimal_Element
7470 @item COBOL_Character
7475 For initialization, see the file @file{i-cobol.ads} in the distributed library.
7480 @strong{81}. Support for access to machine instructions. See C.1(1).
7483 See documentation in file @file{s-maccod.ads} in the distributed library.
7488 @strong{82}. Implementation-defined aspects of access to machine
7489 operations. See C.1(9).
7492 See documentation in file @file{s-maccod.ads} in the distributed library.
7497 @strong{83}. Implementation-defined aspects of interrupts. See C.3(2).
7500 Interrupts are mapped to signals or conditions as appropriate. See
7502 @code{Ada.Interrupt_Names} in source file @file{a-intnam.ads} for details
7503 on the interrupts supported on a particular target.
7508 @strong{84}. Implementation-defined aspects of pre-elaboration. See
7512 GNAT does not permit a partition to be restarted without reloading,
7513 except under control of the debugger.
7518 @strong{85}. The semantics of pragma @code{Discard_Names}. See C.5(7).
7521 Pragma @code{Discard_Names} causes names of enumeration literals to
7522 be suppressed. In the presence of this pragma, the Image attribute
7523 provides the image of the Pos of the literal, and Value accepts
7529 @strong{86}. The result of the @code{Task_Identification.Image}
7530 attribute. See C.7.1(7).
7533 The result of this attribute is an 8-digit hexadecimal string
7534 representing the virtual address of the task control block.
7539 @strong{87}. The value of @code{Current_Task} when in a protected entry
7540 or interrupt handler. See C.7.1(17).
7543 Protected entries or interrupt handlers can be executed by any
7544 convenient thread, so the value of @code{Current_Task} is undefined.
7549 @strong{88}. The effect of calling @code{Current_Task} from an entry
7550 body or interrupt handler. See C.7.1(19).
7553 The effect of calling @code{Current_Task} from an entry body or
7554 interrupt handler is to return the identification of the task currently
7560 @strong{89}. Implementation-defined aspects of
7561 @code{Task_Attributes}. See C.7.2(19).
7564 There are no implementation-defined aspects of @code{Task_Attributes}.
7569 @strong{90}. Values of all @code{Metrics}. See D(2).
7572 The metrics information for GNAT depends on the performance of the
7573 underlying operating system. The sources of the run-time for tasking
7574 implementation, together with the output from @code{-gnatG} can be
7575 used to determine the exact sequence of operating systems calls made
7576 to implement various tasking constructs. Together with appropriate
7577 information on the performance of the underlying operating system,
7578 on the exact target in use, this information can be used to determine
7579 the required metrics.
7584 @strong{91}. The declarations of @code{Any_Priority} and
7585 @code{Priority}. See D.1(11).
7588 See declarations in file @file{system.ads}.
7593 @strong{92}. Implementation-defined execution resources. See D.1(15).
7596 There are no implementation-defined execution resources.
7601 @strong{93}. Whether, on a multiprocessor, a task that is waiting for
7602 access to a protected object keeps its processor busy. See D.2.1(3).
7605 On a multi-processor, a task that is waiting for access to a protected
7606 object does not keep its processor busy.
7611 @strong{94}. The affect of implementation defined execution resources
7612 on task dispatching. See D.2.1(9).
7617 Tasks map to IRIX threads, and the dispatching policy is as defined by
7618 the IRIX implementation of threads.
7620 Tasks map to threads in the threads package used by GNAT@. Where possible
7621 and appropriate, these threads correspond to native threads of the
7622 underlying operating system.
7627 @strong{95}. Implementation-defined @code{policy_identifiers} allowed
7628 in a pragma @code{Task_Dispatching_Policy}. See D.2.2(3).
7631 There are no implementation-defined policy-identifiers allowed in this
7637 @strong{96}. Implementation-defined aspects of priority inversion. See
7641 Execution of a task cannot be preempted by the implementation processing
7642 of delay expirations for lower priority tasks.
7647 @strong{97}. Implementation defined task dispatching. See D.2.2(18).
7652 Tasks map to IRIX threads, and the dispatching policy is as defied by
7653 the IRIX implementation of threads.
7655 The policy is the same as that of the underlying threads implementation.
7660 @strong{98}. Implementation-defined @code{policy_identifiers} allowed
7661 in a pragma @code{Locking_Policy}. See D.3(4).
7664 The only implementation defined policy permitted in GNAT is
7665 @code{Inheritance_Locking}. On targets that support this policy, locking
7666 is implemented by inheritance, i.e.@: the task owning the lock operates
7667 at a priority equal to the highest priority of any task currently
7668 requesting the lock.
7673 @strong{99}. Default ceiling priorities. See D.3(10).
7676 The ceiling priority of protected objects of the type
7677 @code{System.Interrupt_Priority'Last} as described in the Ada 95
7678 Reference Manual D.3(10),
7683 @strong{100}. The ceiling of any protected object used internally by
7684 the implementation. See D.3(16).
7687 The ceiling priority of internal protected objects is
7688 @code{System.Priority'Last}.
7693 @strong{101}. Implementation-defined queuing policies. See D.4(1).
7696 There are no implementation-defined queueing policies.
7701 @strong{102}. On a multiprocessor, any conditions that cause the
7702 completion of an aborted construct to be delayed later than what is
7703 specified for a single processor. See D.6(3).
7706 The semantics for abort on a multi-processor is the same as on a single
7707 processor, there are no further delays.
7712 @strong{103}. Any operations that implicitly require heap storage
7713 allocation. See D.7(8).
7716 The only operation that implicitly requires heap storage allocation is
7722 @strong{104}. Implementation-defined aspects of pragma
7723 @code{Restrictions}. See D.7(20).
7726 There are no such implementation-defined aspects.
7731 @strong{105}. Implementation-defined aspects of package
7732 @code{Real_Time}. See D.8(17).
7735 There are no implementation defined aspects of package @code{Real_Time}.
7740 @strong{106}. Implementation-defined aspects of
7741 @code{delay_statements}. See D.9(8).
7744 Any difference greater than one microsecond will cause the task to be
7745 delayed (see D.9(7)).
7750 @strong{107}. The upper bound on the duration of interrupt blocking
7751 caused by the implementation. See D.12(5).
7754 The upper bound is determined by the underlying operating system. In
7755 no cases is it more than 10 milliseconds.
7760 @strong{108}. The means for creating and executing distributed
7764 The GLADE package provides a utility GNATDIST for creating and executing
7765 distributed programs. See the GLADE reference manual for further details.
7770 @strong{109}. Any events that can result in a partition becoming
7771 inaccessible. See E.1(7).
7774 See the GLADE reference manual for full details on such events.
7779 @strong{110}. The scheduling policies, treatment of priorities, and
7780 management of shared resources between partitions in certain cases. See
7784 See the GLADE reference manual for full details on these aspects of
7785 multi-partition execution.
7790 @strong{111}. Events that cause the version of a compilation unit to
7794 Editing the source file of a compilation unit, or the source files of
7795 any units on which it is dependent in a significant way cause the version
7796 to change. No other actions cause the version number to change. All changes
7797 are significant except those which affect only layout, capitalization or
7803 @strong{112}. Whether the execution of the remote subprogram is
7804 immediately aborted as a result of cancellation. See E.4(13).
7807 See the GLADE reference manual for details on the effect of abort in
7808 a distributed application.
7813 @strong{113}. Implementation-defined aspects of the PCS@. See E.5(25).
7816 See the GLADE reference manual for a full description of all implementation
7817 defined aspects of the PCS@.
7822 @strong{114}. Implementation-defined interfaces in the PCS@. See
7826 See the GLADE reference manual for a full description of all
7827 implementation defined interfaces.
7832 @strong{115}. The values of named numbers in the package
7833 @code{Decimal}. See F.2(7).
7845 @item Max_Decimal_Digits
7852 @strong{116}. The value of @code{Max_Picture_Length} in the package
7853 @code{Text_IO.Editing}. See F.3.3(16).
7861 @strong{117}. The value of @code{Max_Picture_Length} in the package
7862 @code{Wide_Text_IO.Editing}. See F.3.4(5).
7870 @strong{118}. The accuracy actually achieved by the complex elementary
7871 functions and by other complex arithmetic operations. See G.1(1).
7874 Standard library functions are used for the complex arithmetic
7875 operations. Only fast math mode is currently supported.
7880 @strong{119}. The sign of a zero result (or a component thereof) from
7881 any operator or function in @code{Numerics.Generic_Complex_Types}, when
7882 @code{Real'Signed_Zeros} is True. See G.1.1(53).
7885 The signs of zero values are as recommended by the relevant
7886 implementation advice.
7891 @strong{120}. The sign of a zero result (or a component thereof) from
7892 any operator or function in
7893 @code{Numerics.Generic_Complex_Elementary_Functions}, when
7894 @code{Real'Signed_Zeros} is @code{True}. See G.1.2(45).
7897 The signs of zero values are as recommended by the relevant
7898 implementation advice.
7903 @strong{121}. Whether the strict mode or the relaxed mode is the
7904 default. See G.2(2).
7907 The strict mode is the default. There is no separate relaxed mode. GNAT
7908 provides a highly efficient implementation of strict mode.
7913 @strong{122}. The result interval in certain cases of fixed-to-float
7914 conversion. See G.2.1(10).
7917 For cases where the result interval is implementation dependent, the
7918 accuracy is that provided by performing all operations in 64-bit IEEE
7919 floating-point format.
7924 @strong{123}. The result of a floating point arithmetic operation in
7925 overflow situations, when the @code{Machine_Overflows} attribute of the
7926 result type is @code{False}. See G.2.1(13).
7929 Infinite and Nan values are produced as dictated by the IEEE
7930 floating-point standard.
7935 @strong{124}. The result interval for division (or exponentiation by a
7936 negative exponent), when the floating point hardware implements division
7937 as multiplication by a reciprocal. See G.2.1(16).
7940 Not relevant, division is IEEE exact.
7945 @strong{125}. The definition of close result set, which determines the
7946 accuracy of certain fixed point multiplications and divisions. See
7950 Operations in the close result set are performed using IEEE long format
7951 floating-point arithmetic. The input operands are converted to
7952 floating-point, the operation is done in floating-point, and the result
7953 is converted to the target type.
7958 @strong{126}. Conditions on a @code{universal_real} operand of a fixed
7959 point multiplication or division for which the result shall be in the
7960 perfect result set. See G.2.3(22).
7963 The result is only defined to be in the perfect result set if the result
7964 can be computed by a single scaling operation involving a scale factor
7965 representable in 64-bits.
7970 @strong{127}. The result of a fixed point arithmetic operation in
7971 overflow situations, when the @code{Machine_Overflows} attribute of the
7972 result type is @code{False}. See G.2.3(27).
7975 Not relevant, @code{Machine_Overflows} is @code{True} for fixed-point
7981 @strong{128}. The result of an elementary function reference in
7982 overflow situations, when the @code{Machine_Overflows} attribute of the
7983 result type is @code{False}. See G.2.4(4).
7986 IEEE infinite and Nan values are produced as appropriate.
7991 @strong{129}. The value of the angle threshold, within which certain
7992 elementary functions, complex arithmetic operations, and complex
7993 elementary functions yield results conforming to a maximum relative
7994 error bound. See G.2.4(10).
7997 Information on this subject is not yet available.
8002 @strong{130}. The accuracy of certain elementary functions for
8003 parameters beyond the angle threshold. See G.2.4(10).
8006 Information on this subject is not yet available.
8011 @strong{131}. The result of a complex arithmetic operation or complex
8012 elementary function reference in overflow situations, when the
8013 @code{Machine_Overflows} attribute of the corresponding real type is
8014 @code{False}. See G.2.6(5).
8017 IEEE infinite and Nan values are produced as appropriate.
8022 @strong{132}. The accuracy of certain complex arithmetic operations and
8023 certain complex elementary functions for parameters (or components
8024 thereof) beyond the angle threshold. See G.2.6(8).
8027 Information on those subjects is not yet available.
8032 @strong{133}. Information regarding bounded errors and erroneous
8033 execution. See H.2(1).
8036 Information on this subject is not yet available.
8041 @strong{134}. Implementation-defined aspects of pragma
8042 @code{Inspection_Point}. See H.3.2(8).
8045 Pragma @code{Inspection_Point} ensures that the variable is live and can
8046 be examined by the debugger at the inspection point.
8051 @strong{135}. Implementation-defined aspects of pragma
8052 @code{Restrictions}. See H.4(25).
8055 There are no implementation-defined aspects of pragma @code{Restrictions}. The
8056 use of pragma @code{Restrictions [No_Exceptions]} has no effect on the
8057 generated code. Checks must suppressed by use of pragma @code{Suppress}.
8062 @strong{136}. Any restrictions on pragma @code{Restrictions}. See
8066 There are no restrictions on pragma @code{Restrictions}.
8068 @node Intrinsic Subprograms
8069 @chapter Intrinsic Subprograms
8070 @cindex Intrinsic Subprograms
8073 * Intrinsic Operators::
8074 * Enclosing_Entity::
8075 * Exception_Information::
8076 * Exception_Message::
8084 * Shift_Right_Arithmetic::
8089 GNAT allows a user application program to write the declaration:
8091 @smallexample @c ada
8092 pragma Import (Intrinsic, name);
8096 providing that the name corresponds to one of the implemented intrinsic
8097 subprograms in GNAT, and that the parameter profile of the referenced
8098 subprogram meets the requirements. This chapter describes the set of
8099 implemented intrinsic subprograms, and the requirements on parameter profiles.
8100 Note that no body is supplied; as with other uses of pragma Import, the
8101 body is supplied elsewhere (in this case by the compiler itself). Note
8102 that any use of this feature is potentially non-portable, since the
8103 Ada standard does not require Ada compilers to implement this feature.
8105 @node Intrinsic Operators
8106 @section Intrinsic Operators
8107 @cindex Intrinsic operator
8110 All the predefined numeric operators in package Standard
8111 in @code{pragma Import (Intrinsic,..)}
8112 declarations. In the binary operator case, the operands must have the same
8113 size. The operand or operands must also be appropriate for
8114 the operator. For example, for addition, the operands must
8115 both be floating-point or both be fixed-point, and the
8116 right operand for @code{"**"} must have a root type of
8117 @code{Standard.Integer'Base}.
8118 You can use an intrinsic operator declaration as in the following example:
8120 @smallexample @c ada
8121 type Int1 is new Integer;
8122 type Int2 is new Integer;
8124 function "+" (X1 : Int1; X2 : Int2) return Int1;
8125 function "+" (X1 : Int1; X2 : Int2) return Int2;
8126 pragma Import (Intrinsic, "+");
8130 This declaration would permit ``mixed mode'' arithmetic on items
8131 of the differing types @code{Int1} and @code{Int2}.
8132 It is also possible to specify such operators for private types, if the
8133 full views are appropriate arithmetic types.
8135 @node Enclosing_Entity
8136 @section Enclosing_Entity
8137 @cindex Enclosing_Entity
8139 This intrinsic subprogram is used in the implementation of the
8140 library routine @code{GNAT.Source_Info}. The only useful use of the
8141 intrinsic import in this case is the one in this unit, so an
8142 application program should simply call the function
8143 @code{GNAT.Source_Info.Enclosing_Entity} to obtain the name of
8144 the current subprogram, package, task, entry, or protected subprogram.
8146 @node Exception_Information
8147 @section Exception_Information
8148 @cindex Exception_Information'
8150 This intrinsic subprogram is used in the implementation of the
8151 library routine @code{GNAT.Current_Exception}. The only useful
8152 use of the intrinsic import in this case is the one in this unit,
8153 so an application program should simply call the function
8154 @code{GNAT.Current_Exception.Exception_Information} to obtain
8155 the exception information associated with the current exception.
8157 @node Exception_Message
8158 @section Exception_Message
8159 @cindex Exception_Message
8161 This intrinsic subprogram is used in the implementation of the
8162 library routine @code{GNAT.Current_Exception}. The only useful
8163 use of the intrinsic import in this case is the one in this unit,
8164 so an application program should simply call the function
8165 @code{GNAT.Current_Exception.Exception_Message} to obtain
8166 the message associated with the current exception.
8168 @node Exception_Name
8169 @section Exception_Name
8170 @cindex Exception_Name
8172 This intrinsic subprogram is used in the implementation of the
8173 library routine @code{GNAT.Current_Exception}. The only useful
8174 use of the intrinsic import in this case is the one in this unit,
8175 so an application program should simply call the function
8176 @code{GNAT.Current_Exception.Exception_Name} to obtain
8177 the name of the current exception.
8183 This intrinsic subprogram is used in the implementation of the
8184 library routine @code{GNAT.Source_Info}. The only useful use of the
8185 intrinsic import in this case is the one in this unit, so an
8186 application program should simply call the function
8187 @code{GNAT.Source_Info.File} to obtain the name of the current
8194 This intrinsic subprogram is used in the implementation of the
8195 library routine @code{GNAT.Source_Info}. The only useful use of the
8196 intrinsic import in this case is the one in this unit, so an
8197 application program should simply call the function
8198 @code{GNAT.Source_Info.Line} to obtain the number of the current
8202 @section Rotate_Left
8205 In standard Ada 95, the @code{Rotate_Left} function is available only
8206 for the predefined modular types in package @code{Interfaces}. However, in
8207 GNAT it is possible to define a Rotate_Left function for a user
8208 defined modular type or any signed integer type as in this example:
8210 @smallexample @c ada
8212 (Value : My_Modular_Type;
8214 return My_Modular_Type;
8218 The requirements are that the profile be exactly as in the example
8219 above. The only modifications allowed are in the formal parameter
8220 names, and in the type of @code{Value} and the return type, which
8221 must be the same, and must be either a signed integer type, or
8222 a modular integer type with a binary modulus, and the size must
8223 be 8. 16, 32 or 64 bits.
8226 @section Rotate_Right
8227 @cindex Rotate_Right
8229 A @code{Rotate_Right} function can be defined for any user defined
8230 binary modular integer type, or signed integer type, as described
8231 above for @code{Rotate_Left}.
8237 A @code{Shift_Left} function can be defined for any user defined
8238 binary modular integer type, or signed integer type, as described
8239 above for @code{Rotate_Left}.
8242 @section Shift_Right
8245 A @code{Shift_Right} function can be defined for any user defined
8246 binary modular integer type, or signed integer type, as described
8247 above for @code{Rotate_Left}.
8249 @node Shift_Right_Arithmetic
8250 @section Shift_Right_Arithmetic
8251 @cindex Shift_Right_Arithmetic
8253 A @code{Shift_Right_Arithmetic} function can be defined for any user
8254 defined binary modular integer type, or signed integer type, as described
8255 above for @code{Rotate_Left}.
8257 @node Source_Location
8258 @section Source_Location
8259 @cindex Source_Location
8261 This intrinsic subprogram is used in the implementation of the
8262 library routine @code{GNAT.Source_Info}. The only useful use of the
8263 intrinsic import in this case is the one in this unit, so an
8264 application program should simply call the function
8265 @code{GNAT.Source_Info.Source_Location} to obtain the current
8266 source file location.
8268 @node Representation Clauses and Pragmas
8269 @chapter Representation Clauses and Pragmas
8270 @cindex Representation Clauses
8273 * Alignment Clauses::
8275 * Storage_Size Clauses::
8276 * Size of Variant Record Objects::
8277 * Biased Representation ::
8278 * Value_Size and Object_Size Clauses::
8279 * Component_Size Clauses::
8280 * Bit_Order Clauses::
8281 * Effect of Bit_Order on Byte Ordering::
8282 * Pragma Pack for Arrays::
8283 * Pragma Pack for Records::
8284 * Record Representation Clauses::
8285 * Enumeration Clauses::
8287 * Effect of Convention on Representation::
8288 * Determining the Representations chosen by GNAT::
8292 @cindex Representation Clause
8293 @cindex Representation Pragma
8294 @cindex Pragma, representation
8295 This section describes the representation clauses accepted by GNAT, and
8296 their effect on the representation of corresponding data objects.
8298 GNAT fully implements Annex C (Systems Programming). This means that all
8299 the implementation advice sections in chapter 13 are fully implemented.
8300 However, these sections only require a minimal level of support for
8301 representation clauses. GNAT provides much more extensive capabilities,
8302 and this section describes the additional capabilities provided.
8304 @node Alignment Clauses
8305 @section Alignment Clauses
8306 @cindex Alignment Clause
8309 GNAT requires that all alignment clauses specify a power of 2, and all
8310 default alignments are always a power of 2. The default alignment
8311 values are as follows:
8314 @item @emph{Primitive Types}.
8315 For primitive types, the alignment is the minimum of the actual size of
8316 objects of the type divided by @code{Storage_Unit},
8317 and the maximum alignment supported by the target.
8318 (This maximum alignment is given by the GNAT-specific attribute
8319 @code{Standard'Maximum_Alignment}; see @ref{Maximum_Alignment}.)
8320 @cindex @code{Maximum_Alignment} attribute
8321 For example, for type @code{Long_Float}, the object size is 8 bytes, and the
8322 default alignment will be 8 on any target that supports alignments
8323 this large, but on some targets, the maximum alignment may be smaller
8324 than 8, in which case objects of type @code{Long_Float} will be maximally
8327 @item @emph{Arrays}.
8328 For arrays, the alignment is equal to the alignment of the component type
8329 for the normal case where no packing or component size is given. If the
8330 array is packed, and the packing is effective (see separate section on
8331 packed arrays), then the alignment will be one for long packed arrays,
8332 or arrays whose length is not known at compile time. For short packed
8333 arrays, which are handled internally as modular types, the alignment
8334 will be as described for primitive types, e.g.@: a packed array of length
8335 31 bits will have an object size of four bytes, and an alignment of 4.
8337 @item @emph{Records}.
8338 For the normal non-packed case, the alignment of a record is equal to
8339 the maximum alignment of any of its components. For tagged records, this
8340 includes the implicit access type used for the tag. If a pragma @code{Pack} is
8341 used and all fields are packable (see separate section on pragma @code{Pack}),
8342 then the resulting alignment is 1.
8344 A special case is when:
8347 the size of the record is given explicitly, or a
8348 full record representation clause is given, and
8350 the size of the record is 2, 4, or 8 bytes.
8353 In this case, an alignment is chosen to match the
8354 size of the record. For example, if we have:
8356 @smallexample @c ada
8357 type Small is record
8360 for Small'Size use 16;
8364 then the default alignment of the record type @code{Small} is 2, not 1. This
8365 leads to more efficient code when the record is treated as a unit, and also
8366 allows the type to specified as @code{Atomic} on architectures requiring
8372 An alignment clause may
8373 always specify a larger alignment than the default value, up to some
8374 maximum value dependent on the target (obtainable by using the
8375 attribute reference @code{Standard'Maximum_Alignment}).
8377 it is permissible to specify a smaller alignment than the default value
8378 is for a record with a record representation clause.
8379 In this case, packable fields for which a component clause is
8380 given still result in a default alignment corresponding to the original
8381 type, but this may be overridden, since these components in fact only
8382 require an alignment of one byte. For example, given
8384 @smallexample @c ada
8390 A at 0 range 0 .. 31;
8393 for V'alignment use 1;
8397 @cindex Alignment, default
8398 The default alignment for the type @code{V} is 4, as a result of the
8399 Integer field in the record, but since this field is placed with a
8400 component clause, it is permissible, as shown, to override the default
8401 alignment of the record with a smaller value.
8404 @section Size Clauses
8408 The default size for a type @code{T} is obtainable through the
8409 language-defined attribute @code{T'Size} and also through the
8410 equivalent GNAT-defined attribute @code{T'Value_Size}.
8411 For objects of type @code{T}, GNAT will generally increase the type size
8412 so that the object size (obtainable through the GNAT-defined attribute
8413 @code{T'Object_Size})
8414 is a multiple of @code{T'Alignment * Storage_Unit}.
8417 @smallexample @c ada
8418 type Smallint is range 1 .. 6;
8427 In this example, @code{Smallint'Size} = @code{Smallint'Value_Size} = 3,
8428 as specified by the RM rules,
8429 but objects of this type will have a size of 8
8430 (@code{Smallint'Object_Size} = 8),
8431 since objects by default occupy an integral number
8432 of storage units. On some targets, notably older
8433 versions of the Digital Alpha, the size of stand
8434 alone objects of this type may be 32, reflecting
8435 the inability of the hardware to do byte load/stores.
8437 Similarly, the size of type @code{Rec} is 40 bits
8438 (@code{Rec'Size} = @code{Rec'Value_Size} = 40), but
8439 the alignment is 4, so objects of this type will have
8440 their size increased to 64 bits so that it is a multiple
8441 of the alignment (in bits). This decision is
8442 in accordance with the specific Implementation Advice in RM 13.3(43):
8445 A @code{Size} clause should be supported for an object if the specified
8446 @code{Size} is at least as large as its subtype's @code{Size}, and corresponds
8447 to a size in storage elements that is a multiple of the object's
8448 @code{Alignment} (if the @code{Alignment} is nonzero).
8452 An explicit size clause may be used to override the default size by
8453 increasing it. For example, if we have:
8455 @smallexample @c ada
8456 type My_Boolean is new Boolean;
8457 for My_Boolean'Size use 32;
8461 then values of this type will always be 32 bits long. In the case of
8462 discrete types, the size can be increased up to 64 bits, with the effect
8463 that the entire specified field is used to hold the value, sign- or
8464 zero-extended as appropriate. If more than 64 bits is specified, then
8465 padding space is allocated after the value, and a warning is issued that
8466 there are unused bits.
8468 Similarly the size of records and arrays may be increased, and the effect
8469 is to add padding bits after the value. This also causes a warning message
8472 The largest Size value permitted in GNAT is 2**31@minus{}1. Since this is a
8473 Size in bits, this corresponds to an object of size 256 megabytes (minus
8474 one). This limitation is true on all targets. The reason for this
8475 limitation is that it improves the quality of the code in many cases
8476 if it is known that a Size value can be accommodated in an object of
8479 @node Storage_Size Clauses
8480 @section Storage_Size Clauses
8481 @cindex Storage_Size Clause
8484 For tasks, the @code{Storage_Size} clause specifies the amount of space
8485 to be allocated for the task stack. This cannot be extended, and if the
8486 stack is exhausted, then @code{Storage_Error} will be raised (if stack
8487 checking is enabled). Use a @code{Storage_Size} attribute definition clause,
8488 or a @code{Storage_Size} pragma in the task definition to set the
8489 appropriate required size. A useful technique is to include in every
8490 task definition a pragma of the form:
8492 @smallexample @c ada
8493 pragma Storage_Size (Default_Stack_Size);
8497 Then @code{Default_Stack_Size} can be defined in a global package, and
8498 modified as required. Any tasks requiring stack sizes different from the
8499 default can have an appropriate alternative reference in the pragma.
8501 For access types, the @code{Storage_Size} clause specifies the maximum
8502 space available for allocation of objects of the type. If this space is
8503 exceeded then @code{Storage_Error} will be raised by an allocation attempt.
8504 In the case where the access type is declared local to a subprogram, the
8505 use of a @code{Storage_Size} clause triggers automatic use of a special
8506 predefined storage pool (@code{System.Pool_Size}) that ensures that all
8507 space for the pool is automatically reclaimed on exit from the scope in
8508 which the type is declared.
8510 A special case recognized by the compiler is the specification of a
8511 @code{Storage_Size} of zero for an access type. This means that no
8512 items can be allocated from the pool, and this is recognized at compile
8513 time, and all the overhead normally associated with maintaining a fixed
8514 size storage pool is eliminated. Consider the following example:
8516 @smallexample @c ada
8518 type R is array (Natural) of Character;
8519 type P is access all R;
8520 for P'Storage_Size use 0;
8521 -- Above access type intended only for interfacing purposes
8525 procedure g (m : P);
8526 pragma Import (C, g);
8537 As indicated in this example, these dummy storage pools are often useful in
8538 connection with interfacing where no object will ever be allocated. If you
8539 compile the above example, you get the warning:
8542 p.adb:16:09: warning: allocation from empty storage pool
8543 p.adb:16:09: warning: Storage_Error will be raised at run time
8547 Of course in practice, there will not be any explicit allocators in the
8548 case of such an access declaration.
8550 @node Size of Variant Record Objects
8551 @section Size of Variant Record Objects
8552 @cindex Size, variant record objects
8553 @cindex Variant record objects, size
8556 In the case of variant record objects, there is a question whether Size gives
8557 information about a particular variant, or the maximum size required
8558 for any variant. Consider the following program
8560 @smallexample @c ada
8561 with Text_IO; use Text_IO;
8563 type R1 (A : Boolean := False) is record
8565 when True => X : Character;
8574 Put_Line (Integer'Image (V1'Size));
8575 Put_Line (Integer'Image (V2'Size));
8580 Here we are dealing with a variant record, where the True variant
8581 requires 16 bits, and the False variant requires 8 bits.
8582 In the above example, both V1 and V2 contain the False variant,
8583 which is only 8 bits long. However, the result of running the
8592 The reason for the difference here is that the discriminant value of
8593 V1 is fixed, and will always be False. It is not possible to assign
8594 a True variant value to V1, therefore 8 bits is sufficient. On the
8595 other hand, in the case of V2, the initial discriminant value is
8596 False (from the default), but it is possible to assign a True
8597 variant value to V2, therefore 16 bits must be allocated for V2
8598 in the general case, even fewer bits may be needed at any particular
8599 point during the program execution.
8601 As can be seen from the output of this program, the @code{'Size}
8602 attribute applied to such an object in GNAT gives the actual allocated
8603 size of the variable, which is the largest size of any of the variants.
8604 The Ada Reference Manual is not completely clear on what choice should
8605 be made here, but the GNAT behavior seems most consistent with the
8606 language in the RM@.
8608 In some cases, it may be desirable to obtain the size of the current
8609 variant, rather than the size of the largest variant. This can be
8610 achieved in GNAT by making use of the fact that in the case of a
8611 subprogram parameter, GNAT does indeed return the size of the current
8612 variant (because a subprogram has no way of knowing how much space
8613 is actually allocated for the actual).
8615 Consider the following modified version of the above program:
8617 @smallexample @c ada
8618 with Text_IO; use Text_IO;
8620 type R1 (A : Boolean := False) is record
8622 when True => X : Character;
8629 function Size (V : R1) return Integer is
8635 Put_Line (Integer'Image (V2'Size));
8636 Put_Line (Integer'IMage (Size (V2)));
8638 Put_Line (Integer'Image (V2'Size));
8639 Put_Line (Integer'IMage (Size (V2)));
8644 The output from this program is
8654 Here we see that while the @code{'Size} attribute always returns
8655 the maximum size, regardless of the current variant value, the
8656 @code{Size} function does indeed return the size of the current
8659 @node Biased Representation
8660 @section Biased Representation
8661 @cindex Size for biased representation
8662 @cindex Biased representation
8665 In the case of scalars with a range starting at other than zero, it is
8666 possible in some cases to specify a size smaller than the default minimum
8667 value, and in such cases, GNAT uses an unsigned biased representation,
8668 in which zero is used to represent the lower bound, and successive values
8669 represent successive values of the type.
8671 For example, suppose we have the declaration:
8673 @smallexample @c ada
8674 type Small is range -7 .. -4;
8675 for Small'Size use 2;
8679 Although the default size of type @code{Small} is 4, the @code{Size}
8680 clause is accepted by GNAT and results in the following representation
8684 -7 is represented as 2#00#
8685 -6 is represented as 2#01#
8686 -5 is represented as 2#10#
8687 -4 is represented as 2#11#
8691 Biased representation is only used if the specified @code{Size} clause
8692 cannot be accepted in any other manner. These reduced sizes that force
8693 biased representation can be used for all discrete types except for
8694 enumeration types for which a representation clause is given.
8696 @node Value_Size and Object_Size Clauses
8697 @section Value_Size and Object_Size Clauses
8700 @cindex Size, of objects
8703 In Ada 95, @code{T'Size} for a type @code{T} is the minimum number of bits
8704 required to hold values of type @code{T}. Although this interpretation was
8705 allowed in Ada 83, it was not required, and this requirement in practice
8706 can cause some significant difficulties. For example, in most Ada 83
8707 compilers, @code{Natural'Size} was 32. However, in Ada 95,
8708 @code{Natural'Size} is
8709 typically 31. This means that code may change in behavior when moving
8710 from Ada 83 to Ada 95. For example, consider:
8712 @smallexample @c ada
8719 at 0 range 0 .. Natural'Size - 1;
8720 at 0 range Natural'Size .. 2 * Natural'Size - 1;
8725 In the above code, since the typical size of @code{Natural} objects
8726 is 32 bits and @code{Natural'Size} is 31, the above code can cause
8727 unexpected inefficient packing in Ada 95, and in general there are
8728 cases where the fact that the object size can exceed the
8729 size of the type causes surprises.
8731 To help get around this problem GNAT provides two implementation
8732 defined attributes, @code{Value_Size} and @code{Object_Size}. When
8733 applied to a type, these attributes yield the size of the type
8734 (corresponding to the RM defined size attribute), and the size of
8735 objects of the type respectively.
8737 The @code{Object_Size} is used for determining the default size of
8738 objects and components. This size value can be referred to using the
8739 @code{Object_Size} attribute. The phrase ``is used'' here means that it is
8740 the basis of the determination of the size. The backend is free to
8741 pad this up if necessary for efficiency, e.g.@: an 8-bit stand-alone
8742 character might be stored in 32 bits on a machine with no efficient
8743 byte access instructions such as the Alpha.
8745 The default rules for the value of @code{Object_Size} for
8746 discrete types are as follows:
8750 The @code{Object_Size} for base subtypes reflect the natural hardware
8751 size in bits (run the compiler with @option{-gnatS} to find those values
8752 for numeric types). Enumeration types and fixed-point base subtypes have
8753 8, 16, 32 or 64 bits for this size, depending on the range of values
8757 The @code{Object_Size} of a subtype is the same as the
8758 @code{Object_Size} of
8759 the type from which it is obtained.
8762 The @code{Object_Size} of a derived base type is copied from the parent
8763 base type, and the @code{Object_Size} of a derived first subtype is copied
8764 from the parent first subtype.
8768 The @code{Value_Size} attribute
8769 is the (minimum) number of bits required to store a value
8771 This value is used to determine how tightly to pack
8772 records or arrays with components of this type, and also affects
8773 the semantics of unchecked conversion (unchecked conversions where
8774 the @code{Value_Size} values differ generate a warning, and are potentially
8777 The default rules for the value of @code{Value_Size} are as follows:
8781 The @code{Value_Size} for a base subtype is the minimum number of bits
8782 required to store all values of the type (including the sign bit
8783 only if negative values are possible).
8786 If a subtype statically matches the first subtype of a given type, then it has
8787 by default the same @code{Value_Size} as the first subtype. This is a
8788 consequence of RM 13.1(14) (``if two subtypes statically match,
8789 then their subtype-specific aspects are the same''.)
8792 All other subtypes have a @code{Value_Size} corresponding to the minimum
8793 number of bits required to store all values of the subtype. For
8794 dynamic bounds, it is assumed that the value can range down or up
8795 to the corresponding bound of the ancestor
8799 The RM defined attribute @code{Size} corresponds to the
8800 @code{Value_Size} attribute.
8802 The @code{Size} attribute may be defined for a first-named subtype. This sets
8803 the @code{Value_Size} of
8804 the first-named subtype to the given value, and the
8805 @code{Object_Size} of this first-named subtype to the given value padded up
8806 to an appropriate boundary. It is a consequence of the default rules
8807 above that this @code{Object_Size} will apply to all further subtypes. On the
8808 other hand, @code{Value_Size} is affected only for the first subtype, any
8809 dynamic subtypes obtained from it directly, and any statically matching
8810 subtypes. The @code{Value_Size} of any other static subtypes is not affected.
8812 @code{Value_Size} and
8813 @code{Object_Size} may be explicitly set for any subtype using
8814 an attribute definition clause. Note that the use of these attributes
8815 can cause the RM 13.1(14) rule to be violated. If two access types
8816 reference aliased objects whose subtypes have differing @code{Object_Size}
8817 values as a result of explicit attribute definition clauses, then it
8818 is erroneous to convert from one access subtype to the other.
8820 At the implementation level, Esize stores the Object_Size and the
8821 RM_Size field stores the @code{Value_Size} (and hence the value of the
8822 @code{Size} attribute,
8823 which, as noted above, is equivalent to @code{Value_Size}).
8825 To get a feel for the difference, consider the following examples (note
8826 that in each case the base is @code{Short_Short_Integer} with a size of 8):
8829 Object_Size Value_Size
8831 type x1 is range 0 .. 5; 8 3
8833 type x2 is range 0 .. 5;
8834 for x2'size use 12; 16 12
8836 subtype x3 is x2 range 0 .. 3; 16 2
8838 subtype x4 is x2'base range 0 .. 10; 8 4
8840 subtype x5 is x2 range 0 .. dynamic; 16 3*
8842 subtype x6 is x2'base range 0 .. dynamic; 8 3*
8847 Note: the entries marked ``3*'' are not actually specified by the Ada 95 RM,
8848 but it seems in the spirit of the RM rules to allocate the minimum number
8849 of bits (here 3, given the range for @code{x2})
8850 known to be large enough to hold the given range of values.
8852 So far, so good, but GNAT has to obey the RM rules, so the question is
8853 under what conditions must the RM @code{Size} be used.
8854 The following is a list
8855 of the occasions on which the RM @code{Size} must be used:
8859 Component size for packed arrays or records
8862 Value of the attribute @code{Size} for a type
8865 Warning about sizes not matching for unchecked conversion
8869 For record types, the @code{Object_Size} is always a multiple of the
8870 alignment of the type (this is true for all types). In some cases the
8871 @code{Value_Size} can be smaller. Consider:
8881 On a typical 32-bit architecture, the X component will be four bytes, and
8882 require four-byte alignment, and the Y component will be one byte. In this
8883 case @code{R'Value_Size} will be 40 (bits) since this is the minimum size
8884 required to store a value of this type, and for example, it is permissible
8885 to have a component of type R in an outer record whose component size is
8886 specified to be 48 bits. However, @code{R'Object_Size} will be 64 (bits),
8887 since it must be rounded up so that this value is a multiple of the
8888 alignment (4 bytes = 32 bits).
8891 For all other types, the @code{Object_Size}
8892 and Value_Size are the same (and equivalent to the RM attribute @code{Size}).
8893 Only @code{Size} may be specified for such types.
8895 @node Component_Size Clauses
8896 @section Component_Size Clauses
8897 @cindex Component_Size Clause
8900 Normally, the value specified in a component clause must be consistent
8901 with the subtype of the array component with regard to size and alignment.
8902 In other words, the value specified must be at least equal to the size
8903 of this subtype, and must be a multiple of the alignment value.
8905 In addition, component size clauses are allowed which cause the array
8906 to be packed, by specifying a smaller value. The cases in which this
8907 is allowed are for component size values in the range 1 through 63. The value
8908 specified must not be smaller than the Size of the subtype. GNAT will
8909 accurately honor all packing requests in this range. For example, if
8912 @smallexample @c ada
8913 type r is array (1 .. 8) of Natural;
8914 for r'Component_Size use 31;
8918 then the resulting array has a length of 31 bytes (248 bits = 8 * 31).
8919 Of course access to the components of such an array is considerably
8920 less efficient than if the natural component size of 32 is used.
8922 @node Bit_Order Clauses
8923 @section Bit_Order Clauses
8924 @cindex Bit_Order Clause
8925 @cindex bit ordering
8926 @cindex ordering, of bits
8929 For record subtypes, GNAT permits the specification of the @code{Bit_Order}
8930 attribute. The specification may either correspond to the default bit
8931 order for the target, in which case the specification has no effect and
8932 places no additional restrictions, or it may be for the non-standard
8933 setting (that is the opposite of the default).
8935 In the case where the non-standard value is specified, the effect is
8936 to renumber bits within each byte, but the ordering of bytes is not
8937 affected. There are certain
8938 restrictions placed on component clauses as follows:
8942 @item Components fitting within a single storage unit.
8944 These are unrestricted, and the effect is merely to renumber bits. For
8945 example if we are on a little-endian machine with @code{Low_Order_First}
8946 being the default, then the following two declarations have exactly
8949 @smallexample @c ada
8952 B : Integer range 1 .. 120;
8956 A at 0 range 0 .. 0;
8957 B at 0 range 1 .. 7;
8962 B : Integer range 1 .. 120;
8965 for R2'Bit_Order use High_Order_First;
8968 A at 0 range 7 .. 7;
8969 B at 0 range 0 .. 6;
8974 The useful application here is to write the second declaration with the
8975 @code{Bit_Order} attribute definition clause, and know that it will be treated
8976 the same, regardless of whether the target is little-endian or big-endian.
8978 @item Components occupying an integral number of bytes.
8980 These are components that exactly fit in two or more bytes. Such component
8981 declarations are allowed, but have no effect, since it is important to realize
8982 that the @code{Bit_Order} specification does not affect the ordering of bytes.
8983 In particular, the following attempt at getting an endian-independent integer
8986 @smallexample @c ada
8991 for R2'Bit_Order use High_Order_First;
8994 A at 0 range 0 .. 31;
8999 This declaration will result in a little-endian integer on a
9000 little-endian machine, and a big-endian integer on a big-endian machine.
9001 If byte flipping is required for interoperability between big- and
9002 little-endian machines, this must be explicitly programmed. This capability
9003 is not provided by @code{Bit_Order}.
9005 @item Components that are positioned across byte boundaries
9007 but do not occupy an integral number of bytes. Given that bytes are not
9008 reordered, such fields would occupy a non-contiguous sequence of bits
9009 in memory, requiring non-trivial code to reassemble. They are for this
9010 reason not permitted, and any component clause specifying such a layout
9011 will be flagged as illegal by GNAT@.
9016 Since the misconception that Bit_Order automatically deals with all
9017 endian-related incompatibilities is a common one, the specification of
9018 a component field that is an integral number of bytes will always
9019 generate a warning. This warning may be suppressed using
9020 @code{pragma Suppress} if desired. The following section contains additional
9021 details regarding the issue of byte ordering.
9023 @node Effect of Bit_Order on Byte Ordering
9024 @section Effect of Bit_Order on Byte Ordering
9025 @cindex byte ordering
9026 @cindex ordering, of bytes
9029 In this section we will review the effect of the @code{Bit_Order} attribute
9030 definition clause on byte ordering. Briefly, it has no effect at all, but
9031 a detailed example will be helpful. Before giving this
9032 example, let us review the precise
9033 definition of the effect of defining @code{Bit_Order}. The effect of a
9034 non-standard bit order is described in section 15.5.3 of the Ada
9038 2 A bit ordering is a method of interpreting the meaning of
9039 the storage place attributes.
9043 To understand the precise definition of storage place attributes in
9044 this context, we visit section 13.5.1 of the manual:
9047 13 A record_representation_clause (without the mod_clause)
9048 specifies the layout. The storage place attributes (see 13.5.2)
9049 are taken from the values of the position, first_bit, and last_bit
9050 expressions after normalizing those values so that first_bit is
9051 less than Storage_Unit.
9055 The critical point here is that storage places are taken from
9056 the values after normalization, not before. So the @code{Bit_Order}
9057 interpretation applies to normalized values. The interpretation
9058 is described in the later part of the 15.5.3 paragraph:
9061 2 A bit ordering is a method of interpreting the meaning of
9062 the storage place attributes. High_Order_First (known in the
9063 vernacular as ``big endian'') means that the first bit of a
9064 storage element (bit 0) is the most significant bit (interpreting
9065 the sequence of bits that represent a component as an unsigned
9066 integer value). Low_Order_First (known in the vernacular as
9067 ``little endian'') means the opposite: the first bit is the
9072 Note that the numbering is with respect to the bits of a storage
9073 unit. In other words, the specification affects only the numbering
9074 of bits within a single storage unit.
9076 We can make the effect clearer by giving an example.
9078 Suppose that we have an external device which presents two bytes, the first
9079 byte presented, which is the first (low addressed byte) of the two byte
9080 record is called Master, and the second byte is called Slave.
9082 The left most (most significant bit is called Control for each byte, and
9083 the remaining 7 bits are called V1, V2, @dots{} V7, where V7 is the rightmost
9084 (least significant) bit.
9086 On a big-endian machine, we can write the following representation clause
9088 @smallexample @c ada
9090 Master_Control : Bit;
9098 Slave_Control : Bit;
9109 Master_Control at 0 range 0 .. 0;
9110 Master_V1 at 0 range 1 .. 1;
9111 Master_V2 at 0 range 2 .. 2;
9112 Master_V3 at 0 range 3 .. 3;
9113 Master_V4 at 0 range 4 .. 4;
9114 Master_V5 at 0 range 5 .. 5;
9115 Master_V6 at 0 range 6 .. 6;
9116 Master_V7 at 0 range 7 .. 7;
9117 Slave_Control at 1 range 0 .. 0;
9118 Slave_V1 at 1 range 1 .. 1;
9119 Slave_V2 at 1 range 2 .. 2;
9120 Slave_V3 at 1 range 3 .. 3;
9121 Slave_V4 at 1 range 4 .. 4;
9122 Slave_V5 at 1 range 5 .. 5;
9123 Slave_V6 at 1 range 6 .. 6;
9124 Slave_V7 at 1 range 7 .. 7;
9129 Now if we move this to a little endian machine, then the bit ordering within
9130 the byte is backwards, so we have to rewrite the record rep clause as:
9132 @smallexample @c ada
9134 Master_Control at 0 range 7 .. 7;
9135 Master_V1 at 0 range 6 .. 6;
9136 Master_V2 at 0 range 5 .. 5;
9137 Master_V3 at 0 range 4 .. 4;
9138 Master_V4 at 0 range 3 .. 3;
9139 Master_V5 at 0 range 2 .. 2;
9140 Master_V6 at 0 range 1 .. 1;
9141 Master_V7 at 0 range 0 .. 0;
9142 Slave_Control at 1 range 7 .. 7;
9143 Slave_V1 at 1 range 6 .. 6;
9144 Slave_V2 at 1 range 5 .. 5;
9145 Slave_V3 at 1 range 4 .. 4;
9146 Slave_V4 at 1 range 3 .. 3;
9147 Slave_V5 at 1 range 2 .. 2;
9148 Slave_V6 at 1 range 1 .. 1;
9149 Slave_V7 at 1 range 0 .. 0;
9154 It is a nuisance to have to rewrite the clause, especially if
9155 the code has to be maintained on both machines. However,
9156 this is a case that we can handle with the
9157 @code{Bit_Order} attribute if it is implemented.
9158 Note that the implementation is not required on byte addressed
9159 machines, but it is indeed implemented in GNAT.
9160 This means that we can simply use the
9161 first record clause, together with the declaration
9163 @smallexample @c ada
9164 for Data'Bit_Order use High_Order_First;
9168 and the effect is what is desired, namely the layout is exactly the same,
9169 independent of whether the code is compiled on a big-endian or little-endian
9172 The important point to understand is that byte ordering is not affected.
9173 A @code{Bit_Order} attribute definition never affects which byte a field
9174 ends up in, only where it ends up in that byte.
9175 To make this clear, let us rewrite the record rep clause of the previous
9178 @smallexample @c ada
9179 for Data'Bit_Order use High_Order_First;
9181 Master_Control at 0 range 0 .. 0;
9182 Master_V1 at 0 range 1 .. 1;
9183 Master_V2 at 0 range 2 .. 2;
9184 Master_V3 at 0 range 3 .. 3;
9185 Master_V4 at 0 range 4 .. 4;
9186 Master_V5 at 0 range 5 .. 5;
9187 Master_V6 at 0 range 6 .. 6;
9188 Master_V7 at 0 range 7 .. 7;
9189 Slave_Control at 0 range 8 .. 8;
9190 Slave_V1 at 0 range 9 .. 9;
9191 Slave_V2 at 0 range 10 .. 10;
9192 Slave_V3 at 0 range 11 .. 11;
9193 Slave_V4 at 0 range 12 .. 12;
9194 Slave_V5 at 0 range 13 .. 13;
9195 Slave_V6 at 0 range 14 .. 14;
9196 Slave_V7 at 0 range 15 .. 15;
9201 This is exactly equivalent to saying (a repeat of the first example):
9203 @smallexample @c ada
9204 for Data'Bit_Order use High_Order_First;
9206 Master_Control at 0 range 0 .. 0;
9207 Master_V1 at 0 range 1 .. 1;
9208 Master_V2 at 0 range 2 .. 2;
9209 Master_V3 at 0 range 3 .. 3;
9210 Master_V4 at 0 range 4 .. 4;
9211 Master_V5 at 0 range 5 .. 5;
9212 Master_V6 at 0 range 6 .. 6;
9213 Master_V7 at 0 range 7 .. 7;
9214 Slave_Control at 1 range 0 .. 0;
9215 Slave_V1 at 1 range 1 .. 1;
9216 Slave_V2 at 1 range 2 .. 2;
9217 Slave_V3 at 1 range 3 .. 3;
9218 Slave_V4 at 1 range 4 .. 4;
9219 Slave_V5 at 1 range 5 .. 5;
9220 Slave_V6 at 1 range 6 .. 6;
9221 Slave_V7 at 1 range 7 .. 7;
9226 Why are they equivalent? Well take a specific field, the @code{Slave_V2}
9227 field. The storage place attributes are obtained by normalizing the
9228 values given so that the @code{First_Bit} value is less than 8. After
9229 normalizing the values (0,10,10) we get (1,2,2) which is exactly what
9230 we specified in the other case.
9232 Now one might expect that the @code{Bit_Order} attribute might affect
9233 bit numbering within the entire record component (two bytes in this
9234 case, thus affecting which byte fields end up in), but that is not
9235 the way this feature is defined, it only affects numbering of bits,
9236 not which byte they end up in.
9238 Consequently it never makes sense to specify a starting bit number
9239 greater than 7 (for a byte addressable field) if an attribute
9240 definition for @code{Bit_Order} has been given, and indeed it
9241 may be actively confusing to specify such a value, so the compiler
9242 generates a warning for such usage.
9244 If you do need to control byte ordering then appropriate conditional
9245 values must be used. If in our example, the slave byte came first on
9246 some machines we might write:
9248 @smallexample @c ada
9249 Master_Byte_First constant Boolean := @dots{};
9251 Master_Byte : constant Natural :=
9252 1 - Boolean'Pos (Master_Byte_First);
9253 Slave_Byte : constant Natural :=
9254 Boolean'Pos (Master_Byte_First);
9256 for Data'Bit_Order use High_Order_First;
9258 Master_Control at Master_Byte range 0 .. 0;
9259 Master_V1 at Master_Byte range 1 .. 1;
9260 Master_V2 at Master_Byte range 2 .. 2;
9261 Master_V3 at Master_Byte range 3 .. 3;
9262 Master_V4 at Master_Byte range 4 .. 4;
9263 Master_V5 at Master_Byte range 5 .. 5;
9264 Master_V6 at Master_Byte range 6 .. 6;
9265 Master_V7 at Master_Byte range 7 .. 7;
9266 Slave_Control at Slave_Byte range 0 .. 0;
9267 Slave_V1 at Slave_Byte range 1 .. 1;
9268 Slave_V2 at Slave_Byte range 2 .. 2;
9269 Slave_V3 at Slave_Byte range 3 .. 3;
9270 Slave_V4 at Slave_Byte range 4 .. 4;
9271 Slave_V5 at Slave_Byte range 5 .. 5;
9272 Slave_V6 at Slave_Byte range 6 .. 6;
9273 Slave_V7 at Slave_Byte range 7 .. 7;
9278 Now to switch between machines, all that is necessary is
9279 to set the boolean constant @code{Master_Byte_First} in
9280 an appropriate manner.
9282 @node Pragma Pack for Arrays
9283 @section Pragma Pack for Arrays
9284 @cindex Pragma Pack (for arrays)
9287 Pragma @code{Pack} applied to an array has no effect unless the component type
9288 is packable. For a component type to be packable, it must be one of the
9295 Any type whose size is specified with a size clause
9297 Any packed array type with a static size
9301 For all these cases, if the component subtype size is in the range
9302 1 through 63, then the effect of the pragma @code{Pack} is exactly as though a
9303 component size were specified giving the component subtype size.
9304 For example if we have:
9306 @smallexample @c ada
9307 type r is range 0 .. 17;
9309 type ar is array (1 .. 8) of r;
9314 Then the component size of @code{ar} will be set to 5 (i.e.@: to @code{r'size},
9315 and the size of the array @code{ar} will be exactly 40 bits.
9317 Note that in some cases this rather fierce approach to packing can produce
9318 unexpected effects. For example, in Ada 95, type Natural typically has a
9319 size of 31, meaning that if you pack an array of Natural, you get 31-bit
9320 close packing, which saves a few bits, but results in far less efficient
9321 access. Since many other Ada compilers will ignore such a packing request,
9322 GNAT will generate a warning on some uses of pragma @code{Pack} that it guesses
9323 might not be what is intended. You can easily remove this warning by
9324 using an explicit @code{Component_Size} setting instead, which never generates
9325 a warning, since the intention of the programmer is clear in this case.
9327 GNAT treats packed arrays in one of two ways. If the size of the array is
9328 known at compile time and is less than 64 bits, then internally the array
9329 is represented as a single modular type, of exactly the appropriate number
9330 of bits. If the length is greater than 63 bits, or is not known at compile
9331 time, then the packed array is represented as an array of bytes, and the
9332 length is always a multiple of 8 bits.
9334 Note that to represent a packed array as a modular type, the alignment must
9335 be suitable for the modular type involved. For example, on typical machines
9336 a 32-bit packed array will be represented by a 32-bit modular integer with
9337 an alignment of four bytes. If you explicitly override the default alignment
9338 with an alignment clause that is too small, the modular representation
9339 cannot be used. For example, consider the following set of declarations:
9341 @smallexample @c ada
9342 type R is range 1 .. 3;
9343 type S is array (1 .. 31) of R;
9344 for S'Component_Size use 2;
9346 for S'Alignment use 1;
9350 If the alignment clause were not present, then a 62-bit modular
9351 representation would be chosen (typically with an alignment of 4 or 8
9352 bytes depending on the target). But the default alignment is overridden
9353 with the explicit alignment clause. This means that the modular
9354 representation cannot be used, and instead the array of bytes
9355 representation must be used, meaning that the length must be a multiple
9356 of 8. Thus the above set of declarations will result in a diagnostic
9357 rejecting the size clause and noting that the minimum size allowed is 64.
9359 @cindex Pragma Pack (for type Natural)
9360 @cindex Pragma Pack warning
9362 One special case that is worth noting occurs when the base type of the
9363 component size is 8/16/32 and the subtype is one bit less. Notably this
9364 occurs with subtype @code{Natural}. Consider:
9366 @smallexample @c ada
9367 type Arr is array (1 .. 32) of Natural;
9372 In all commonly used Ada 83 compilers, this pragma Pack would be ignored,
9373 since typically @code{Natural'Size} is 32 in Ada 83, and in any case most
9374 Ada 83 compilers did not attempt 31 bit packing.
9376 In Ada 95, @code{Natural'Size} is required to be 31. Furthermore, GNAT really
9377 does pack 31-bit subtype to 31 bits. This may result in a substantial
9378 unintended performance penalty when porting legacy Ada 83 code. To help
9379 prevent this, GNAT generates a warning in such cases. If you really want 31
9380 bit packing in a case like this, you can set the component size explicitly:
9382 @smallexample @c ada
9383 type Arr is array (1 .. 32) of Natural;
9384 for Arr'Component_Size use 31;
9388 Here 31-bit packing is achieved as required, and no warning is generated,
9389 since in this case the programmer intention is clear.
9391 @node Pragma Pack for Records
9392 @section Pragma Pack for Records
9393 @cindex Pragma Pack (for records)
9396 Pragma @code{Pack} applied to a record will pack the components to reduce
9397 wasted space from alignment gaps and by reducing the amount of space
9398 taken by components. We distinguish between @emph{packable} components and
9399 @emph{non-packable} components.
9400 Components of the following types are considered packable:
9403 All primitive types are packable.
9406 Small packed arrays, whose size does not exceed 64 bits, and where the
9407 size is statically known at compile time, are represented internally
9408 as modular integers, and so they are also packable.
9413 All packable components occupy the exact number of bits corresponding to
9414 their @code{Size} value, and are packed with no padding bits, i.e.@: they
9415 can start on an arbitrary bit boundary.
9417 All other types are non-packable, they occupy an integral number of
9419 are placed at a boundary corresponding to their alignment requirements.
9421 For example, consider the record
9423 @smallexample @c ada
9424 type Rb1 is array (1 .. 13) of Boolean;
9427 type Rb2 is array (1 .. 65) of Boolean;
9442 The representation for the record x2 is as follows:
9444 @smallexample @c ada
9445 for x2'Size use 224;
9447 l1 at 0 range 0 .. 0;
9448 l2 at 0 range 1 .. 64;
9449 l3 at 12 range 0 .. 31;
9450 l4 at 16 range 0 .. 0;
9451 l5 at 16 range 1 .. 13;
9452 l6 at 18 range 0 .. 71;
9457 Studying this example, we see that the packable fields @code{l1}
9459 of length equal to their sizes, and placed at specific bit boundaries (and
9460 not byte boundaries) to
9461 eliminate padding. But @code{l3} is of a non-packable float type, so
9462 it is on the next appropriate alignment boundary.
9464 The next two fields are fully packable, so @code{l4} and @code{l5} are
9465 minimally packed with no gaps. However, type @code{Rb2} is a packed
9466 array that is longer than 64 bits, so it is itself non-packable. Thus
9467 the @code{l6} field is aligned to the next byte boundary, and takes an
9468 integral number of bytes, i.e.@: 72 bits.
9470 @node Record Representation Clauses
9471 @section Record Representation Clauses
9472 @cindex Record Representation Clause
9475 Record representation clauses may be given for all record types, including
9476 types obtained by record extension. Component clauses are allowed for any
9477 static component. The restrictions on component clauses depend on the type
9480 @cindex Component Clause
9481 For all components of an elementary type, the only restriction on component
9482 clauses is that the size must be at least the 'Size value of the type
9483 (actually the Value_Size). There are no restrictions due to alignment,
9484 and such components may freely cross storage boundaries.
9486 Packed arrays with a size up to and including 64 bits are represented
9487 internally using a modular type with the appropriate number of bits, and
9488 thus the same lack of restriction applies. For example, if you declare:
9490 @smallexample @c ada
9491 type R is array (1 .. 49) of Boolean;
9497 then a component clause for a component of type R may start on any
9498 specified bit boundary, and may specify a value of 49 bits or greater.
9500 Packed bit arrays that are longer than 64 bits must always be placed
9501 on a storage unit (byte) boundary. Any component clause that does not
9502 meet this requirement will be rejected.
9504 The rules for other types are different for GNAT 3 and GNAT 5 versions
9505 (based on GCC 2 and GCC 3 respectively). In GNAT 5, larger components
9506 (other than packed arrays)
9507 may also be placed on arbitrary boundaries, so for example, the following
9510 @smallexample @c ada
9511 type R is array (1 .. 10) of Boolean;
9520 G at 0 range 0 .. 0;
9521 H at 0 range 1 .. 1;
9522 L at 0 range 2 .. 81;
9523 R at 0 range 82 .. 161;
9528 In GNAT 3, there are more severe restrictions on larger components.
9529 For non-primitive types, including packed arrays with a size greater than
9530 64 bits, component clauses must respect the alignment requirement of the
9531 type, in particular, always starting on a byte boundary, and the length
9532 must be a multiple of the storage unit.
9534 The following rules regarding tagged types are enforced in both GNAT 3 and
9537 The tag field of a tagged type always occupies an address sized field at
9538 the start of the record. No component clause may attempt to overlay this
9541 In the case of a record extension T1, of a type T, no component clause applied
9542 to the type T1 can specify a storage location that would overlap the first
9543 T'Size bytes of the record.
9545 @node Enumeration Clauses
9546 @section Enumeration Clauses
9548 The only restriction on enumeration clauses is that the range of values
9549 must be representable. For the signed case, if one or more of the
9550 representation values are negative, all values must be in the range:
9552 @smallexample @c ada
9553 System.Min_Int .. System.Max_Int
9557 For the unsigned case, where all values are non negative, the values must
9560 @smallexample @c ada
9561 0 .. System.Max_Binary_Modulus;
9565 A @emph{confirming} representation clause is one in which the values range
9566 from 0 in sequence, i.e.@: a clause that confirms the default representation
9567 for an enumeration type.
9568 Such a confirming representation
9569 is permitted by these rules, and is specially recognized by the compiler so
9570 that no extra overhead results from the use of such a clause.
9572 If an array has an index type which is an enumeration type to which an
9573 enumeration clause has been applied, then the array is stored in a compact
9574 manner. Consider the declarations:
9576 @smallexample @c ada
9577 type r is (A, B, C);
9578 for r use (A => 1, B => 5, C => 10);
9579 type t is array (r) of Character;
9583 The array type t corresponds to a vector with exactly three elements and
9584 has a default size equal to @code{3*Character'Size}. This ensures efficient
9585 use of space, but means that accesses to elements of the array will incur
9586 the overhead of converting representation values to the corresponding
9587 positional values, (i.e.@: the value delivered by the @code{Pos} attribute).
9589 @node Address Clauses
9590 @section Address Clauses
9591 @cindex Address Clause
9593 The reference manual allows a general restriction on representation clauses,
9594 as found in RM 13.1(22):
9597 An implementation need not support representation
9598 items containing nonstatic expressions, except that
9599 an implementation should support a representation item
9600 for a given entity if each nonstatic expression in the
9601 representation item is a name that statically denotes
9602 a constant declared before the entity.
9606 In practice this is applicable only to address clauses, since this is the
9607 only case in which a non-static expression is permitted by the syntax. As
9608 the AARM notes in sections 13.1 (22.a-22.h):
9611 22.a Reason: This is to avoid the following sort of thing:
9613 22.b X : Integer := F(@dots{});
9614 Y : Address := G(@dots{});
9615 for X'Address use Y;
9617 22.c In the above, we have to evaluate the
9618 initialization expression for X before we
9619 know where to put the result. This seems
9620 like an unreasonable implementation burden.
9622 22.d The above code should instead be written
9625 22.e Y : constant Address := G(@dots{});
9626 X : Integer := F(@dots{});
9627 for X'Address use Y;
9629 22.f This allows the expression ``Y'' to be safely
9630 evaluated before X is created.
9632 22.g The constant could be a formal parameter of mode in.
9634 22.h An implementation can support other nonstatic
9635 expressions if it wants to. Expressions of type
9636 Address are hardly ever static, but their value
9637 might be known at compile time anyway in many
9642 GNAT does indeed permit many additional cases of non-static expressions. In
9643 particular, if the type involved is elementary there are no restrictions
9644 (since in this case, holding a temporary copy of the initialization value,
9645 if one is present, is inexpensive). In addition, if there is no implicit or
9646 explicit initialization, then there are no restrictions. GNAT will reject
9647 only the case where all three of these conditions hold:
9652 The type of the item is non-elementary (e.g.@: a record or array).
9655 There is explicit or implicit initialization required for the object.
9656 Note that access values are always implicitly initialized, and also
9657 in GNAT, certain bit-packed arrays (those having a dynamic length or
9658 a length greater than 64) will also be implicitly initialized to zero.
9661 The address value is non-static. Here GNAT is more permissive than the
9662 RM, and allows the address value to be the address of a previously declared
9663 stand-alone variable, as long as it does not itself have an address clause.
9665 @smallexample @c ada
9666 Anchor : Some_Initialized_Type;
9667 Overlay : Some_Initialized_Type;
9668 for Overlay'Address use Anchor'Address;
9672 However, the prefix of the address clause cannot be an array component, or
9673 a component of a discriminated record.
9678 As noted above in section 22.h, address values are typically non-static. In
9679 particular the To_Address function, even if applied to a literal value, is
9680 a non-static function call. To avoid this minor annoyance, GNAT provides
9681 the implementation defined attribute 'To_Address. The following two
9682 expressions have identical values:
9686 @smallexample @c ada
9687 To_Address (16#1234_0000#)
9688 System'To_Address (16#1234_0000#);
9692 except that the second form is considered to be a static expression, and
9693 thus when used as an address clause value is always permitted.
9696 Additionally, GNAT treats as static an address clause that is an
9697 unchecked_conversion of a static integer value. This simplifies the porting
9698 of legacy code, and provides a portable equivalent to the GNAT attribute
9701 Another issue with address clauses is the interaction with alignment
9702 requirements. When an address clause is given for an object, the address
9703 value must be consistent with the alignment of the object (which is usually
9704 the same as the alignment of the type of the object). If an address clause
9705 is given that specifies an inappropriately aligned address value, then the
9706 program execution is erroneous.
9708 Since this source of erroneous behavior can have unfortunate effects, GNAT
9709 checks (at compile time if possible, generating a warning, or at execution
9710 time with a run-time check) that the alignment is appropriate. If the
9711 run-time check fails, then @code{Program_Error} is raised. This run-time
9712 check is suppressed if range checks are suppressed, or if
9713 @code{pragma Restrictions (No_Elaboration_Code)} is in effect.
9716 An address clause cannot be given for an exported object. More
9717 understandably the real restriction is that objects with an address
9718 clause cannot be exported. This is because such variables are not
9719 defined by the Ada program, so there is no external object to export.
9722 It is permissible to give an address clause and a pragma Import for the
9723 same object. In this case, the variable is not really defined by the
9724 Ada program, so there is no external symbol to be linked. The link name
9725 and the external name are ignored in this case. The reason that we allow this
9726 combination is that it provides a useful idiom to avoid unwanted
9727 initializations on objects with address clauses.
9729 When an address clause is given for an object that has implicit or
9730 explicit initialization, then by default initialization takes place. This
9731 means that the effect of the object declaration is to overwrite the
9732 memory at the specified address. This is almost always not what the
9733 programmer wants, so GNAT will output a warning:
9743 for Ext'Address use System'To_Address (16#1234_1234#);
9745 >>> warning: implicit initialization of "Ext" may
9746 modify overlaid storage
9747 >>> warning: use pragma Import for "Ext" to suppress
9748 initialization (RM B(24))
9754 As indicated by the warning message, the solution is to use a (dummy) pragma
9755 Import to suppress this initialization. The pragma tell the compiler that the
9756 object is declared and initialized elsewhere. The following package compiles
9757 without warnings (and the initialization is suppressed):
9759 @smallexample @c ada
9767 for Ext'Address use System'To_Address (16#1234_1234#);
9768 pragma Import (Ada, Ext);
9773 A final issue with address clauses involves their use for overlaying
9774 variables, as in the following example:
9775 @cindex Overlaying of objects
9777 @smallexample @c ada
9780 for B'Address use A'Address;
9784 or alternatively, using the form recommended by the RM:
9786 @smallexample @c ada
9788 Addr : constant Address := A'Address;
9790 for B'Address use Addr;
9794 In both of these cases, @code{A}
9795 and @code{B} become aliased to one another via the
9796 address clause. This use of address clauses to overlay
9797 variables, achieving an effect similar to unchecked
9798 conversion was erroneous in Ada 83, but in Ada 95
9799 the effect is implementation defined. Furthermore, the
9800 Ada 95 RM specifically recommends that in a situation
9801 like this, @code{B} should be subject to the following
9802 implementation advice (RM 13.3(19)):
9805 19 If the Address of an object is specified, or it is imported
9806 or exported, then the implementation should not perform
9807 optimizations based on assumptions of no aliases.
9811 GNAT follows this recommendation, and goes further by also applying
9812 this recommendation to the overlaid variable (@code{A}
9813 in the above example) in this case. This means that the overlay
9814 works "as expected", in that a modification to one of the variables
9815 will affect the value of the other.
9817 @node Effect of Convention on Representation
9818 @section Effect of Convention on Representation
9819 @cindex Convention, effect on representation
9822 Normally the specification of a foreign language convention for a type or
9823 an object has no effect on the chosen representation. In particular, the
9824 representation chosen for data in GNAT generally meets the standard system
9825 conventions, and for example records are laid out in a manner that is
9826 consistent with C@. This means that specifying convention C (for example)
9829 There are three exceptions to this general rule:
9833 @item Convention Fortran and array subtypes
9834 If pragma Convention Fortran is specified for an array subtype, then in
9835 accordance with the implementation advice in section 3.6.2(11) of the
9836 Ada Reference Manual, the array will be stored in a Fortran-compatible
9837 column-major manner, instead of the normal default row-major order.
9839 @item Convention C and enumeration types
9840 GNAT normally stores enumeration types in 8, 16, or 32 bits as required
9841 to accommodate all values of the type. For example, for the enumeration
9844 @smallexample @c ada
9845 type Color is (Red, Green, Blue);
9849 8 bits is sufficient to store all values of the type, so by default, objects
9850 of type @code{Color} will be represented using 8 bits. However, normal C
9851 convention is to use 32 bits for all enum values in C, since enum values
9852 are essentially of type int. If pragma @code{Convention C} is specified for an
9853 Ada enumeration type, then the size is modified as necessary (usually to
9854 32 bits) to be consistent with the C convention for enum values.
9856 @item Convention C/Fortran and Boolean types
9857 In C, the usual convention for boolean values, that is values used for
9858 conditions, is that zero represents false, and nonzero values represent
9859 true. In Ada, the normal convention is that two specific values, typically
9860 0/1, are used to represent false/true respectively.
9862 Fortran has a similar convention for @code{LOGICAL} values (any nonzero
9863 value represents true).
9865 To accommodate the Fortran and C conventions, if a pragma Convention specifies
9866 C or Fortran convention for a derived Boolean, as in the following example:
9868 @smallexample @c ada
9869 type C_Switch is new Boolean;
9870 pragma Convention (C, C_Switch);
9874 then the GNAT generated code will treat any nonzero value as true. For truth
9875 values generated by GNAT, the conventional value 1 will be used for True, but
9876 when one of these values is read, any nonzero value is treated as True.
9880 @node Determining the Representations chosen by GNAT
9881 @section Determining the Representations chosen by GNAT
9882 @cindex Representation, determination of
9883 @cindex @code{-gnatR} switch
9886 Although the descriptions in this section are intended to be complete, it is
9887 often easier to simply experiment to see what GNAT accepts and what the
9888 effect is on the layout of types and objects.
9890 As required by the Ada RM, if a representation clause is not accepted, then
9891 it must be rejected as illegal by the compiler. However, when a
9892 representation clause or pragma is accepted, there can still be questions
9893 of what the compiler actually does. For example, if a partial record
9894 representation clause specifies the location of some components and not
9895 others, then where are the non-specified components placed? Or if pragma
9896 @code{Pack} is used on a record, then exactly where are the resulting
9897 fields placed? The section on pragma @code{Pack} in this chapter can be
9898 used to answer the second question, but it is often easier to just see
9899 what the compiler does.
9901 For this purpose, GNAT provides the option @code{-gnatR}. If you compile
9902 with this option, then the compiler will output information on the actual
9903 representations chosen, in a format similar to source representation
9904 clauses. For example, if we compile the package:
9906 @smallexample @c ada
9908 type r (x : boolean) is tagged record
9910 when True => S : String (1 .. 100);
9915 type r2 is new r (false) with record
9920 y2 at 16 range 0 .. 31;
9927 type x1 is array (1 .. 10) of x;
9928 for x1'component_size use 11;
9930 type ia is access integer;
9932 type Rb1 is array (1 .. 13) of Boolean;
9935 type Rb2 is array (1 .. 65) of Boolean;
9951 using the switch @code{-gnatR} we obtain the following output:
9954 Representation information for unit q
9955 -------------------------------------
9958 for r'Alignment use 4;
9960 x at 4 range 0 .. 7;
9961 _tag at 0 range 0 .. 31;
9962 s at 5 range 0 .. 799;
9965 for r2'Size use 160;
9966 for r2'Alignment use 4;
9968 x at 4 range 0 .. 7;
9969 _tag at 0 range 0 .. 31;
9970 _parent at 0 range 0 .. 63;
9971 y2 at 16 range 0 .. 31;
9975 for x'Alignment use 1;
9977 y at 0 range 0 .. 7;
9980 for x1'Size use 112;
9981 for x1'Alignment use 1;
9982 for x1'Component_Size use 11;
9984 for rb1'Size use 13;
9985 for rb1'Alignment use 2;
9986 for rb1'Component_Size use 1;
9988 for rb2'Size use 72;
9989 for rb2'Alignment use 1;
9990 for rb2'Component_Size use 1;
9992 for x2'Size use 224;
9993 for x2'Alignment use 4;
9995 l1 at 0 range 0 .. 0;
9996 l2 at 0 range 1 .. 64;
9997 l3 at 12 range 0 .. 31;
9998 l4 at 16 range 0 .. 0;
9999 l5 at 16 range 1 .. 13;
10000 l6 at 18 range 0 .. 71;
10005 The Size values are actually the Object_Size, i.e.@: the default size that
10006 will be allocated for objects of the type.
10007 The ?? size for type r indicates that we have a variant record, and the
10008 actual size of objects will depend on the discriminant value.
10010 The Alignment values show the actual alignment chosen by the compiler
10011 for each record or array type.
10013 The record representation clause for type r shows where all fields
10014 are placed, including the compiler generated tag field (whose location
10015 cannot be controlled by the programmer).
10017 The record representation clause for the type extension r2 shows all the
10018 fields present, including the parent field, which is a copy of the fields
10019 of the parent type of r2, i.e.@: r1.
10021 The component size and size clauses for types rb1 and rb2 show
10022 the exact effect of pragma @code{Pack} on these arrays, and the record
10023 representation clause for type x2 shows how pragma @code{Pack} affects
10026 In some cases, it may be useful to cut and paste the representation clauses
10027 generated by the compiler into the original source to fix and guarantee
10028 the actual representation to be used.
10030 @node Standard Library Routines
10031 @chapter Standard Library Routines
10034 The Ada 95 Reference Manual contains in Annex A a full description of an
10035 extensive set of standard library routines that can be used in any Ada
10036 program, and which must be provided by all Ada compilers. They are
10037 analogous to the standard C library used by C programs.
10039 GNAT implements all of the facilities described in annex A, and for most
10040 purposes the description in the Ada 95
10041 reference manual, or appropriate Ada
10042 text book, will be sufficient for making use of these facilities.
10044 In the case of the input-output facilities, @xref{The Implementation of
10045 Standard I/O}, gives details on exactly how GNAT interfaces to the
10046 file system. For the remaining packages, the Ada 95 reference manual
10047 should be sufficient. The following is a list of the packages included,
10048 together with a brief description of the functionality that is provided.
10050 For completeness, references are included to other predefined library
10051 routines defined in other sections of the Ada 95 reference manual (these are
10052 cross-indexed from annex A).
10056 This is a parent package for all the standard library packages. It is
10057 usually included implicitly in your program, and itself contains no
10058 useful data or routines.
10060 @item Ada.Calendar (9.6)
10061 @code{Calendar} provides time of day access, and routines for
10062 manipulating times and durations.
10064 @item Ada.Characters (A.3.1)
10065 This is a dummy parent package that contains no useful entities
10067 @item Ada.Characters.Handling (A.3.2)
10068 This package provides some basic character handling capabilities,
10069 including classification functions for classes of characters (e.g.@: test
10070 for letters, or digits).
10072 @item Ada.Characters.Latin_1 (A.3.3)
10073 This package includes a complete set of definitions of the characters
10074 that appear in type CHARACTER@. It is useful for writing programs that
10075 will run in international environments. For example, if you want an
10076 upper case E with an acute accent in a string, it is often better to use
10077 the definition of @code{UC_E_Acute} in this package. Then your program
10078 will print in an understandable manner even if your environment does not
10079 support these extended characters.
10081 @item Ada.Command_Line (A.15)
10082 This package provides access to the command line parameters and the name
10083 of the current program (analogous to the use of @code{argc} and @code{argv}
10084 in C), and also allows the exit status for the program to be set in a
10085 system-independent manner.
10087 @item Ada.Decimal (F.2)
10088 This package provides constants describing the range of decimal numbers
10089 implemented, and also a decimal divide routine (analogous to the COBOL
10090 verb DIVIDE .. GIVING .. REMAINDER ..)
10092 @item Ada.Direct_IO (A.8.4)
10093 This package provides input-output using a model of a set of records of
10094 fixed-length, containing an arbitrary definite Ada type, indexed by an
10095 integer record number.
10097 @item Ada.Dynamic_Priorities (D.5)
10098 This package allows the priorities of a task to be adjusted dynamically
10099 as the task is running.
10101 @item Ada.Exceptions (11.4.1)
10102 This package provides additional information on exceptions, and also
10103 contains facilities for treating exceptions as data objects, and raising
10104 exceptions with associated messages.
10106 @item Ada.Finalization (7.6)
10107 This package contains the declarations and subprograms to support the
10108 use of controlled types, providing for automatic initialization and
10109 finalization (analogous to the constructors and destructors of C++)
10111 @item Ada.Interrupts (C.3.2)
10112 This package provides facilities for interfacing to interrupts, which
10113 includes the set of signals or conditions that can be raised and
10114 recognized as interrupts.
10116 @item Ada.Interrupts.Names (C.3.2)
10117 This package provides the set of interrupt names (actually signal
10118 or condition names) that can be handled by GNAT@.
10120 @item Ada.IO_Exceptions (A.13)
10121 This package defines the set of exceptions that can be raised by use of
10122 the standard IO packages.
10125 This package contains some standard constants and exceptions used
10126 throughout the numerics packages. Note that the constants pi and e are
10127 defined here, and it is better to use these definitions than rolling
10130 @item Ada.Numerics.Complex_Elementary_Functions
10131 Provides the implementation of standard elementary functions (such as
10132 log and trigonometric functions) operating on complex numbers using the
10133 standard @code{Float} and the @code{Complex} and @code{Imaginary} types
10134 created by the package @code{Numerics.Complex_Types}.
10136 @item Ada.Numerics.Complex_Types
10137 This is a predefined instantiation of
10138 @code{Numerics.Generic_Complex_Types} using @code{Standard.Float} to
10139 build the type @code{Complex} and @code{Imaginary}.
10141 @item Ada.Numerics.Discrete_Random
10142 This package provides a random number generator suitable for generating
10143 random integer values from a specified range.
10145 @item Ada.Numerics.Float_Random
10146 This package provides a random number generator suitable for generating
10147 uniformly distributed floating point values.
10149 @item Ada.Numerics.Generic_Complex_Elementary_Functions
10150 This is a generic version of the package that provides the
10151 implementation of standard elementary functions (such as log and
10152 trigonometric functions) for an arbitrary complex type.
10154 The following predefined instantiations of this package are provided:
10158 @code{Ada.Numerics.Short_Complex_Elementary_Functions}
10160 @code{Ada.Numerics.Complex_Elementary_Functions}
10162 @code{Ada.Numerics.
10163 Long_Complex_Elementary_Functions}
10166 @item Ada.Numerics.Generic_Complex_Types
10167 This is a generic package that allows the creation of complex types,
10168 with associated complex arithmetic operations.
10170 The following predefined instantiations of this package exist
10173 @code{Ada.Numerics.Short_Complex_Complex_Types}
10175 @code{Ada.Numerics.Complex_Complex_Types}
10177 @code{Ada.Numerics.Long_Complex_Complex_Types}
10180 @item Ada.Numerics.Generic_Elementary_Functions
10181 This is a generic package that provides the implementation of standard
10182 elementary functions (such as log an trigonometric functions) for an
10183 arbitrary float type.
10185 The following predefined instantiations of this package exist
10189 @code{Ada.Numerics.Short_Elementary_Functions}
10191 @code{Ada.Numerics.Elementary_Functions}
10193 @code{Ada.Numerics.Long_Elementary_Functions}
10196 @item Ada.Real_Time (D.8)
10197 This package provides facilities similar to those of @code{Calendar}, but
10198 operating with a finer clock suitable for real time control. Note that
10199 annex D requires that there be no backward clock jumps, and GNAT generally
10200 guarantees this behavior, but of course if the external clock on which
10201 the GNAT runtime depends is deliberately reset by some external event,
10202 then such a backward jump may occur.
10204 @item Ada.Sequential_IO (A.8.1)
10205 This package provides input-output facilities for sequential files,
10206 which can contain a sequence of values of a single type, which can be
10207 any Ada type, including indefinite (unconstrained) types.
10209 @item Ada.Storage_IO (A.9)
10210 This package provides a facility for mapping arbitrary Ada types to and
10211 from a storage buffer. It is primarily intended for the creation of new
10214 @item Ada.Streams (13.13.1)
10215 This is a generic package that provides the basic support for the
10216 concept of streams as used by the stream attributes (@code{Input},
10217 @code{Output}, @code{Read} and @code{Write}).
10219 @item Ada.Streams.Stream_IO (A.12.1)
10220 This package is a specialization of the type @code{Streams} defined in
10221 package @code{Streams} together with a set of operations providing
10222 Stream_IO capability. The Stream_IO model permits both random and
10223 sequential access to a file which can contain an arbitrary set of values
10224 of one or more Ada types.
10226 @item Ada.Strings (A.4.1)
10227 This package provides some basic constants used by the string handling
10230 @item Ada.Strings.Bounded (A.4.4)
10231 This package provides facilities for handling variable length
10232 strings. The bounded model requires a maximum length. It is thus
10233 somewhat more limited than the unbounded model, but avoids the use of
10234 dynamic allocation or finalization.
10236 @item Ada.Strings.Fixed (A.4.3)
10237 This package provides facilities for handling fixed length strings.
10239 @item Ada.Strings.Maps (A.4.2)
10240 This package provides facilities for handling character mappings and
10241 arbitrarily defined subsets of characters. For instance it is useful in
10242 defining specialized translation tables.
10244 @item Ada.Strings.Maps.Constants (A.4.6)
10245 This package provides a standard set of predefined mappings and
10246 predefined character sets. For example, the standard upper to lower case
10247 conversion table is found in this package. Note that upper to lower case
10248 conversion is non-trivial if you want to take the entire set of
10249 characters, including extended characters like E with an acute accent,
10250 into account. You should use the mappings in this package (rather than
10251 adding 32 yourself) to do case mappings.
10253 @item Ada.Strings.Unbounded (A.4.5)
10254 This package provides facilities for handling variable length
10255 strings. The unbounded model allows arbitrary length strings, but
10256 requires the use of dynamic allocation and finalization.
10258 @item Ada.Strings.Wide_Bounded (A.4.7)
10259 @itemx Ada.Strings.Wide_Fixed (A.4.7)
10260 @itemx Ada.Strings.Wide_Maps (A.4.7)
10261 @itemx Ada.Strings.Wide_Maps.Constants (A.4.7)
10262 @itemx Ada.Strings.Wide_Unbounded (A.4.7)
10263 These packages provide analogous capabilities to the corresponding
10264 packages without @samp{Wide_} in the name, but operate with the types
10265 @code{Wide_String} and @code{Wide_Character} instead of @code{String}
10266 and @code{Character}.
10268 @item Ada.Synchronous_Task_Control (D.10)
10269 This package provides some standard facilities for controlling task
10270 communication in a synchronous manner.
10273 This package contains definitions for manipulation of the tags of tagged
10276 @item Ada.Task_Attributes
10277 This package provides the capability of associating arbitrary
10278 task-specific data with separate tasks.
10281 This package provides basic text input-output capabilities for
10282 character, string and numeric data. The subpackages of this
10283 package are listed next.
10285 @item Ada.Text_IO.Decimal_IO
10286 Provides input-output facilities for decimal fixed-point types
10288 @item Ada.Text_IO.Enumeration_IO
10289 Provides input-output facilities for enumeration types.
10291 @item Ada.Text_IO.Fixed_IO
10292 Provides input-output facilities for ordinary fixed-point types.
10294 @item Ada.Text_IO.Float_IO
10295 Provides input-output facilities for float types. The following
10296 predefined instantiations of this generic package are available:
10300 @code{Short_Float_Text_IO}
10302 @code{Float_Text_IO}
10304 @code{Long_Float_Text_IO}
10307 @item Ada.Text_IO.Integer_IO
10308 Provides input-output facilities for integer types. The following
10309 predefined instantiations of this generic package are available:
10312 @item Short_Short_Integer
10313 @code{Ada.Short_Short_Integer_Text_IO}
10314 @item Short_Integer
10315 @code{Ada.Short_Integer_Text_IO}
10317 @code{Ada.Integer_Text_IO}
10319 @code{Ada.Long_Integer_Text_IO}
10320 @item Long_Long_Integer
10321 @code{Ada.Long_Long_Integer_Text_IO}
10324 @item Ada.Text_IO.Modular_IO
10325 Provides input-output facilities for modular (unsigned) types
10327 @item Ada.Text_IO.Complex_IO (G.1.3)
10328 This package provides basic text input-output capabilities for complex
10331 @item Ada.Text_IO.Editing (F.3.3)
10332 This package contains routines for edited output, analogous to the use
10333 of pictures in COBOL@. The picture formats used by this package are a
10334 close copy of the facility in COBOL@.
10336 @item Ada.Text_IO.Text_Streams (A.12.2)
10337 This package provides a facility that allows Text_IO files to be treated
10338 as streams, so that the stream attributes can be used for writing
10339 arbitrary data, including binary data, to Text_IO files.
10341 @item Ada.Unchecked_Conversion (13.9)
10342 This generic package allows arbitrary conversion from one type to
10343 another of the same size, providing for breaking the type safety in
10344 special circumstances.
10346 If the types have the same Size (more accurately the same Value_Size),
10347 then the effect is simply to transfer the bits from the source to the
10348 target type without any modification. This usage is well defined, and
10349 for simple types whose representation is typically the same across
10350 all implementations, gives a portable method of performing such
10353 If the types do not have the same size, then the result is implementation
10354 defined, and thus may be non-portable. The following describes how GNAT
10355 handles such unchecked conversion cases.
10357 If the types are of different sizes, and are both discrete types, then
10358 the effect is of a normal type conversion without any constraint checking.
10359 In particular if the result type has a larger size, the result will be
10360 zero or sign extended. If the result type has a smaller size, the result
10361 will be truncated by ignoring high order bits.
10363 If the types are of different sizes, and are not both discrete types,
10364 then the conversion works as though pointers were created to the source
10365 and target, and the pointer value is converted. The effect is that bits
10366 are copied from successive low order storage units and bits of the source
10367 up to the length of the target type.
10369 A warning is issued if the lengths differ, since the effect in this
10370 case is implementation dependent, and the above behavior may not match
10371 that of some other compiler.
10373 A pointer to one type may be converted to a pointer to another type using
10374 unchecked conversion. The only case in which the effect is undefined is
10375 when one or both pointers are pointers to unconstrained array types. In
10376 this case, the bounds information may get incorrectly transferred, and in
10377 particular, GNAT uses double size pointers for such types, and it is
10378 meaningless to convert between such pointer types. GNAT will issue a
10379 warning if the alignment of the target designated type is more strict
10380 than the alignment of the source designated type (since the result may
10381 be unaligned in this case).
10383 A pointer other than a pointer to an unconstrained array type may be
10384 converted to and from System.Address. Such usage is common in Ada 83
10385 programs, but note that Ada.Address_To_Access_Conversions is the
10386 preferred method of performing such conversions in Ada 95. Neither
10387 unchecked conversion nor Ada.Address_To_Access_Conversions should be
10388 used in conjunction with pointers to unconstrained objects, since
10389 the bounds information cannot be handled correctly in this case.
10391 @item Ada.Unchecked_Deallocation (13.11.2)
10392 This generic package allows explicit freeing of storage previously
10393 allocated by use of an allocator.
10395 @item Ada.Wide_Text_IO (A.11)
10396 This package is similar to @code{Ada.Text_IO}, except that the external
10397 file supports wide character representations, and the internal types are
10398 @code{Wide_Character} and @code{Wide_String} instead of @code{Character}
10399 and @code{String}. It contains generic subpackages listed next.
10401 @item Ada.Wide_Text_IO.Decimal_IO
10402 Provides input-output facilities for decimal fixed-point types
10404 @item Ada.Wide_Text_IO.Enumeration_IO
10405 Provides input-output facilities for enumeration types.
10407 @item Ada.Wide_Text_IO.Fixed_IO
10408 Provides input-output facilities for ordinary fixed-point types.
10410 @item Ada.Wide_Text_IO.Float_IO
10411 Provides input-output facilities for float types. The following
10412 predefined instantiations of this generic package are available:
10416 @code{Short_Float_Wide_Text_IO}
10418 @code{Float_Wide_Text_IO}
10420 @code{Long_Float_Wide_Text_IO}
10423 @item Ada.Wide_Text_IO.Integer_IO
10424 Provides input-output facilities for integer types. The following
10425 predefined instantiations of this generic package are available:
10428 @item Short_Short_Integer
10429 @code{Ada.Short_Short_Integer_Wide_Text_IO}
10430 @item Short_Integer
10431 @code{Ada.Short_Integer_Wide_Text_IO}
10433 @code{Ada.Integer_Wide_Text_IO}
10435 @code{Ada.Long_Integer_Wide_Text_IO}
10436 @item Long_Long_Integer
10437 @code{Ada.Long_Long_Integer_Wide_Text_IO}
10440 @item Ada.Wide_Text_IO.Modular_IO
10441 Provides input-output facilities for modular (unsigned) types
10443 @item Ada.Wide_Text_IO.Complex_IO (G.1.3)
10444 This package is similar to @code{Ada.Text_IO.Complex_IO}, except that the
10445 external file supports wide character representations.
10447 @item Ada.Wide_Text_IO.Editing (F.3.4)
10448 This package is similar to @code{Ada.Text_IO.Editing}, except that the
10449 types are @code{Wide_Character} and @code{Wide_String} instead of
10450 @code{Character} and @code{String}.
10452 @item Ada.Wide_Text_IO.Streams (A.12.3)
10453 This package is similar to @code{Ada.Text_IO.Streams}, except that the
10454 types are @code{Wide_Character} and @code{Wide_String} instead of
10455 @code{Character} and @code{String}.
10458 @node The Implementation of Standard I/O
10459 @chapter The Implementation of Standard I/O
10462 GNAT implements all the required input-output facilities described in
10463 A.6 through A.14. These sections of the Ada 95 reference manual describe the
10464 required behavior of these packages from the Ada point of view, and if
10465 you are writing a portable Ada program that does not need to know the
10466 exact manner in which Ada maps to the outside world when it comes to
10467 reading or writing external files, then you do not need to read this
10468 chapter. As long as your files are all regular files (not pipes or
10469 devices), and as long as you write and read the files only from Ada, the
10470 description in the Ada 95 reference manual is sufficient.
10472 However, if you want to do input-output to pipes or other devices, such
10473 as the keyboard or screen, or if the files you are dealing with are
10474 either generated by some other language, or to be read by some other
10475 language, then you need to know more about the details of how the GNAT
10476 implementation of these input-output facilities behaves.
10478 In this chapter we give a detailed description of exactly how GNAT
10479 interfaces to the file system. As always, the sources of the system are
10480 available to you for answering questions at an even more detailed level,
10481 but for most purposes the information in this chapter will suffice.
10483 Another reason that you may need to know more about how input-output is
10484 implemented arises when you have a program written in mixed languages
10485 where, for example, files are shared between the C and Ada sections of
10486 the same program. GNAT provides some additional facilities, in the form
10487 of additional child library packages, that facilitate this sharing, and
10488 these additional facilities are also described in this chapter.
10491 * Standard I/O Packages::
10500 * Operations on C Streams::
10501 * Interfacing to C Streams::
10504 @node Standard I/O Packages
10505 @section Standard I/O Packages
10508 The Standard I/O packages described in Annex A for
10514 Ada.Text_IO.Complex_IO
10516 Ada.Text_IO.Text_Streams,
10520 Ada.Wide_Text_IO.Complex_IO,
10522 Ada.Wide_Text_IO.Text_Streams
10532 are implemented using the C
10533 library streams facility; where
10537 All files are opened using @code{fopen}.
10539 All input/output operations use @code{fread}/@code{fwrite}.
10543 There is no internal buffering of any kind at the Ada library level. The
10544 only buffering is that provided at the system level in the
10545 implementation of the C library routines that support streams. This
10546 facilitates shared use of these streams by mixed language programs.
10549 @section FORM Strings
10552 The format of a FORM string in GNAT is:
10555 "keyword=value,keyword=value,@dots{},keyword=value"
10559 where letters may be in upper or lower case, and there are no spaces
10560 between values. The order of the entries is not important. Currently
10561 there are two keywords defined.
10569 The use of these parameters is described later in this section.
10575 Direct_IO can only be instantiated for definite types. This is a
10576 restriction of the Ada language, which means that the records are fixed
10577 length (the length being determined by @code{@var{type}'Size}, rounded
10578 up to the next storage unit boundary if necessary).
10580 The records of a Direct_IO file are simply written to the file in index
10581 sequence, with the first record starting at offset zero, and subsequent
10582 records following. There is no control information of any kind. For
10583 example, if 32-bit integers are being written, each record takes
10584 4-bytes, so the record at index @var{K} starts at offset
10585 (@var{K}@minus{}1)*4.
10587 There is no limit on the size of Direct_IO files, they are expanded as
10588 necessary to accommodate whatever records are written to the file.
10590 @node Sequential_IO
10591 @section Sequential_IO
10594 Sequential_IO may be instantiated with either a definite (constrained)
10595 or indefinite (unconstrained) type.
10597 For the definite type case, the elements written to the file are simply
10598 the memory images of the data values with no control information of any
10599 kind. The resulting file should be read using the same type, no validity
10600 checking is performed on input.
10602 For the indefinite type case, the elements written consist of two
10603 parts. First is the size of the data item, written as the memory image
10604 of a @code{Interfaces.C.size_t} value, followed by the memory image of
10605 the data value. The resulting file can only be read using the same
10606 (unconstrained) type. Normal assignment checks are performed on these
10607 read operations, and if these checks fail, @code{Data_Error} is
10608 raised. In particular, in the array case, the lengths must match, and in
10609 the variant record case, if the variable for a particular read operation
10610 is constrained, the discriminants must match.
10612 Note that it is not possible to use Sequential_IO to write variable
10613 length array items, and then read the data back into different length
10614 arrays. For example, the following will raise @code{Data_Error}:
10616 @smallexample @c ada
10617 package IO is new Sequential_IO (String);
10622 IO.Write (F, "hello!")
10623 IO.Reset (F, Mode=>In_File);
10630 On some Ada implementations, this will print @code{hell}, but the program is
10631 clearly incorrect, since there is only one element in the file, and that
10632 element is the string @code{hello!}.
10634 In Ada 95, this kind of behavior can be legitimately achieved using
10635 Stream_IO, and this is the preferred mechanism. In particular, the above
10636 program fragment rewritten to use Stream_IO will work correctly.
10642 Text_IO files consist of a stream of characters containing the following
10643 special control characters:
10646 LF (line feed, 16#0A#) Line Mark
10647 FF (form feed, 16#0C#) Page Mark
10651 A canonical Text_IO file is defined as one in which the following
10652 conditions are met:
10656 The character @code{LF} is used only as a line mark, i.e.@: to mark the end
10660 The character @code{FF} is used only as a page mark, i.e.@: to mark the
10661 end of a page and consequently can appear only immediately following a
10662 @code{LF} (line mark) character.
10665 The file ends with either @code{LF} (line mark) or @code{LF}-@code{FF}
10666 (line mark, page mark). In the former case, the page mark is implicitly
10667 assumed to be present.
10671 A file written using Text_IO will be in canonical form provided that no
10672 explicit @code{LF} or @code{FF} characters are written using @code{Put}
10673 or @code{Put_Line}. There will be no @code{FF} character at the end of
10674 the file unless an explicit @code{New_Page} operation was performed
10675 before closing the file.
10677 A canonical Text_IO file that is a regular file, i.e.@: not a device or a
10678 pipe, can be read using any of the routines in Text_IO@. The
10679 semantics in this case will be exactly as defined in the Ada 95 reference
10680 manual and all the routines in Text_IO are fully implemented.
10682 A text file that does not meet the requirements for a canonical Text_IO
10683 file has one of the following:
10687 The file contains @code{FF} characters not immediately following a
10688 @code{LF} character.
10691 The file contains @code{LF} or @code{FF} characters written by
10692 @code{Put} or @code{Put_Line}, which are not logically considered to be
10693 line marks or page marks.
10696 The file ends in a character other than @code{LF} or @code{FF},
10697 i.e.@: there is no explicit line mark or page mark at the end of the file.
10701 Text_IO can be used to read such non-standard text files but subprograms
10702 to do with line or page numbers do not have defined meanings. In
10703 particular, a @code{FF} character that does not follow a @code{LF}
10704 character may or may not be treated as a page mark from the point of
10705 view of page and line numbering. Every @code{LF} character is considered
10706 to end a line, and there is an implied @code{LF} character at the end of
10710 * Text_IO Stream Pointer Positioning::
10711 * Text_IO Reading and Writing Non-Regular Files::
10713 * Treating Text_IO Files as Streams::
10714 * Text_IO Extensions::
10715 * Text_IO Facilities for Unbounded Strings::
10718 @node Text_IO Stream Pointer Positioning
10719 @subsection Stream Pointer Positioning
10722 @code{Ada.Text_IO} has a definition of current position for a file that
10723 is being read. No internal buffering occurs in Text_IO, and usually the
10724 physical position in the stream used to implement the file corresponds
10725 to this logical position defined by Text_IO@. There are two exceptions:
10729 After a call to @code{End_Of_Page} that returns @code{True}, the stream
10730 is positioned past the @code{LF} (line mark) that precedes the page
10731 mark. Text_IO maintains an internal flag so that subsequent read
10732 operations properly handle the logical position which is unchanged by
10733 the @code{End_Of_Page} call.
10736 After a call to @code{End_Of_File} that returns @code{True}, if the
10737 Text_IO file was positioned before the line mark at the end of file
10738 before the call, then the logical position is unchanged, but the stream
10739 is physically positioned right at the end of file (past the line mark,
10740 and past a possible page mark following the line mark. Again Text_IO
10741 maintains internal flags so that subsequent read operations properly
10742 handle the logical position.
10746 These discrepancies have no effect on the observable behavior of
10747 Text_IO, but if a single Ada stream is shared between a C program and
10748 Ada program, or shared (using @samp{shared=yes} in the form string)
10749 between two Ada files, then the difference may be observable in some
10752 @node Text_IO Reading and Writing Non-Regular Files
10753 @subsection Reading and Writing Non-Regular Files
10756 A non-regular file is a device (such as a keyboard), or a pipe. Text_IO
10757 can be used for reading and writing. Writing is not affected and the
10758 sequence of characters output is identical to the normal file case, but
10759 for reading, the behavior of Text_IO is modified to avoid undesirable
10760 look-ahead as follows:
10762 An input file that is not a regular file is considered to have no page
10763 marks. Any @code{Ascii.FF} characters (the character normally used for a
10764 page mark) appearing in the file are considered to be data
10765 characters. In particular:
10769 @code{Get_Line} and @code{Skip_Line} do not test for a page mark
10770 following a line mark. If a page mark appears, it will be treated as a
10774 This avoids the need to wait for an extra character to be typed or
10775 entered from the pipe to complete one of these operations.
10778 @code{End_Of_Page} always returns @code{False}
10781 @code{End_Of_File} will return @code{False} if there is a page mark at
10782 the end of the file.
10786 Output to non-regular files is the same as for regular files. Page marks
10787 may be written to non-regular files using @code{New_Page}, but as noted
10788 above they will not be treated as page marks on input if the output is
10789 piped to another Ada program.
10791 Another important discrepancy when reading non-regular files is that the end
10792 of file indication is not ``sticky''. If an end of file is entered, e.g.@: by
10793 pressing the @key{EOT} key,
10795 is signaled once (i.e.@: the test @code{End_Of_File}
10796 will yield @code{True}, or a read will
10797 raise @code{End_Error}), but then reading can resume
10798 to read data past that end of
10799 file indication, until another end of file indication is entered.
10801 @node Get_Immediate
10802 @subsection Get_Immediate
10803 @cindex Get_Immediate
10806 Get_Immediate returns the next character (including control characters)
10807 from the input file. In particular, Get_Immediate will return LF or FF
10808 characters used as line marks or page marks. Such operations leave the
10809 file positioned past the control character, and it is thus not treated
10810 as having its normal function. This means that page, line and column
10811 counts after this kind of Get_Immediate call are set as though the mark
10812 did not occur. In the case where a Get_Immediate leaves the file
10813 positioned between the line mark and page mark (which is not normally
10814 possible), it is undefined whether the FF character will be treated as a
10817 @node Treating Text_IO Files as Streams
10818 @subsection Treating Text_IO Files as Streams
10819 @cindex Stream files
10822 The package @code{Text_IO.Streams} allows a Text_IO file to be treated
10823 as a stream. Data written to a Text_IO file in this stream mode is
10824 binary data. If this binary data contains bytes 16#0A# (@code{LF}) or
10825 16#0C# (@code{FF}), the resulting file may have non-standard
10826 format. Similarly if read operations are used to read from a Text_IO
10827 file treated as a stream, then @code{LF} and @code{FF} characters may be
10828 skipped and the effect is similar to that described above for
10829 @code{Get_Immediate}.
10831 @node Text_IO Extensions
10832 @subsection Text_IO Extensions
10833 @cindex Text_IO extensions
10836 A package GNAT.IO_Aux in the GNAT library provides some useful extensions
10837 to the standard @code{Text_IO} package:
10840 @item function File_Exists (Name : String) return Boolean;
10841 Determines if a file of the given name exists.
10843 @item function Get_Line return String;
10844 Reads a string from the standard input file. The value returned is exactly
10845 the length of the line that was read.
10847 @item function Get_Line (File : Ada.Text_IO.File_Type) return String;
10848 Similar, except that the parameter File specifies the file from which
10849 the string is to be read.
10853 @node Text_IO Facilities for Unbounded Strings
10854 @subsection Text_IO Facilities for Unbounded Strings
10855 @cindex Text_IO for unbounded strings
10856 @cindex Unbounded_String, Text_IO operations
10859 The package @code{Ada.Strings.Unbounded.Text_IO}
10860 in library files @code{a-suteio.ads/adb} contains some GNAT-specific
10861 subprograms useful for Text_IO operations on unbounded strings:
10865 @item function Get_Line (File : File_Type) return Unbounded_String;
10866 Reads a line from the specified file
10867 and returns the result as an unbounded string.
10869 @item procedure Put (File : File_Type; U : Unbounded_String);
10870 Writes the value of the given unbounded string to the specified file
10871 Similar to the effect of
10872 @code{Put (To_String (U))} except that an extra copy is avoided.
10874 @item procedure Put_Line (File : File_Type; U : Unbounded_String);
10875 Writes the value of the given unbounded string to the specified file,
10876 followed by a @code{New_Line}.
10877 Similar to the effect of @code{Put_Line (To_String (U))} except
10878 that an extra copy is avoided.
10882 In the above procedures, @code{File} is of type @code{Ada.Text_IO.File_Type}
10883 and is optional. If the parameter is omitted, then the standard input or
10884 output file is referenced as appropriate.
10886 The package @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} in library
10887 files @file{a-swuwti.ads} and @file{a-swuwti.adb} provides similar extended
10888 @code{Wide_Text_IO} functionality for unbounded wide strings.
10891 @section Wide_Text_IO
10894 @code{Wide_Text_IO} is similar in most respects to Text_IO, except that
10895 both input and output files may contain special sequences that represent
10896 wide character values. The encoding scheme for a given file may be
10897 specified using a FORM parameter:
10904 as part of the FORM string (WCEM = wide character encoding method),
10905 where @var{x} is one of the following characters
10911 Upper half encoding
10923 The encoding methods match those that
10924 can be used in a source
10925 program, but there is no requirement that the encoding method used for
10926 the source program be the same as the encoding method used for files,
10927 and different files may use different encoding methods.
10929 The default encoding method for the standard files, and for opened files
10930 for which no WCEM parameter is given in the FORM string matches the
10931 wide character encoding specified for the main program (the default
10932 being brackets encoding if no coding method was specified with -gnatW).
10936 In this encoding, a wide character is represented by a five character
10944 where @var{a}, @var{b}, @var{c}, @var{d} are the four hexadecimal
10945 characters (using upper case letters) of the wide character code. For
10946 example, ESC A345 is used to represent the wide character with code
10947 16#A345#. This scheme is compatible with use of the full
10948 @code{Wide_Character} set.
10950 @item Upper Half Coding
10951 The wide character with encoding 16#abcd#, where the upper bit is on
10952 (i.e.@: a is in the range 8-F) is represented as two bytes 16#ab# and
10953 16#cd#. The second byte may never be a format control character, but is
10954 not required to be in the upper half. This method can be also used for
10955 shift-JIS or EUC where the internal coding matches the external coding.
10957 @item Shift JIS Coding
10958 A wide character is represented by a two character sequence 16#ab# and
10959 16#cd#, with the restrictions described for upper half encoding as
10960 described above. The internal character code is the corresponding JIS
10961 character according to the standard algorithm for Shift-JIS
10962 conversion. Only characters defined in the JIS code set table can be
10963 used with this encoding method.
10966 A wide character is represented by a two character sequence 16#ab# and
10967 16#cd#, with both characters being in the upper half. The internal
10968 character code is the corresponding JIS character according to the EUC
10969 encoding algorithm. Only characters defined in the JIS code set table
10970 can be used with this encoding method.
10973 A wide character is represented using
10974 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
10975 10646-1/Am.2. Depending on the character value, the representation
10976 is a one, two, or three byte sequence:
10979 16#0000#-16#007f#: 2#0xxxxxxx#
10980 16#0080#-16#07ff#: 2#110xxxxx# 2#10xxxxxx#
10981 16#0800#-16#ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
10985 where the xxx bits correspond to the left-padded bits of the
10986 16-bit character value. Note that all lower half ASCII characters
10987 are represented as ASCII bytes and all upper half characters and
10988 other wide characters are represented as sequences of upper-half
10989 (The full UTF-8 scheme allows for encoding 31-bit characters as
10990 6-byte sequences, but in this implementation, all UTF-8 sequences
10991 of four or more bytes length will raise a Constraint_Error, as
10992 will all invalid UTF-8 sequences.)
10994 @item Brackets Coding
10995 In this encoding, a wide character is represented by the following eight
10996 character sequence:
11003 where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
11004 characters (using uppercase letters) of the wide character code. For
11005 example, @code{["A345"]} is used to represent the wide character with code
11007 This scheme is compatible with use of the full Wide_Character set.
11008 On input, brackets coding can also be used for upper half characters,
11009 e.g.@: @code{["C1"]} for lower case a. However, on output, brackets notation
11010 is only used for wide characters with a code greater than @code{16#FF#}.
11015 For the coding schemes other than Hex and Brackets encoding,
11016 not all wide character
11017 values can be represented. An attempt to output a character that cannot
11018 be represented using the encoding scheme for the file causes
11019 Constraint_Error to be raised. An invalid wide character sequence on
11020 input also causes Constraint_Error to be raised.
11023 * Wide_Text_IO Stream Pointer Positioning::
11024 * Wide_Text_IO Reading and Writing Non-Regular Files::
11027 @node Wide_Text_IO Stream Pointer Positioning
11028 @subsection Stream Pointer Positioning
11031 @code{Ada.Wide_Text_IO} is similar to @code{Ada.Text_IO} in its handling
11032 of stream pointer positioning (@pxref{Text_IO}). There is one additional
11035 If @code{Ada.Wide_Text_IO.Look_Ahead} reads a character outside the
11036 normal lower ASCII set (i.e.@: a character in the range:
11038 @smallexample @c ada
11039 Wide_Character'Val (16#0080#) .. Wide_Character'Val (16#FFFF#)
11043 then although the logical position of the file pointer is unchanged by
11044 the @code{Look_Ahead} call, the stream is physically positioned past the
11045 wide character sequence. Again this is to avoid the need for buffering
11046 or backup, and all @code{Wide_Text_IO} routines check the internal
11047 indication that this situation has occurred so that this is not visible
11048 to a normal program using @code{Wide_Text_IO}. However, this discrepancy
11049 can be observed if the wide text file shares a stream with another file.
11051 @node Wide_Text_IO Reading and Writing Non-Regular Files
11052 @subsection Reading and Writing Non-Regular Files
11055 As in the case of Text_IO, when a non-regular file is read, it is
11056 assumed that the file contains no page marks (any form characters are
11057 treated as data characters), and @code{End_Of_Page} always returns
11058 @code{False}. Similarly, the end of file indication is not sticky, so
11059 it is possible to read beyond an end of file.
11065 A stream file is a sequence of bytes, where individual elements are
11066 written to the file as described in the Ada 95 reference manual. The type
11067 @code{Stream_Element} is simply a byte. There are two ways to read or
11068 write a stream file.
11072 The operations @code{Read} and @code{Write} directly read or write a
11073 sequence of stream elements with no control information.
11076 The stream attributes applied to a stream file transfer data in the
11077 manner described for stream attributes.
11081 @section Shared Files
11084 Section A.14 of the Ada 95 Reference Manual allows implementations to
11085 provide a wide variety of behavior if an attempt is made to access the
11086 same external file with two or more internal files.
11088 To provide a full range of functionality, while at the same time
11089 minimizing the problems of portability caused by this implementation
11090 dependence, GNAT handles file sharing as follows:
11094 In the absence of a @samp{shared=@var{xxx}} form parameter, an attempt
11095 to open two or more files with the same full name is considered an error
11096 and is not supported. The exception @code{Use_Error} will be
11097 raised. Note that a file that is not explicitly closed by the program
11098 remains open until the program terminates.
11101 If the form parameter @samp{shared=no} appears in the form string, the
11102 file can be opened or created with its own separate stream identifier,
11103 regardless of whether other files sharing the same external file are
11104 opened. The exact effect depends on how the C stream routines handle
11105 multiple accesses to the same external files using separate streams.
11108 If the form parameter @samp{shared=yes} appears in the form string for
11109 each of two or more files opened using the same full name, the same
11110 stream is shared between these files, and the semantics are as described
11111 in Ada 95 Reference Manual, Section A.14.
11115 When a program that opens multiple files with the same name is ported
11116 from another Ada compiler to GNAT, the effect will be that
11117 @code{Use_Error} is raised.
11119 The documentation of the original compiler and the documentation of the
11120 program should then be examined to determine if file sharing was
11121 expected, and @samp{shared=@var{xxx}} parameters added to @code{Open}
11122 and @code{Create} calls as required.
11124 When a program is ported from GNAT to some other Ada compiler, no
11125 special attention is required unless the @samp{shared=@var{xxx}} form
11126 parameter is used in the program. In this case, you must examine the
11127 documentation of the new compiler to see if it supports the required
11128 file sharing semantics, and form strings modified appropriately. Of
11129 course it may be the case that the program cannot be ported if the
11130 target compiler does not support the required functionality. The best
11131 approach in writing portable code is to avoid file sharing (and hence
11132 the use of the @samp{shared=@var{xxx}} parameter in the form string)
11135 One common use of file sharing in Ada 83 is the use of instantiations of
11136 Sequential_IO on the same file with different types, to achieve
11137 heterogeneous input-output. Although this approach will work in GNAT if
11138 @samp{shared=yes} is specified, it is preferable in Ada 95 to use Stream_IO
11139 for this purpose (using the stream attributes)
11142 @section Open Modes
11145 @code{Open} and @code{Create} calls result in a call to @code{fopen}
11146 using the mode shown in the following table:
11149 @center @code{Open} and @code{Create} Call Modes
11151 @b{OPEN } @b{CREATE}
11152 Append_File "r+" "w+"
11154 Out_File (Direct_IO) "r+" "w"
11155 Out_File (all other cases) "w" "w"
11156 Inout_File "r+" "w+"
11160 If text file translation is required, then either @samp{b} or @samp{t}
11161 is added to the mode, depending on the setting of Text. Text file
11162 translation refers to the mapping of CR/LF sequences in an external file
11163 to LF characters internally. This mapping only occurs in DOS and
11164 DOS-like systems, and is not relevant to other systems.
11166 A special case occurs with Stream_IO@. As shown in the above table, the
11167 file is initially opened in @samp{r} or @samp{w} mode for the
11168 @code{In_File} and @code{Out_File} cases. If a @code{Set_Mode} operation
11169 subsequently requires switching from reading to writing or vice-versa,
11170 then the file is reopened in @samp{r+} mode to permit the required operation.
11172 @node Operations on C Streams
11173 @section Operations on C Streams
11174 The package @code{Interfaces.C_Streams} provides an Ada program with direct
11175 access to the C library functions for operations on C streams:
11177 @smallexample @c adanocomment
11178 package Interfaces.C_Streams is
11179 -- Note: the reason we do not use the types that are in
11180 -- Interfaces.C is that we want to avoid dragging in the
11181 -- code in this unit if possible.
11182 subtype chars is System.Address;
11183 -- Pointer to null-terminated array of characters
11184 subtype FILEs is System.Address;
11185 -- Corresponds to the C type FILE*
11186 subtype voids is System.Address;
11187 -- Corresponds to the C type void*
11188 subtype int is Integer;
11189 subtype long is Long_Integer;
11190 -- Note: the above types are subtypes deliberately, and it
11191 -- is part of this spec that the above correspondences are
11192 -- guaranteed. This means that it is legitimate to, for
11193 -- example, use Integer instead of int. We provide these
11194 -- synonyms for clarity, but in some cases it may be
11195 -- convenient to use the underlying types (for example to
11196 -- avoid an unnecessary dependency of a spec on the spec
11198 type size_t is mod 2 ** Standard'Address_Size;
11199 NULL_Stream : constant FILEs;
11200 -- Value returned (NULL in C) to indicate an
11201 -- fdopen/fopen/tmpfile error
11202 ----------------------------------
11203 -- Constants Defined in stdio.h --
11204 ----------------------------------
11205 EOF : constant int;
11206 -- Used by a number of routines to indicate error or
11208 IOFBF : constant int;
11209 IOLBF : constant int;
11210 IONBF : constant int;
11211 -- Used to indicate buffering mode for setvbuf call
11212 SEEK_CUR : constant int;
11213 SEEK_END : constant int;
11214 SEEK_SET : constant int;
11215 -- Used to indicate origin for fseek call
11216 function stdin return FILEs;
11217 function stdout return FILEs;
11218 function stderr return FILEs;
11219 -- Streams associated with standard files
11220 --------------------------
11221 -- Standard C functions --
11222 --------------------------
11223 -- The functions selected below are ones that are
11224 -- available in DOS, OS/2, UNIX and Xenix (but not
11225 -- necessarily in ANSI C). These are very thin interfaces
11226 -- which copy exactly the C headers. For more
11227 -- documentation on these functions, see the Microsoft C
11228 -- "Run-Time Library Reference" (Microsoft Press, 1990,
11229 -- ISBN 1-55615-225-6), which includes useful information
11230 -- on system compatibility.
11231 procedure clearerr (stream : FILEs);
11232 function fclose (stream : FILEs) return int;
11233 function fdopen (handle : int; mode : chars) return FILEs;
11234 function feof (stream : FILEs) return int;
11235 function ferror (stream : FILEs) return int;
11236 function fflush (stream : FILEs) return int;
11237 function fgetc (stream : FILEs) return int;
11238 function fgets (strng : chars; n : int; stream : FILEs)
11240 function fileno (stream : FILEs) return int;
11241 function fopen (filename : chars; Mode : chars)
11243 -- Note: to maintain target independence, use
11244 -- text_translation_required, a boolean variable defined in
11245 -- a-sysdep.c to deal with the target dependent text
11246 -- translation requirement. If this variable is set,
11247 -- then b/t should be appended to the standard mode
11248 -- argument to set the text translation mode off or on
11250 function fputc (C : int; stream : FILEs) return int;
11251 function fputs (Strng : chars; Stream : FILEs) return int;
11268 function ftell (stream : FILEs) return long;
11275 function isatty (handle : int) return int;
11276 procedure mktemp (template : chars);
11277 -- The return value (which is just a pointer to template)
11279 procedure rewind (stream : FILEs);
11280 function rmtmp return int;
11288 function tmpfile return FILEs;
11289 function ungetc (c : int; stream : FILEs) return int;
11290 function unlink (filename : chars) return int;
11291 ---------------------
11292 -- Extra functions --
11293 ---------------------
11294 -- These functions supply slightly thicker bindings than
11295 -- those above. They are derived from functions in the
11296 -- C Run-Time Library, but may do a bit more work than
11297 -- just directly calling one of the Library functions.
11298 function is_regular_file (handle : int) return int;
11299 -- Tests if given handle is for a regular file (result 1)
11300 -- or for a non-regular file (pipe or device, result 0).
11301 ---------------------------------
11302 -- Control of Text/Binary Mode --
11303 ---------------------------------
11304 -- If text_translation_required is true, then the following
11305 -- functions may be used to dynamically switch a file from
11306 -- binary to text mode or vice versa. These functions have
11307 -- no effect if text_translation_required is false (i.e. in
11308 -- normal UNIX mode). Use fileno to get a stream handle.
11309 procedure set_binary_mode (handle : int);
11310 procedure set_text_mode (handle : int);
11311 ----------------------------
11312 -- Full Path Name support --
11313 ----------------------------
11314 procedure full_name (nam : chars; buffer : chars);
11315 -- Given a NUL terminated string representing a file
11316 -- name, returns in buffer a NUL terminated string
11317 -- representing the full path name for the file name.
11318 -- On systems where it is relevant the drive is also
11319 -- part of the full path name. It is the responsibility
11320 -- of the caller to pass an actual parameter for buffer
11321 -- that is big enough for any full path name. Use
11322 -- max_path_len given below as the size of buffer.
11323 max_path_len : integer;
11324 -- Maximum length of an allowable full path name on the
11325 -- system, including a terminating NUL character.
11326 end Interfaces.C_Streams;
11329 @node Interfacing to C Streams
11330 @section Interfacing to C Streams
11333 The packages in this section permit interfacing Ada files to C Stream
11336 @smallexample @c ada
11337 with Interfaces.C_Streams;
11338 package Ada.Sequential_IO.C_Streams is
11339 function C_Stream (F : File_Type)
11340 return Interfaces.C_Streams.FILEs;
11342 (File : in out File_Type;
11343 Mode : in File_Mode;
11344 C_Stream : in Interfaces.C_Streams.FILEs;
11345 Form : in String := "");
11346 end Ada.Sequential_IO.C_Streams;
11348 with Interfaces.C_Streams;
11349 package Ada.Direct_IO.C_Streams is
11350 function C_Stream (F : File_Type)
11351 return Interfaces.C_Streams.FILEs;
11353 (File : in out File_Type;
11354 Mode : in File_Mode;
11355 C_Stream : in Interfaces.C_Streams.FILEs;
11356 Form : in String := "");
11357 end Ada.Direct_IO.C_Streams;
11359 with Interfaces.C_Streams;
11360 package Ada.Text_IO.C_Streams is
11361 function C_Stream (F : File_Type)
11362 return Interfaces.C_Streams.FILEs;
11364 (File : in out File_Type;
11365 Mode : in File_Mode;
11366 C_Stream : in Interfaces.C_Streams.FILEs;
11367 Form : in String := "");
11368 end Ada.Text_IO.C_Streams;
11370 with Interfaces.C_Streams;
11371 package Ada.Wide_Text_IO.C_Streams is
11372 function C_Stream (F : File_Type)
11373 return Interfaces.C_Streams.FILEs;
11375 (File : in out File_Type;
11376 Mode : in File_Mode;
11377 C_Stream : in Interfaces.C_Streams.FILEs;
11378 Form : in String := "");
11379 end Ada.Wide_Text_IO.C_Streams;
11381 with Interfaces.C_Streams;
11382 package Ada.Stream_IO.C_Streams is
11383 function C_Stream (F : File_Type)
11384 return Interfaces.C_Streams.FILEs;
11386 (File : in out File_Type;
11387 Mode : in File_Mode;
11388 C_Stream : in Interfaces.C_Streams.FILEs;
11389 Form : in String := "");
11390 end Ada.Stream_IO.C_Streams;
11394 In each of these five packages, the @code{C_Stream} function obtains the
11395 @code{FILE} pointer from a currently opened Ada file. It is then
11396 possible to use the @code{Interfaces.C_Streams} package to operate on
11397 this stream, or the stream can be passed to a C program which can
11398 operate on it directly. Of course the program is responsible for
11399 ensuring that only appropriate sequences of operations are executed.
11401 One particular use of relevance to an Ada program is that the
11402 @code{setvbuf} function can be used to control the buffering of the
11403 stream used by an Ada file. In the absence of such a call the standard
11404 default buffering is used.
11406 The @code{Open} procedures in these packages open a file giving an
11407 existing C Stream instead of a file name. Typically this stream is
11408 imported from a C program, allowing an Ada file to operate on an
11411 @node The GNAT Library
11412 @chapter The GNAT Library
11415 The GNAT library contains a number of general and special purpose packages.
11416 It represents functionality that the GNAT developers have found useful, and
11417 which is made available to GNAT users. The packages described here are fully
11418 supported, and upwards compatibility will be maintained in future releases,
11419 so you can use these facilities with the confidence that the same functionality
11420 will be available in future releases.
11422 The chapter here simply gives a brief summary of the facilities available.
11423 The full documentation is found in the spec file for the package. The full
11424 sources of these library packages, including both spec and body, are provided
11425 with all GNAT releases. For example, to find out the full specifications of
11426 the SPITBOL pattern matching capability, including a full tutorial and
11427 extensive examples, look in the @file{g-spipat.ads} file in the library.
11429 For each entry here, the package name (as it would appear in a @code{with}
11430 clause) is given, followed by the name of the corresponding spec file in
11431 parentheses. The packages are children in four hierarchies, @code{Ada},
11432 @code{Interfaces}, @code{System}, and @code{GNAT}, the latter being a
11433 GNAT-specific hierarchy.
11435 Note that an application program should only use packages in one of these
11436 four hierarchies if the package is defined in the Ada Reference Manual,
11437 or is listed in this section of the GNAT Programmers Reference Manual.
11438 All other units should be considered internal implementation units and
11439 should not be directly @code{with}'ed by application code. The use of
11440 a @code{with} statement that references one of these internal implementation
11441 units makes an application potentially dependent on changes in versions
11442 of GNAT, and will generate a warning message.
11445 * Ada.Characters.Latin_9 (a-chlat9.ads)::
11446 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
11447 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
11448 * Ada.Command_Line.Remove (a-colire.ads)::
11449 * Ada.Command_Line.Environment (a-colien.ads)::
11450 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
11451 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
11452 * Ada.Exceptions.Traceback (a-exctra.ads)::
11453 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
11454 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
11455 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
11456 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
11457 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
11458 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
11459 * GNAT.Array_Split (g-arrspl.ads)::
11460 * GNAT.AWK (g-awk.ads)::
11461 * GNAT.Bounded_Buffers (g-boubuf.ads)::
11462 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
11463 * GNAT.Bubble_Sort (g-bubsor.ads)::
11464 * GNAT.Bubble_Sort_A (g-busora.ads)::
11465 * GNAT.Bubble_Sort_G (g-busorg.ads)::
11466 * GNAT.Calendar (g-calend.ads)::
11467 * GNAT.Calendar.Time_IO (g-catiio.ads)::
11468 * GNAT.CRC32 (g-crc32.ads)::
11469 * GNAT.Case_Util (g-casuti.ads)::
11470 * GNAT.CGI (g-cgi.ads)::
11471 * GNAT.CGI.Cookie (g-cgicoo.ads)::
11472 * GNAT.CGI.Debug (g-cgideb.ads)::
11473 * GNAT.Command_Line (g-comlin.ads)::
11474 * GNAT.Compiler_Version (g-comver.ads)::
11475 * GNAT.Ctrl_C (g-ctrl_c.ads)::
11476 * GNAT.Current_Exception (g-curexc.ads)::
11477 * GNAT.Debug_Pools (g-debpoo.ads)::
11478 * GNAT.Debug_Utilities (g-debuti.ads)::
11479 * GNAT.Directory_Operations (g-dirope.ads)::
11480 * GNAT.Dynamic_HTables (g-dynhta.ads)::
11481 * GNAT.Dynamic_Tables (g-dyntab.ads)::
11482 * GNAT.Exception_Actions (g-excact.ads)::
11483 * GNAT.Exception_Traces (g-exctra.ads)::
11484 * GNAT.Exceptions (g-except.ads)::
11485 * GNAT.Expect (g-expect.ads)::
11486 * GNAT.Float_Control (g-flocon.ads)::
11487 * GNAT.Heap_Sort (g-heasor.ads)::
11488 * GNAT.Heap_Sort_A (g-hesora.ads)::
11489 * GNAT.Heap_Sort_G (g-hesorg.ads)::
11490 * GNAT.HTable (g-htable.ads)::
11491 * GNAT.IO (g-io.ads)::
11492 * GNAT.IO_Aux (g-io_aux.ads)::
11493 * GNAT.Lock_Files (g-locfil.ads)::
11494 * GNAT.MD5 (g-md5.ads)::
11495 * GNAT.Memory_Dump (g-memdum.ads)::
11496 * GNAT.Most_Recent_Exception (g-moreex.ads)::
11497 * GNAT.OS_Lib (g-os_lib.ads)::
11498 * GNAT.Perfect_Hash.Generators (g-pehage.ads)::
11499 * GNAT.Regexp (g-regexp.ads)::
11500 * GNAT.Registry (g-regist.ads)::
11501 * GNAT.Regpat (g-regpat.ads)::
11502 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
11503 * GNAT.Semaphores (g-semaph.ads)::
11504 * GNAT.Signals (g-signal.ads)::
11505 * GNAT.Sockets (g-socket.ads)::
11506 * GNAT.Source_Info (g-souinf.ads)::
11507 * GNAT.Spell_Checker (g-speche.ads)::
11508 * GNAT.Spitbol.Patterns (g-spipat.ads)::
11509 * GNAT.Spitbol (g-spitbo.ads)::
11510 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
11511 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
11512 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
11513 * GNAT.Strings (g-string.ads)::
11514 * GNAT.String_Split (g-strspl.ads)::
11515 * GNAT.Table (g-table.ads)::
11516 * GNAT.Task_Lock (g-tasloc.ads)::
11517 * GNAT.Threads (g-thread.ads)::
11518 * GNAT.Traceback (g-traceb.ads)::
11519 * GNAT.Traceback.Symbolic (g-trasym.ads)::
11520 * GNAT.Wide_String_Split (g-wistsp.ads)::
11521 * Interfaces.C.Extensions (i-cexten.ads)::
11522 * Interfaces.C.Streams (i-cstrea.ads)::
11523 * Interfaces.CPP (i-cpp.ads)::
11524 * Interfaces.Os2lib (i-os2lib.ads)::
11525 * Interfaces.Os2lib.Errors (i-os2err.ads)::
11526 * Interfaces.Os2lib.Synchronization (i-os2syn.ads)::
11527 * Interfaces.Os2lib.Threads (i-os2thr.ads)::
11528 * Interfaces.Packed_Decimal (i-pacdec.ads)::
11529 * Interfaces.VxWorks (i-vxwork.ads)::
11530 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
11531 * System.Address_Image (s-addima.ads)::
11532 * System.Assertions (s-assert.ads)::
11533 * System.Memory (s-memory.ads)::
11534 * System.Partition_Interface (s-parint.ads)::
11535 * System.Restrictions (s-restri.ads)::
11536 * System.Rident (s-rident.ads)::
11537 * System.Task_Info (s-tasinf.ads)::
11538 * System.Wch_Cnv (s-wchcnv.ads)::
11539 * System.Wch_Con (s-wchcon.ads)::
11542 @node Ada.Characters.Latin_9 (a-chlat9.ads)
11543 @section @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
11544 @cindex @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
11545 @cindex Latin_9 constants for Character
11548 This child of @code{Ada.Characters}
11549 provides a set of definitions corresponding to those in the
11550 RM-defined package @code{Ada.Characters.Latin_1} but with the
11551 few modifications required for @code{Latin-9}
11552 The provision of such a package
11553 is specifically authorized by the Ada Reference Manual
11556 @node Ada.Characters.Wide_Latin_1 (a-cwila1.ads)
11557 @section @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
11558 @cindex @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
11559 @cindex Latin_1 constants for Wide_Character
11562 This child of @code{Ada.Characters}
11563 provides a set of definitions corresponding to those in the
11564 RM-defined package @code{Ada.Characters.Latin_1} but with the
11565 types of the constants being @code{Wide_Character}
11566 instead of @code{Character}. The provision of such a package
11567 is specifically authorized by the Ada Reference Manual
11570 @node Ada.Characters.Wide_Latin_9 (a-cwila9.ads)
11571 @section @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
11572 @cindex @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
11573 @cindex Latin_9 constants for Wide_Character
11576 This child of @code{Ada.Characters}
11577 provides a set of definitions corresponding to those in the
11578 GNAT defined package @code{Ada.Characters.Latin_9} but with the
11579 types of the constants being @code{Wide_Character}
11580 instead of @code{Character}. The provision of such a package
11581 is specifically authorized by the Ada Reference Manual
11584 @node Ada.Command_Line.Remove (a-colire.ads)
11585 @section @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
11586 @cindex @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
11587 @cindex Removing command line arguments
11588 @cindex Command line, argument removal
11591 This child of @code{Ada.Command_Line}
11592 provides a mechanism for logically removing
11593 arguments from the argument list. Once removed, an argument is not visible
11594 to further calls on the subprograms in @code{Ada.Command_Line} will not
11595 see the removed argument.
11597 @node Ada.Command_Line.Environment (a-colien.ads)
11598 @section @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
11599 @cindex @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
11600 @cindex Environment entries
11603 This child of @code{Ada.Command_Line}
11604 provides a mechanism for obtaining environment values on systems
11605 where this concept makes sense.
11607 @node Ada.Direct_IO.C_Streams (a-diocst.ads)
11608 @section @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
11609 @cindex @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
11610 @cindex C Streams, Interfacing with Direct_IO
11613 This package provides subprograms that allow interfacing between
11614 C streams and @code{Direct_IO}. The stream identifier can be
11615 extracted from a file opened on the Ada side, and an Ada file
11616 can be constructed from a stream opened on the C side.
11618 @node Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)
11619 @section @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
11620 @cindex @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
11621 @cindex Null_Occurrence, testing for
11624 This child subprogram provides a way of testing for the null
11625 exception occurrence (@code{Null_Occurrence}) without raising
11628 @node Ada.Exceptions.Traceback (a-exctra.ads)
11629 @section @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
11630 @cindex @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
11631 @cindex Traceback for Exception Occurrence
11634 This child package provides the subprogram (@code{Tracebacks}) to
11635 give a traceback array of addresses based on an exception
11638 @node Ada.Sequential_IO.C_Streams (a-siocst.ads)
11639 @section @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
11640 @cindex @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
11641 @cindex C Streams, Interfacing with Sequential_IO
11644 This package provides subprograms that allow interfacing between
11645 C streams and @code{Sequential_IO}. The stream identifier can be
11646 extracted from a file opened on the Ada side, and an Ada file
11647 can be constructed from a stream opened on the C side.
11649 @node Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)
11650 @section @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
11651 @cindex @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
11652 @cindex C Streams, Interfacing with Stream_IO
11655 This package provides subprograms that allow interfacing between
11656 C streams and @code{Stream_IO}. The stream identifier can be
11657 extracted from a file opened on the Ada side, and an Ada file
11658 can be constructed from a stream opened on the C side.
11660 @node Ada.Strings.Unbounded.Text_IO (a-suteio.ads)
11661 @section @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
11662 @cindex @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
11663 @cindex @code{Unbounded_String}, IO support
11664 @cindex @code{Text_IO}, extensions for unbounded strings
11667 This package provides subprograms for Text_IO for unbounded
11668 strings, avoiding the necessity for an intermediate operation
11669 with ordinary strings.
11671 @node Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)
11672 @section @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
11673 @cindex @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
11674 @cindex @code{Unbounded_Wide_String}, IO support
11675 @cindex @code{Text_IO}, extensions for unbounded wide strings
11678 This package provides subprograms for Text_IO for unbounded
11679 wide strings, avoiding the necessity for an intermediate operation
11680 with ordinary wide strings.
11682 @node Ada.Text_IO.C_Streams (a-tiocst.ads)
11683 @section @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
11684 @cindex @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
11685 @cindex C Streams, Interfacing with @code{Text_IO}
11688 This package provides subprograms that allow interfacing between
11689 C streams and @code{Text_IO}. The stream identifier can be
11690 extracted from a file opened on the Ada side, and an Ada file
11691 can be constructed from a stream opened on the C side.
11693 @node Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)
11694 @section @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
11695 @cindex @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
11696 @cindex C Streams, Interfacing with @code{Wide_Text_IO}
11699 This package provides subprograms that allow interfacing between
11700 C streams and @code{Wide_Text_IO}. The stream identifier can be
11701 extracted from a file opened on the Ada side, and an Ada file
11702 can be constructed from a stream opened on the C side.
11704 @node GNAT.Array_Split (g-arrspl.ads)
11705 @section @code{GNAT.Array_Split} (@file{g-arrspl.ads})
11706 @cindex @code{GNAT.Array_Split} (@file{g-arrspl.ads})
11707 @cindex Array splitter
11710 Useful array-manipulation routines: given a set of separators, split
11711 an array wherever the separators appear, and provide direct access
11712 to the resulting slices.
11714 @node GNAT.AWK (g-awk.ads)
11715 @section @code{GNAT.AWK} (@file{g-awk.ads})
11716 @cindex @code{GNAT.AWK} (@file{g-awk.ads})
11721 Provides AWK-like parsing functions, with an easy interface for parsing one
11722 or more files containing formatted data. The file is viewed as a database
11723 where each record is a line and a field is a data element in this line.
11725 @node GNAT.Bounded_Buffers (g-boubuf.ads)
11726 @section @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
11727 @cindex @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
11729 @cindex Bounded Buffers
11732 Provides a concurrent generic bounded buffer abstraction. Instances are
11733 useful directly or as parts of the implementations of other abstractions,
11736 @node GNAT.Bounded_Mailboxes (g-boumai.ads)
11737 @section @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
11738 @cindex @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
11743 Provides a thread-safe asynchronous intertask mailbox communication facility.
11745 @node GNAT.Bubble_Sort (g-bubsor.ads)
11746 @section @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
11747 @cindex @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
11749 @cindex Bubble sort
11752 Provides a general implementation of bubble sort usable for sorting arbitrary
11753 data items. Exchange and comparison procedures are provided by passing
11754 access-to-procedure values.
11756 @node GNAT.Bubble_Sort_A (g-busora.ads)
11757 @section @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
11758 @cindex @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
11760 @cindex Bubble sort
11763 Provides a general implementation of bubble sort usable for sorting arbitrary
11764 data items. Move and comparison procedures are provided by passing
11765 access-to-procedure values. This is an older version, retained for
11766 compatibility. Usually @code{GNAT.Bubble_Sort} will be preferable.
11768 @node GNAT.Bubble_Sort_G (g-busorg.ads)
11769 @section @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
11770 @cindex @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
11772 @cindex Bubble sort
11775 Similar to @code{Bubble_Sort_A} except that the move and sorting procedures
11776 are provided as generic parameters, this improves efficiency, especially
11777 if the procedures can be inlined, at the expense of duplicating code for
11778 multiple instantiations.
11780 @node GNAT.Calendar (g-calend.ads)
11781 @section @code{GNAT.Calendar} (@file{g-calend.ads})
11782 @cindex @code{GNAT.Calendar} (@file{g-calend.ads})
11783 @cindex @code{Calendar}
11786 Extends the facilities provided by @code{Ada.Calendar} to include handling
11787 of days of the week, an extended @code{Split} and @code{Time_Of} capability.
11788 Also provides conversion of @code{Ada.Calendar.Time} values to and from the
11789 C @code{timeval} format.
11791 @node GNAT.Calendar.Time_IO (g-catiio.ads)
11792 @section @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
11793 @cindex @code{Calendar}
11795 @cindex @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
11797 @node GNAT.CRC32 (g-crc32.ads)
11798 @section @code{GNAT.CRC32} (@file{g-crc32.ads})
11799 @cindex @code{GNAT.CRC32} (@file{g-crc32.ads})
11801 @cindex Cyclic Redundancy Check
11804 This package implements the CRC-32 algorithm. For a full description
11805 of this algorithm see
11806 ``Computation of Cyclic Redundancy Checks via Table Look-Up'',
11807 @cite{Communications of the ACM}, Vol.@: 31 No.@: 8, pp.@: 1008-1013,
11808 Aug.@: 1988. Sarwate, D.V@.
11811 Provides an extended capability for formatted output of time values with
11812 full user control over the format. Modeled on the GNU Date specification.
11814 @node GNAT.Case_Util (g-casuti.ads)
11815 @section @code{GNAT.Case_Util} (@file{g-casuti.ads})
11816 @cindex @code{GNAT.Case_Util} (@file{g-casuti.ads})
11817 @cindex Casing utilities
11818 @cindex Character handling (@code{GNAT.Case_Util})
11821 A set of simple routines for handling upper and lower casing of strings
11822 without the overhead of the full casing tables
11823 in @code{Ada.Characters.Handling}.
11825 @node GNAT.CGI (g-cgi.ads)
11826 @section @code{GNAT.CGI} (@file{g-cgi.ads})
11827 @cindex @code{GNAT.CGI} (@file{g-cgi.ads})
11828 @cindex CGI (Common Gateway Interface)
11831 This is a package for interfacing a GNAT program with a Web server via the
11832 Common Gateway Interface (CGI)@. Basically this package parses the CGI
11833 parameters, which are a set of key/value pairs sent by the Web server. It
11834 builds a table whose index is the key and provides some services to deal
11837 @node GNAT.CGI.Cookie (g-cgicoo.ads)
11838 @section @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
11839 @cindex @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
11840 @cindex CGI (Common Gateway Interface) cookie support
11841 @cindex Cookie support in CGI
11844 This is a package to interface a GNAT program with a Web server via the
11845 Common Gateway Interface (CGI). It exports services to deal with Web
11846 cookies (piece of information kept in the Web client software).
11848 @node GNAT.CGI.Debug (g-cgideb.ads)
11849 @section @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
11850 @cindex @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
11851 @cindex CGI (Common Gateway Interface) debugging
11854 This is a package to help debugging CGI (Common Gateway Interface)
11855 programs written in Ada.
11857 @node GNAT.Command_Line (g-comlin.ads)
11858 @section @code{GNAT.Command_Line} (@file{g-comlin.ads})
11859 @cindex @code{GNAT.Command_Line} (@file{g-comlin.ads})
11860 @cindex Command line
11863 Provides a high level interface to @code{Ada.Command_Line} facilities,
11864 including the ability to scan for named switches with optional parameters
11865 and expand file names using wild card notations.
11867 @node GNAT.Compiler_Version (g-comver.ads)
11868 @section @code{GNAT.Compiler_Version} (@file{g-comver.ads})
11869 @cindex @code{GNAT.Compiler_Version} (@file{g-comver.ads})
11870 @cindex Compiler Version
11871 @cindex Version, of compiler
11874 Provides a routine for obtaining the version of the compiler used to
11875 compile the program. More accurately this is the version of the binder
11876 used to bind the program (this will normally be the same as the version
11877 of the compiler if a consistent tool set is used to compile all units
11880 @node GNAT.Ctrl_C (g-ctrl_c.ads)
11881 @section @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
11882 @cindex @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
11886 Provides a simple interface to handle Ctrl-C keyboard events.
11888 @node GNAT.Current_Exception (g-curexc.ads)
11889 @section @code{GNAT.Current_Exception} (@file{g-curexc.ads})
11890 @cindex @code{GNAT.Current_Exception} (@file{g-curexc.ads})
11891 @cindex Current exception
11892 @cindex Exception retrieval
11895 Provides access to information on the current exception that has been raised
11896 without the need for using the Ada-95 exception choice parameter specification
11897 syntax. This is particularly useful in simulating typical facilities for
11898 obtaining information about exceptions provided by Ada 83 compilers.
11900 @node GNAT.Debug_Pools (g-debpoo.ads)
11901 @section @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
11902 @cindex @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
11904 @cindex Debug pools
11905 @cindex Memory corruption debugging
11908 Provide a debugging storage pools that helps tracking memory corruption
11909 problems. See section ``Finding memory problems with GNAT Debug Pool'' in
11910 the @cite{GNAT User's Guide}.
11912 @node GNAT.Debug_Utilities (g-debuti.ads)
11913 @section @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
11914 @cindex @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
11918 Provides a few useful utilities for debugging purposes, including conversion
11919 to and from string images of address values. Supports both C and Ada formats
11920 for hexadecimal literals.
11922 @node GNAT.Directory_Operations (g-dirope.ads)
11923 @section @code{GNAT.Directory_Operations} (g-dirope.ads)
11924 @cindex @code{GNAT.Directory_Operations} (g-dirope.ads)
11925 @cindex Directory operations
11928 Provides a set of routines for manipulating directories, including changing
11929 the current directory, making new directories, and scanning the files in a
11932 @node GNAT.Dynamic_HTables (g-dynhta.ads)
11933 @section @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
11934 @cindex @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
11935 @cindex Hash tables
11938 A generic implementation of hash tables that can be used to hash arbitrary
11939 data. Provided in two forms, a simple form with built in hash functions,
11940 and a more complex form in which the hash function is supplied.
11943 This package provides a facility similar to that of @code{GNAT.HTable},
11944 except that this package declares a type that can be used to define
11945 dynamic instances of the hash table, while an instantiation of
11946 @code{GNAT.HTable} creates a single instance of the hash table.
11948 @node GNAT.Dynamic_Tables (g-dyntab.ads)
11949 @section @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
11950 @cindex @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
11951 @cindex Table implementation
11952 @cindex Arrays, extendable
11955 A generic package providing a single dimension array abstraction where the
11956 length of the array can be dynamically modified.
11959 This package provides a facility similar to that of @code{GNAT.Table},
11960 except that this package declares a type that can be used to define
11961 dynamic instances of the table, while an instantiation of
11962 @code{GNAT.Table} creates a single instance of the table type.
11964 @node GNAT.Exception_Actions (g-excact.ads)
11965 @section @code{GNAT.Exception_Actions} (@file{g-excact.ads})
11966 @cindex @code{GNAT.Exception_Actions} (@file{g-excact.ads})
11967 @cindex Exception actions
11970 Provides callbacks when an exception is raised. Callbacks can be registered
11971 for specific exceptions, or when any exception is raised. This
11972 can be used for instance to force a core dump to ease debugging.
11974 @node GNAT.Exception_Traces (g-exctra.ads)
11975 @section @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
11976 @cindex @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
11977 @cindex Exception traces
11981 Provides an interface allowing to control automatic output upon exception
11984 @node GNAT.Exceptions (g-except.ads)
11985 @section @code{GNAT.Exceptions} (@file{g-expect.ads})
11986 @cindex @code{GNAT.Exceptions} (@file{g-expect.ads})
11987 @cindex Exceptions, Pure
11988 @cindex Pure packages, exceptions
11991 Normally it is not possible to raise an exception with
11992 a message from a subprogram in a pure package, since the
11993 necessary types and subprograms are in @code{Ada.Exceptions}
11994 which is not a pure unit. @code{GNAT.Exceptions} provides a
11995 facility for getting around this limitation for a few
11996 predefined exceptions, and for example allow raising
11997 @code{Constraint_Error} with a message from a pure subprogram.
11999 @node GNAT.Expect (g-expect.ads)
12000 @section @code{GNAT.Expect} (@file{g-expect.ads})
12001 @cindex @code{GNAT.Expect} (@file{g-expect.ads})
12004 Provides a set of subprograms similar to what is available
12005 with the standard Tcl Expect tool.
12006 It allows you to easily spawn and communicate with an external process.
12007 You can send commands or inputs to the process, and compare the output
12008 with some expected regular expression. Currently @code{GNAT.Expect}
12009 is implemented on all native GNAT ports except for OpenVMS@.
12010 It is not implemented for cross ports, and in particular is not
12011 implemented for VxWorks or LynxOS@.
12013 @node GNAT.Float_Control (g-flocon.ads)
12014 @section @code{GNAT.Float_Control} (@file{g-flocon.ads})
12015 @cindex @code{GNAT.Float_Control} (@file{g-flocon.ads})
12016 @cindex Floating-Point Processor
12019 Provides an interface for resetting the floating-point processor into the
12020 mode required for correct semantic operation in Ada. Some third party
12021 library calls may cause this mode to be modified, and the Reset procedure
12022 in this package can be used to reestablish the required mode.
12024 @node GNAT.Heap_Sort (g-heasor.ads)
12025 @section @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
12026 @cindex @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
12030 Provides a general implementation of heap sort usable for sorting arbitrary
12031 data items. Exchange and comparison procedures are provided by passing
12032 access-to-procedure values. The algorithm used is a modified heap sort
12033 that performs approximately N*log(N) comparisons in the worst case.
12035 @node GNAT.Heap_Sort_A (g-hesora.ads)
12036 @section @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
12037 @cindex @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
12041 Provides a general implementation of heap sort usable for sorting arbitrary
12042 data items. Move and comparison procedures are provided by passing
12043 access-to-procedure values. The algorithm used is a modified heap sort
12044 that performs approximately N*log(N) comparisons in the worst case.
12045 This differs from @code{GNAT.Heap_Sort} in having a less convenient
12046 interface, but may be slightly more efficient.
12048 @node GNAT.Heap_Sort_G (g-hesorg.ads)
12049 @section @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
12050 @cindex @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
12054 Similar to @code{Heap_Sort_A} except that the move and sorting procedures
12055 are provided as generic parameters, this improves efficiency, especially
12056 if the procedures can be inlined, at the expense of duplicating code for
12057 multiple instantiations.
12059 @node GNAT.HTable (g-htable.ads)
12060 @section @code{GNAT.HTable} (@file{g-htable.ads})
12061 @cindex @code{GNAT.HTable} (@file{g-htable.ads})
12062 @cindex Hash tables
12065 A generic implementation of hash tables that can be used to hash arbitrary
12066 data. Provides two approaches, one a simple static approach, and the other
12067 allowing arbitrary dynamic hash tables.
12069 @node GNAT.IO (g-io.ads)
12070 @section @code{GNAT.IO} (@file{g-io.ads})
12071 @cindex @code{GNAT.IO} (@file{g-io.ads})
12073 @cindex Input/Output facilities
12076 A simple preelaborable input-output package that provides a subset of
12077 simple Text_IO functions for reading characters and strings from
12078 Standard_Input, and writing characters, strings and integers to either
12079 Standard_Output or Standard_Error.
12081 @node GNAT.IO_Aux (g-io_aux.ads)
12082 @section @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
12083 @cindex @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
12085 @cindex Input/Output facilities
12087 Provides some auxiliary functions for use with Text_IO, including a test
12088 for whether a file exists, and functions for reading a line of text.
12090 @node GNAT.Lock_Files (g-locfil.ads)
12091 @section @code{GNAT.Lock_Files} (@file{g-locfil.ads})
12092 @cindex @code{GNAT.Lock_Files} (@file{g-locfil.ads})
12093 @cindex File locking
12094 @cindex Locking using files
12097 Provides a general interface for using files as locks. Can be used for
12098 providing program level synchronization.
12100 @node GNAT.MD5 (g-md5.ads)
12101 @section @code{GNAT.MD5} (@file{g-md5.ads})
12102 @cindex @code{GNAT.MD5} (@file{g-md5.ads})
12103 @cindex Message Digest MD5
12106 Implements the MD5 Message-Digest Algorithm as described in RFC 1321.
12108 @node GNAT.Memory_Dump (g-memdum.ads)
12109 @section @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
12110 @cindex @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
12111 @cindex Dump Memory
12114 Provides a convenient routine for dumping raw memory to either the
12115 standard output or standard error files. Uses GNAT.IO for actual
12118 @node GNAT.Most_Recent_Exception (g-moreex.ads)
12119 @section @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
12120 @cindex @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
12121 @cindex Exception, obtaining most recent
12124 Provides access to the most recently raised exception. Can be used for
12125 various logging purposes, including duplicating functionality of some
12126 Ada 83 implementation dependent extensions.
12128 @node GNAT.OS_Lib (g-os_lib.ads)
12129 @section @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
12130 @cindex @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
12131 @cindex Operating System interface
12132 @cindex Spawn capability
12135 Provides a range of target independent operating system interface functions,
12136 including time/date management, file operations, subprocess management,
12137 including a portable spawn procedure, and access to environment variables
12138 and error return codes.
12140 @node GNAT.Perfect_Hash.Generators (g-pehage.ads)
12141 @section @code{GNAT.Perfect_Hash.Generators} (@file{g-pehage.ads})
12142 @cindex @code{GNAT.Perfect_Hash.Generators} (@file{g-pehage.ads})
12143 @cindex Hash functions
12146 Provides a generator of static minimal perfect hash functions. No
12147 collisions occur and each item can be retrieved from the table in one
12148 probe (perfect property). The hash table size corresponds to the exact
12149 size of the key set and no larger (minimal property). The key set has to
12150 be know in advance (static property). The hash functions are also order
12151 preservering. If w2 is inserted after w1 in the generator, their
12152 hashcode are in the same order. These hashing functions are very
12153 convenient for use with realtime applications.
12155 @node GNAT.Regexp (g-regexp.ads)
12156 @section @code{GNAT.Regexp} (@file{g-regexp.ads})
12157 @cindex @code{GNAT.Regexp} (@file{g-regexp.ads})
12158 @cindex Regular expressions
12159 @cindex Pattern matching
12162 A simple implementation of regular expressions, using a subset of regular
12163 expression syntax copied from familiar Unix style utilities. This is the
12164 simples of the three pattern matching packages provided, and is particularly
12165 suitable for ``file globbing'' applications.
12167 @node GNAT.Registry (g-regist.ads)
12168 @section @code{GNAT.Registry} (@file{g-regist.ads})
12169 @cindex @code{GNAT.Registry} (@file{g-regist.ads})
12170 @cindex Windows Registry
12173 This is a high level binding to the Windows registry. It is possible to
12174 do simple things like reading a key value, creating a new key. For full
12175 registry API, but at a lower level of abstraction, refer to the Win32.Winreg
12176 package provided with the Win32Ada binding
12178 @node GNAT.Regpat (g-regpat.ads)
12179 @section @code{GNAT.Regpat} (@file{g-regpat.ads})
12180 @cindex @code{GNAT.Regpat} (@file{g-regpat.ads})
12181 @cindex Regular expressions
12182 @cindex Pattern matching
12185 A complete implementation of Unix-style regular expression matching, copied
12186 from the original V7 style regular expression library written in C by
12187 Henry Spencer (and binary compatible with this C library).
12189 @node GNAT.Secondary_Stack_Info (g-sestin.ads)
12190 @section @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
12191 @cindex @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
12192 @cindex Secondary Stack Info
12195 Provide the capability to query the high water mark of the current task's
12198 @node GNAT.Semaphores (g-semaph.ads)
12199 @section @code{GNAT.Semaphores} (@file{g-semaph.ads})
12200 @cindex @code{GNAT.Semaphores} (@file{g-semaph.ads})
12204 Provides classic counting and binary semaphores using protected types.
12206 @node GNAT.Signals (g-signal.ads)
12207 @section @code{GNAT.Signals} (@file{g-signal.ads})
12208 @cindex @code{GNAT.Signals} (@file{g-signal.ads})
12212 Provides the ability to manipulate the blocked status of signals on supported
12215 @node GNAT.Sockets (g-socket.ads)
12216 @section @code{GNAT.Sockets} (@file{g-socket.ads})
12217 @cindex @code{GNAT.Sockets} (@file{g-socket.ads})
12221 A high level and portable interface to develop sockets based applications.
12222 This package is based on the sockets thin binding found in
12223 @code{GNAT.Sockets.Thin}. Currently @code{GNAT.Sockets} is implemented
12224 on all native GNAT ports except for OpenVMS@. It is not implemented
12225 for the LynxOS@ cross port.
12227 @node GNAT.Source_Info (g-souinf.ads)
12228 @section @code{GNAT.Source_Info} (@file{g-souinf.ads})
12229 @cindex @code{GNAT.Source_Info} (@file{g-souinf.ads})
12230 @cindex Source Information
12233 Provides subprograms that give access to source code information known at
12234 compile time, such as the current file name and line number.
12236 @node GNAT.Spell_Checker (g-speche.ads)
12237 @section @code{GNAT.Spell_Checker} (@file{g-speche.ads})
12238 @cindex @code{GNAT.Spell_Checker} (@file{g-speche.ads})
12239 @cindex Spell checking
12242 Provides a function for determining whether one string is a plausible
12243 near misspelling of another string.
12245 @node GNAT.Spitbol.Patterns (g-spipat.ads)
12246 @section @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
12247 @cindex @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
12248 @cindex SPITBOL pattern matching
12249 @cindex Pattern matching
12252 A complete implementation of SNOBOL4 style pattern matching. This is the
12253 most elaborate of the pattern matching packages provided. It fully duplicates
12254 the SNOBOL4 dynamic pattern construction and matching capabilities, using the
12255 efficient algorithm developed by Robert Dewar for the SPITBOL system.
12257 @node GNAT.Spitbol (g-spitbo.ads)
12258 @section @code{GNAT.Spitbol} (@file{g-spitbo.ads})
12259 @cindex @code{GNAT.Spitbol} (@file{g-spitbo.ads})
12260 @cindex SPITBOL interface
12263 The top level package of the collection of SPITBOL-style functionality, this
12264 package provides basic SNOBOL4 string manipulation functions, such as
12265 Pad, Reverse, Trim, Substr capability, as well as a generic table function
12266 useful for constructing arbitrary mappings from strings in the style of
12267 the SNOBOL4 TABLE function.
12269 @node GNAT.Spitbol.Table_Boolean (g-sptabo.ads)
12270 @section @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
12271 @cindex @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
12272 @cindex Sets of strings
12273 @cindex SPITBOL Tables
12276 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
12277 for type @code{Standard.Boolean}, giving an implementation of sets of
12280 @node GNAT.Spitbol.Table_Integer (g-sptain.ads)
12281 @section @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
12282 @cindex @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
12283 @cindex Integer maps
12285 @cindex SPITBOL Tables
12288 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
12289 for type @code{Standard.Integer}, giving an implementation of maps
12290 from string to integer values.
12292 @node GNAT.Spitbol.Table_VString (g-sptavs.ads)
12293 @section @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
12294 @cindex @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
12295 @cindex String maps
12297 @cindex SPITBOL Tables
12300 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table} for
12301 a variable length string type, giving an implementation of general
12302 maps from strings to strings.
12304 @node GNAT.Strings (g-string.ads)
12305 @section @code{GNAT.Strings} (@file{g-string.ads})
12306 @cindex @code{GNAT.Strings} (@file{g-string.ads})
12309 Common String access types and related subprograms. Basically it
12310 defines a string access and an array of string access types.
12312 @node GNAT.String_Split (g-strspl.ads)
12313 @section @code{GNAT.String_Split} (@file{g-strspl.ads})
12314 @cindex @code{GNAT.String_Split} (@file{g-strspl.ads})
12315 @cindex String splitter
12318 Useful string-manipulation routines: given a set of separators, split
12319 a string wherever the separators appear, and provide direct access
12320 to the resulting slices. This package is instantiated from
12321 @code{GNAT.Array_Split}.
12323 @node GNAT.Table (g-table.ads)
12324 @section @code{GNAT.Table} (@file{g-table.ads})
12325 @cindex @code{GNAT.Table} (@file{g-table.ads})
12326 @cindex Table implementation
12327 @cindex Arrays, extendable
12330 A generic package providing a single dimension array abstraction where the
12331 length of the array can be dynamically modified.
12334 This package provides a facility similar to that of @code{GNAT.Dynamic_Tables},
12335 except that this package declares a single instance of the table type,
12336 while an instantiation of @code{GNAT.Dynamic_Tables} creates a type that can be
12337 used to define dynamic instances of the table.
12339 @node GNAT.Task_Lock (g-tasloc.ads)
12340 @section @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
12341 @cindex @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
12342 @cindex Task synchronization
12343 @cindex Task locking
12347 A very simple facility for locking and unlocking sections of code using a
12348 single global task lock. Appropriate for use in situations where contention
12349 between tasks is very rarely expected.
12351 @node GNAT.Threads (g-thread.ads)
12352 @section @code{GNAT.Threads} (@file{g-thread.ads})
12353 @cindex @code{GNAT.Threads} (@file{g-thread.ads})
12354 @cindex Foreign threads
12355 @cindex Threads, foreign
12358 Provides facilities for creating and destroying threads with explicit calls.
12359 These threads are known to the GNAT run-time system. These subprograms are
12360 exported C-convention procedures intended to be called from foreign code.
12361 By using these primitives rather than directly calling operating systems
12362 routines, compatibility with the Ada tasking runt-time is provided.
12364 @node GNAT.Traceback (g-traceb.ads)
12365 @section @code{GNAT.Traceback} (@file{g-traceb.ads})
12366 @cindex @code{GNAT.Traceback} (@file{g-traceb.ads})
12367 @cindex Trace back facilities
12370 Provides a facility for obtaining non-symbolic traceback information, useful
12371 in various debugging situations.
12373 @node GNAT.Traceback.Symbolic (g-trasym.ads)
12374 @section @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
12375 @cindex @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
12376 @cindex Trace back facilities
12379 Provides symbolic traceback information that includes the subprogram
12380 name and line number information.
12382 @node GNAT.Wide_String_Split (g-wistsp.ads)
12383 @section @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
12384 @cindex @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
12385 @cindex Wide_String splitter
12388 Useful wide_string-manipulation routines: given a set of separators, split
12389 a wide_string wherever the separators appear, and provide direct access
12390 to the resulting slices. This package is instantiated from
12391 @code{GNAT.Array_Split}.
12393 @node Interfaces.C.Extensions (i-cexten.ads)
12394 @section @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
12395 @cindex @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
12398 This package contains additional C-related definitions, intended
12399 for use with either manually or automatically generated bindings
12402 @node Interfaces.C.Streams (i-cstrea.ads)
12403 @section @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
12404 @cindex @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
12405 @cindex C streams, interfacing
12408 This package is a binding for the most commonly used operations
12411 @node Interfaces.CPP (i-cpp.ads)
12412 @section @code{Interfaces.CPP} (@file{i-cpp.ads})
12413 @cindex @code{Interfaces.CPP} (@file{i-cpp.ads})
12414 @cindex C++ interfacing
12415 @cindex Interfacing, to C++
12418 This package provides facilities for use in interfacing to C++. It
12419 is primarily intended to be used in connection with automated tools
12420 for the generation of C++ interfaces.
12422 @node Interfaces.Os2lib (i-os2lib.ads)
12423 @section @code{Interfaces.Os2lib} (@file{i-os2lib.ads})
12424 @cindex @code{Interfaces.Os2lib} (@file{i-os2lib.ads})
12425 @cindex Interfacing, to OS/2
12426 @cindex OS/2 interfacing
12429 This package provides interface definitions to the OS/2 library.
12430 It is a thin binding which is a direct translation of the
12431 various @file{<bse@.h>} files.
12433 @node Interfaces.Os2lib.Errors (i-os2err.ads)
12434 @section @code{Interfaces.Os2lib.Errors} (@file{i-os2err.ads})
12435 @cindex @code{Interfaces.Os2lib.Errors} (@file{i-os2err.ads})
12436 @cindex OS/2 Error codes
12437 @cindex Interfacing, to OS/2
12438 @cindex OS/2 interfacing
12441 This package provides definitions of the OS/2 error codes.
12443 @node Interfaces.Os2lib.Synchronization (i-os2syn.ads)
12444 @section @code{Interfaces.Os2lib.Synchronization} (@file{i-os2syn.ads})
12445 @cindex @code{Interfaces.Os2lib.Synchronization} (@file{i-os2syn.ads})
12446 @cindex Interfacing, to OS/2
12447 @cindex Synchronization, OS/2
12448 @cindex OS/2 synchronization primitives
12451 This is a child package that provides definitions for interfacing
12452 to the @code{OS/2} synchronization primitives.
12454 @node Interfaces.Os2lib.Threads (i-os2thr.ads)
12455 @section @code{Interfaces.Os2lib.Threads} (@file{i-os2thr.ads})
12456 @cindex @code{Interfaces.Os2lib.Threads} (@file{i-os2thr.ads})
12457 @cindex Interfacing, to OS/2
12458 @cindex Thread control, OS/2
12459 @cindex OS/2 thread interfacing
12462 This is a child package that provides definitions for interfacing
12463 to the @code{OS/2} thread primitives.
12465 @node Interfaces.Packed_Decimal (i-pacdec.ads)
12466 @section @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
12467 @cindex @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
12468 @cindex IBM Packed Format
12469 @cindex Packed Decimal
12472 This package provides a set of routines for conversions to and
12473 from a packed decimal format compatible with that used on IBM
12476 @node Interfaces.VxWorks (i-vxwork.ads)
12477 @section @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
12478 @cindex @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
12479 @cindex Interfacing to VxWorks
12480 @cindex VxWorks, interfacing
12483 This package provides a limited binding to the VxWorks API.
12484 In particular, it interfaces with the
12485 VxWorks hardware interrupt facilities.
12487 @node Interfaces.VxWorks.IO (i-vxwoio.ads)
12488 @section @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
12489 @cindex @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
12490 @cindex Interfacing to VxWorks' I/O
12491 @cindex VxWorks, I/O interfacing
12492 @cindex VxWorks, Get_Immediate
12493 @cindex Get_Immediate, VxWorks
12496 This package provides a binding to the ioctl (IO/Control)
12497 function of VxWorks, defining a set of option values and
12498 function codes. A particular use of this package is
12499 to enable the use of Get_Immediate under VxWorks.
12501 @node System.Address_Image (s-addima.ads)
12502 @section @code{System.Address_Image} (@file{s-addima.ads})
12503 @cindex @code{System.Address_Image} (@file{s-addima.ads})
12504 @cindex Address image
12505 @cindex Image, of an address
12508 This function provides a useful debugging
12509 function that gives an (implementation dependent)
12510 string which identifies an address.
12512 @node System.Assertions (s-assert.ads)
12513 @section @code{System.Assertions} (@file{s-assert.ads})
12514 @cindex @code{System.Assertions} (@file{s-assert.ads})
12516 @cindex Assert_Failure, exception
12519 This package provides the declaration of the exception raised
12520 by an run-time assertion failure, as well as the routine that
12521 is used internally to raise this assertion.
12523 @node System.Memory (s-memory.ads)
12524 @section @code{System.Memory} (@file{s-memory.ads})
12525 @cindex @code{System.Memory} (@file{s-memory.ads})
12526 @cindex Memory allocation
12529 This package provides the interface to the low level routines used
12530 by the generated code for allocation and freeing storage for the
12531 default storage pool (analogous to the C routines malloc and free.
12532 It also provides a reallocation interface analogous to the C routine
12533 realloc. The body of this unit may be modified to provide alternative
12534 allocation mechanisms for the default pool, and in addition, direct
12535 calls to this unit may be made for low level allocation uses (for
12536 example see the body of @code{GNAT.Tables}).
12538 @node System.Partition_Interface (s-parint.ads)
12539 @section @code{System.Partition_Interface} (@file{s-parint.ads})
12540 @cindex @code{System.Partition_Interface} (@file{s-parint.ads})
12541 @cindex Partition intefacing functions
12544 This package provides facilities for partition interfacing. It
12545 is used primarily in a distribution context when using Annex E
12548 @node System.Restrictions (s-restri.ads)
12549 @section @code{System.Restrictions} (@file{s-restri.ads})
12550 @cindex @code{System.Restrictions} (@file{s-restri.ads})
12551 @cindex Run-time restrictions access
12554 This package provides facilities for accessing at run-time
12555 the status of restrictions specified at compile time for
12556 the partition. Information is available both with regard
12557 to actual restrictions specified, and with regard to
12558 compiler determined information on which restrictions
12559 are violated by one or more packages in the partition.
12561 @node System.Rident (s-rident.ads)
12562 @section @code{System.Rident} (@file{s-rident.ads})
12563 @cindex @code{System.Rident} (@file{s-rident.ads})
12564 @cindex Restrictions definitions
12567 This package provides definitions of the restrictions
12568 identifiers supported by GNAT, and also the format of
12569 the restrictions provided in package System.Restrictions.
12570 It is not normally necessary to @code{with} this generic package
12571 since the necessary instantiation is included in
12572 package System.Restrictions.
12574 @node System.Task_Info (s-tasinf.ads)
12575 @section @code{System.Task_Info} (@file{s-tasinf.ads})
12576 @cindex @code{System.Task_Info} (@file{s-tasinf.ads})
12577 @cindex Task_Info pragma
12580 This package provides target dependent functionality that is used
12581 to support the @code{Task_Info} pragma
12583 @node System.Wch_Cnv (s-wchcnv.ads)
12584 @section @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
12585 @cindex @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
12586 @cindex Wide Character, Representation
12587 @cindex Wide String, Conversion
12588 @cindex Representation of wide characters
12591 This package provides routines for converting between
12592 wide characters and a representation as a value of type
12593 @code{Standard.String}, using a specified wide character
12594 encoding method. It uses definitions in
12595 package @code{System.Wch_Con}.
12597 @node System.Wch_Con (s-wchcon.ads)
12598 @section @code{System.Wch_Con} (@file{s-wchcon.ads})
12599 @cindex @code{System.Wch_Con} (@file{s-wchcon.ads})
12602 This package provides definitions and descriptions of
12603 the various methods used for encoding wide characters
12604 in ordinary strings. These definitions are used by
12605 the package @code{System.Wch_Cnv}.
12607 @node Interfacing to Other Languages
12608 @chapter Interfacing to Other Languages
12610 The facilities in annex B of the Ada 95 Reference Manual are fully
12611 implemented in GNAT, and in addition, a full interface to C++ is
12615 * Interfacing to C::
12616 * Interfacing to C++::
12617 * Interfacing to COBOL::
12618 * Interfacing to Fortran::
12619 * Interfacing to non-GNAT Ada code::
12622 @node Interfacing to C
12623 @section Interfacing to C
12626 Interfacing to C with GNAT can use one of two approaches:
12630 The types in the package @code{Interfaces.C} may be used.
12632 Standard Ada types may be used directly. This may be less portable to
12633 other compilers, but will work on all GNAT compilers, which guarantee
12634 correspondence between the C and Ada types.
12638 Pragma @code{Convention C} may be applied to Ada types, but mostly has no
12639 effect, since this is the default. The following table shows the
12640 correspondence between Ada scalar types and the corresponding C types.
12645 @item Short_Integer
12647 @item Short_Short_Integer
12651 @item Long_Long_Integer
12659 @item Long_Long_Float
12660 This is the longest floating-point type supported by the hardware.
12664 Additionally, there are the following general correspondences between Ada
12668 Ada enumeration types map to C enumeration types directly if pragma
12669 @code{Convention C} is specified, which causes them to have int
12670 length. Without pragma @code{Convention C}, Ada enumeration types map to
12671 8, 16, or 32 bits (i.e.@: C types @code{signed char}, @code{short},
12672 @code{int}, respectively) depending on the number of values passed.
12673 This is the only case in which pragma @code{Convention C} affects the
12674 representation of an Ada type.
12677 Ada access types map to C pointers, except for the case of pointers to
12678 unconstrained types in Ada, which have no direct C equivalent.
12681 Ada arrays map directly to C arrays.
12684 Ada records map directly to C structures.
12687 Packed Ada records map to C structures where all members are bit fields
12688 of the length corresponding to the @code{@var{type}'Size} value in Ada.
12691 @node Interfacing to C++
12692 @section Interfacing to C++
12695 The interface to C++ makes use of the following pragmas, which are
12696 primarily intended to be constructed automatically using a binding generator
12697 tool, although it is possible to construct them by hand. No suitable binding
12698 generator tool is supplied with GNAT though.
12700 Using these pragmas it is possible to achieve complete
12701 inter-operability between Ada tagged types and C class definitions.
12702 See @ref{Implementation Defined Pragmas}, for more details.
12705 @item pragma CPP_Class ([Entity =>] @var{local_name})
12706 The argument denotes an entity in the current declarative region that is
12707 declared as a tagged or untagged record type. It indicates that the type
12708 corresponds to an externally declared C++ class type, and is to be laid
12709 out the same way that C++ would lay out the type.
12711 @item pragma CPP_Constructor ([Entity =>] @var{local_name})
12712 This pragma identifies an imported function (imported in the usual way
12713 with pragma @code{Import}) as corresponding to a C++ constructor.
12715 @item pragma CPP_Vtable @dots{}
12716 One @code{CPP_Vtable} pragma can be present for each component of type
12717 @code{CPP.Interfaces.Vtable_Ptr} in a record to which pragma @code{CPP_Class}
12721 @node Interfacing to COBOL
12722 @section Interfacing to COBOL
12725 Interfacing to COBOL is achieved as described in section B.4 of
12726 the Ada 95 reference manual.
12728 @node Interfacing to Fortran
12729 @section Interfacing to Fortran
12732 Interfacing to Fortran is achieved as described in section B.5 of the
12733 reference manual. The pragma @code{Convention Fortran}, applied to a
12734 multi-dimensional array causes the array to be stored in column-major
12735 order as required for convenient interface to Fortran.
12737 @node Interfacing to non-GNAT Ada code
12738 @section Interfacing to non-GNAT Ada code
12740 It is possible to specify the convention @code{Ada} in a pragma
12741 @code{Import} or pragma @code{Export}. However this refers to
12742 the calling conventions used by GNAT, which may or may not be
12743 similar enough to those used by some other Ada 83 or Ada 95
12744 compiler to allow interoperation.
12746 If arguments types are kept simple, and if the foreign compiler generally
12747 follows system calling conventions, then it may be possible to integrate
12748 files compiled by other Ada compilers, provided that the elaboration
12749 issues are adequately addressed (for example by eliminating the
12750 need for any load time elaboration).
12752 In particular, GNAT running on VMS is designed to
12753 be highly compatible with the DEC Ada 83 compiler, so this is one
12754 case in which it is possible to import foreign units of this type,
12755 provided that the data items passed are restricted to simple scalar
12756 values or simple record types without variants, or simple array
12757 types with fixed bounds.
12759 @node Specialized Needs Annexes
12760 @chapter Specialized Needs Annexes
12763 Ada 95 defines a number of specialized needs annexes, which are not
12764 required in all implementations. However, as described in this chapter,
12765 GNAT implements all of these special needs annexes:
12768 @item Systems Programming (Annex C)
12769 The Systems Programming Annex is fully implemented.
12771 @item Real-Time Systems (Annex D)
12772 The Real-Time Systems Annex is fully implemented.
12774 @item Distributed Systems (Annex E)
12775 Stub generation is fully implemented in the GNAT compiler. In addition,
12776 a complete compatible PCS is available as part of the GLADE system,
12777 a separate product. When the two
12778 products are used in conjunction, this annex is fully implemented.
12780 @item Information Systems (Annex F)
12781 The Information Systems annex is fully implemented.
12783 @item Numerics (Annex G)
12784 The Numerics Annex is fully implemented.
12786 @item Safety and Security (Annex H)
12787 The Safety and Security annex is fully implemented.
12790 @node Implementation of Specific Ada Features
12791 @chapter Implementation of Specific Ada Features
12794 This chapter describes the GNAT implementation of several Ada language
12798 * Machine Code Insertions::
12799 * GNAT Implementation of Tasking::
12800 * GNAT Implementation of Shared Passive Packages::
12801 * Code Generation for Array Aggregates::
12802 * The Size of Discriminated Records with Default Discriminants::
12805 @node Machine Code Insertions
12806 @section Machine Code Insertions
12809 Package @code{Machine_Code} provides machine code support as described
12810 in the Ada 95 Reference Manual in two separate forms:
12813 Machine code statements, consisting of qualified expressions that
12814 fit the requirements of RM section 13.8.
12816 An intrinsic callable procedure, providing an alternative mechanism of
12817 including machine instructions in a subprogram.
12821 The two features are similar, and both are closely related to the mechanism
12822 provided by the asm instruction in the GNU C compiler. Full understanding
12823 and use of the facilities in this package requires understanding the asm
12824 instruction as described in @cite{Using the GNU Compiler Collection (GCC)}
12825 by Richard Stallman. The relevant section is titled ``Extensions to the C
12826 Language Family'' -> ``Assembler Instructions with C Expression Operands''.
12828 Calls to the function @code{Asm} and the procedure @code{Asm} have identical
12829 semantic restrictions and effects as described below. Both are provided so
12830 that the procedure call can be used as a statement, and the function call
12831 can be used to form a code_statement.
12833 The first example given in the GCC documentation is the C @code{asm}
12836 asm ("fsinx %1 %0" : "=f" (result) : "f" (angle));
12840 The equivalent can be written for GNAT as:
12842 @smallexample @c ada
12843 Asm ("fsinx %1 %0",
12844 My_Float'Asm_Output ("=f", result),
12845 My_Float'Asm_Input ("f", angle));
12849 The first argument to @code{Asm} is the assembler template, and is
12850 identical to what is used in GNU C@. This string must be a static
12851 expression. The second argument is the output operand list. It is
12852 either a single @code{Asm_Output} attribute reference, or a list of such
12853 references enclosed in parentheses (technically an array aggregate of
12856 The @code{Asm_Output} attribute denotes a function that takes two
12857 parameters. The first is a string, the second is the name of a variable
12858 of the type designated by the attribute prefix. The first (string)
12859 argument is required to be a static expression and designates the
12860 constraint for the parameter (e.g.@: what kind of register is
12861 required). The second argument is the variable to be updated with the
12862 result. The possible values for constraint are the same as those used in
12863 the RTL, and are dependent on the configuration file used to build the
12864 GCC back end. If there are no output operands, then this argument may
12865 either be omitted, or explicitly given as @code{No_Output_Operands}.
12867 The second argument of @code{@var{my_float}'Asm_Output} functions as
12868 though it were an @code{out} parameter, which is a little curious, but
12869 all names have the form of expressions, so there is no syntactic
12870 irregularity, even though normally functions would not be permitted
12871 @code{out} parameters. The third argument is the list of input
12872 operands. It is either a single @code{Asm_Input} attribute reference, or
12873 a list of such references enclosed in parentheses (technically an array
12874 aggregate of such references).
12876 The @code{Asm_Input} attribute denotes a function that takes two
12877 parameters. The first is a string, the second is an expression of the
12878 type designated by the prefix. The first (string) argument is required
12879 to be a static expression, and is the constraint for the parameter,
12880 (e.g.@: what kind of register is required). The second argument is the
12881 value to be used as the input argument. The possible values for the
12882 constant are the same as those used in the RTL, and are dependent on
12883 the configuration file used to built the GCC back end.
12885 If there are no input operands, this argument may either be omitted, or
12886 explicitly given as @code{No_Input_Operands}. The fourth argument, not
12887 present in the above example, is a list of register names, called the
12888 @dfn{clobber} argument. This argument, if given, must be a static string
12889 expression, and is a space or comma separated list of names of registers
12890 that must be considered destroyed as a result of the @code{Asm} call. If
12891 this argument is the null string (the default value), then the code
12892 generator assumes that no additional registers are destroyed.
12894 The fifth argument, not present in the above example, called the
12895 @dfn{volatile} argument, is by default @code{False}. It can be set to
12896 the literal value @code{True} to indicate to the code generator that all
12897 optimizations with respect to the instruction specified should be
12898 suppressed, and that in particular, for an instruction that has outputs,
12899 the instruction will still be generated, even if none of the outputs are
12900 used. See the full description in the GCC manual for further details.
12902 The @code{Asm} subprograms may be used in two ways. First the procedure
12903 forms can be used anywhere a procedure call would be valid, and
12904 correspond to what the RM calls ``intrinsic'' routines. Such calls can
12905 be used to intersperse machine instructions with other Ada statements.
12906 Second, the function forms, which return a dummy value of the limited
12907 private type @code{Asm_Insn}, can be used in code statements, and indeed
12908 this is the only context where such calls are allowed. Code statements
12909 appear as aggregates of the form:
12911 @smallexample @c ada
12912 Asm_Insn'(Asm (@dots{}));
12913 Asm_Insn'(Asm_Volatile (@dots{}));
12917 In accordance with RM rules, such code statements are allowed only
12918 within subprograms whose entire body consists of such statements. It is
12919 not permissible to intermix such statements with other Ada statements.
12921 Typically the form using intrinsic procedure calls is more convenient
12922 and more flexible. The code statement form is provided to meet the RM
12923 suggestion that such a facility should be made available. The following
12924 is the exact syntax of the call to @code{Asm}. As usual, if named notation
12925 is used, the arguments may be given in arbitrary order, following the
12926 normal rules for use of positional and named arguments)
12930 [Template =>] static_string_EXPRESSION
12931 [,[Outputs =>] OUTPUT_OPERAND_LIST ]
12932 [,[Inputs =>] INPUT_OPERAND_LIST ]
12933 [,[Clobber =>] static_string_EXPRESSION ]
12934 [,[Volatile =>] static_boolean_EXPRESSION] )
12936 OUTPUT_OPERAND_LIST ::=
12937 [PREFIX.]No_Output_Operands
12938 | OUTPUT_OPERAND_ATTRIBUTE
12939 | (OUTPUT_OPERAND_ATTRIBUTE @{,OUTPUT_OPERAND_ATTRIBUTE@})
12941 OUTPUT_OPERAND_ATTRIBUTE ::=
12942 SUBTYPE_MARK'Asm_Output (static_string_EXPRESSION, NAME)
12944 INPUT_OPERAND_LIST ::=
12945 [PREFIX.]No_Input_Operands
12946 | INPUT_OPERAND_ATTRIBUTE
12947 | (INPUT_OPERAND_ATTRIBUTE @{,INPUT_OPERAND_ATTRIBUTE@})
12949 INPUT_OPERAND_ATTRIBUTE ::=
12950 SUBTYPE_MARK'Asm_Input (static_string_EXPRESSION, EXPRESSION)
12954 The identifiers @code{No_Input_Operands} and @code{No_Output_Operands}
12955 are declared in the package @code{Machine_Code} and must be referenced
12956 according to normal visibility rules. In particular if there is no
12957 @code{use} clause for this package, then appropriate package name
12958 qualification is required.
12960 @node GNAT Implementation of Tasking
12961 @section GNAT Implementation of Tasking
12964 This chapter outlines the basic GNAT approach to tasking (in particular,
12965 a multi-layered library for portability) and discusses issues related
12966 to compliance with the Real-Time Systems Annex.
12969 * Mapping Ada Tasks onto the Underlying Kernel Threads::
12970 * Ensuring Compliance with the Real-Time Annex::
12973 @node Mapping Ada Tasks onto the Underlying Kernel Threads
12974 @subsection Mapping Ada Tasks onto the Underlying Kernel Threads
12977 GNAT's run-time support comprises two layers:
12980 @item GNARL (GNAT Run-time Layer)
12981 @item GNULL (GNAT Low-level Library)
12985 In GNAT, Ada's tasking services rely on a platform and OS independent
12986 layer known as GNARL@. This code is responsible for implementing the
12987 correct semantics of Ada's task creation, rendezvous, protected
12990 GNARL decomposes Ada's tasking semantics into simpler lower level
12991 operations such as create a thread, set the priority of a thread,
12992 yield, create a lock, lock/unlock, etc. The spec for these low-level
12993 operations constitutes GNULLI, the GNULL Interface. This interface is
12994 directly inspired from the POSIX real-time API@.
12996 If the underlying executive or OS implements the POSIX standard
12997 faithfully, the GNULL Interface maps as is to the services offered by
12998 the underlying kernel. Otherwise, some target dependent glue code maps
12999 the services offered by the underlying kernel to the semantics expected
13002 Whatever the underlying OS (VxWorks, UNIX, OS/2, Windows NT, etc.) the
13003 key point is that each Ada task is mapped on a thread in the underlying
13004 kernel. For example, in the case of VxWorks, one Ada task = one VxWorks task.
13006 In addition Ada task priorities map onto the underlying thread priorities.
13007 Mapping Ada tasks onto the underlying kernel threads has several advantages:
13011 The underlying scheduler is used to schedule the Ada tasks. This
13012 makes Ada tasks as efficient as kernel threads from a scheduling
13016 Interaction with code written in C containing threads is eased
13017 since at the lowest level Ada tasks and C threads map onto the same
13018 underlying kernel concept.
13021 When an Ada task is blocked during I/O the remaining Ada tasks are
13025 On multiprocessor systems Ada tasks can execute in parallel.
13029 Some threads libraries offer a mechanism to fork a new process, with the
13030 child process duplicating the threads from the parent.
13032 support this functionality when the parent contains more than one task.
13033 @cindex Forking a new process
13035 @node Ensuring Compliance with the Real-Time Annex
13036 @subsection Ensuring Compliance with the Real-Time Annex
13037 @cindex Real-Time Systems Annex compliance
13040 Although mapping Ada tasks onto
13041 the underlying threads has significant advantages, it does create some
13042 complications when it comes to respecting the scheduling semantics
13043 specified in the real-time annex (Annex D).
13045 For instance the Annex D requirement for the @code{FIFO_Within_Priorities}
13046 scheduling policy states:
13049 @emph{When the active priority of a ready task that is not running
13050 changes, or the setting of its base priority takes effect, the
13051 task is removed from the ready queue for its old active priority
13052 and is added at the tail of the ready queue for its new active
13053 priority, except in the case where the active priority is lowered
13054 due to the loss of inherited priority, in which case the task is
13055 added at the head of the ready queue for its new active priority.}
13059 While most kernels do put tasks at the end of the priority queue when
13060 a task changes its priority, (which respects the main
13061 FIFO_Within_Priorities requirement), almost none keep a thread at the
13062 beginning of its priority queue when its priority drops from the loss
13063 of inherited priority.
13065 As a result most vendors have provided incomplete Annex D implementations.
13067 The GNAT run-time, has a nice cooperative solution to this problem
13068 which ensures that accurate FIFO_Within_Priorities semantics are
13071 The principle is as follows. When an Ada task T is about to start
13072 running, it checks whether some other Ada task R with the same
13073 priority as T has been suspended due to the loss of priority
13074 inheritance. If this is the case, T yields and is placed at the end of
13075 its priority queue. When R arrives at the front of the queue it
13078 Note that this simple scheme preserves the relative order of the tasks
13079 that were ready to execute in the priority queue where R has been
13082 @node GNAT Implementation of Shared Passive Packages
13083 @section GNAT Implementation of Shared Passive Packages
13084 @cindex Shared passive packages
13087 GNAT fully implements the pragma @code{Shared_Passive} for
13088 @cindex pragma @code{Shared_Passive}
13089 the purpose of designating shared passive packages.
13090 This allows the use of passive partitions in the
13091 context described in the Ada Reference Manual; i.e. for communication
13092 between separate partitions of a distributed application using the
13093 features in Annex E.
13095 @cindex Distribution Systems Annex
13097 However, the implementation approach used by GNAT provides for more
13098 extensive usage as follows:
13101 @item Communication between separate programs
13103 This allows separate programs to access the data in passive
13104 partitions, using protected objects for synchronization where
13105 needed. The only requirement is that the two programs have a
13106 common shared file system. It is even possible for programs
13107 running on different machines with different architectures
13108 (e.g. different endianness) to communicate via the data in
13109 a passive partition.
13111 @item Persistence between program runs
13113 The data in a passive package can persist from one run of a
13114 program to another, so that a later program sees the final
13115 values stored by a previous run of the same program.
13120 The implementation approach used is to store the data in files. A
13121 separate stream file is created for each object in the package, and
13122 an access to an object causes the corresponding file to be read or
13125 The environment variable @code{SHARED_MEMORY_DIRECTORY} should be
13126 @cindex @code{SHARED_MEMORY_DIRECTORY} environment variable
13127 set to the directory to be used for these files.
13128 The files in this directory
13129 have names that correspond to their fully qualified names. For
13130 example, if we have the package
13132 @smallexample @c ada
13134 pragma Shared_Passive (X);
13141 and the environment variable is set to @code{/stemp/}, then the files created
13142 will have the names:
13150 These files are created when a value is initially written to the object, and
13151 the files are retained until manually deleted. This provides the persistence
13152 semantics. If no file exists, it means that no partition has assigned a value
13153 to the variable; in this case the initial value declared in the package
13154 will be used. This model ensures that there are no issues in synchronizing
13155 the elaboration process, since elaboration of passive packages elaborates the
13156 initial values, but does not create the files.
13158 The files are written using normal @code{Stream_IO} access.
13159 If you want to be able
13160 to communicate between programs or partitions running on different
13161 architectures, then you should use the XDR versions of the stream attribute
13162 routines, since these are architecture independent.
13164 If active synchronization is required for access to the variables in the
13165 shared passive package, then as described in the Ada Reference Manual, the
13166 package may contain protected objects used for this purpose. In this case
13167 a lock file (whose name is @file{___lock} (three underscores)
13168 is created in the shared memory directory.
13169 @cindex @file{___lock} file (for shared passive packages)
13170 This is used to provide the required locking
13171 semantics for proper protected object synchronization.
13173 As of January 2003, GNAT supports shared passive packages on all platforms
13174 except for OpenVMS.
13176 @node Code Generation for Array Aggregates
13177 @section Code Generation for Array Aggregates
13180 * Static constant aggregates with static bounds::
13181 * Constant aggregates with an unconstrained nominal types::
13182 * Aggregates with static bounds::
13183 * Aggregates with non-static bounds::
13184 * Aggregates in assignment statements::
13188 Aggregate have a rich syntax and allow the user to specify the values of
13189 complex data structures by means of a single construct. As a result, the
13190 code generated for aggregates can be quite complex and involve loops, case
13191 statements and multiple assignments. In the simplest cases, however, the
13192 compiler will recognize aggregates whose components and constraints are
13193 fully static, and in those cases the compiler will generate little or no
13194 executable code. The following is an outline of the code that GNAT generates
13195 for various aggregate constructs. For further details, the user will find it
13196 useful to examine the output produced by the -gnatG flag to see the expanded
13197 source that is input to the code generator. The user will also want to examine
13198 the assembly code generated at various levels of optimization.
13200 The code generated for aggregates depends on the context, the component values,
13201 and the type. In the context of an object declaration the code generated is
13202 generally simpler than in the case of an assignment. As a general rule, static
13203 component values and static subtypes also lead to simpler code.
13205 @node Static constant aggregates with static bounds
13206 @subsection Static constant aggregates with static bounds
13209 For the declarations:
13210 @smallexample @c ada
13211 type One_Dim is array (1..10) of integer;
13212 ar0 : constant One_Dim := ( 1, 2, 3, 4, 5, 6, 7, 8, 9, 0);
13216 GNAT generates no executable code: the constant ar0 is placed in static memory.
13217 The same is true for constant aggregates with named associations:
13219 @smallexample @c ada
13220 Cr1 : constant One_Dim := (4 => 16, 2 => 4, 3 => 9, 1=> 1);
13221 Cr3 : constant One_Dim := (others => 7777);
13225 The same is true for multidimensional constant arrays such as:
13227 @smallexample @c ada
13228 type two_dim is array (1..3, 1..3) of integer;
13229 Unit : constant two_dim := ( (1,0,0), (0,1,0), (0,0,1));
13233 The same is true for arrays of one-dimensional arrays: the following are
13236 @smallexample @c ada
13237 type ar1b is array (1..3) of boolean;
13238 type ar_ar is array (1..3) of ar1b;
13239 None : constant ar1b := (others => false); -- fully static
13240 None2 : constant ar_ar := (1..3 => None); -- fully static
13244 However, for multidimensional aggregates with named associations, GNAT will
13245 generate assignments and loops, even if all associations are static. The
13246 following two declarations generate a loop for the first dimension, and
13247 individual component assignments for the second dimension:
13249 @smallexample @c ada
13250 Zero1: constant two_dim := (1..3 => (1..3 => 0));
13251 Zero2: constant two_dim := (others => (others => 0));
13254 @node Constant aggregates with an unconstrained nominal types
13255 @subsection Constant aggregates with an unconstrained nominal types
13258 In such cases the aggregate itself establishes the subtype, so that
13259 associations with @code{others} cannot be used. GNAT determines the
13260 bounds for the actual subtype of the aggregate, and allocates the
13261 aggregate statically as well. No code is generated for the following:
13263 @smallexample @c ada
13264 type One_Unc is array (natural range <>) of integer;
13265 Cr_Unc : constant One_Unc := (12,24,36);
13268 @node Aggregates with static bounds
13269 @subsection Aggregates with static bounds
13272 In all previous examples the aggregate was the initial (and immutable) value
13273 of a constant. If the aggregate initializes a variable, then code is generated
13274 for it as a combination of individual assignments and loops over the target
13275 object. The declarations
13277 @smallexample @c ada
13278 Cr_Var1 : One_Dim := (2, 5, 7, 11);
13279 Cr_Var2 : One_Dim := (others > -1);
13283 generate the equivalent of
13285 @smallexample @c ada
13291 for I in Cr_Var2'range loop
13292 Cr_Var2 (I) := =-1;
13296 @node Aggregates with non-static bounds
13297 @subsection Aggregates with non-static bounds
13300 If the bounds of the aggregate are not statically compatible with the bounds
13301 of the nominal subtype of the target, then constraint checks have to be
13302 generated on the bounds. For a multidimensional array, constraint checks may
13303 have to be applied to sub-arrays individually, if they do not have statically
13304 compatible subtypes.
13306 @node Aggregates in assignment statements
13307 @subsection Aggregates in assignment statements
13310 In general, aggregate assignment requires the construction of a temporary,
13311 and a copy from the temporary to the target of the assignment. This is because
13312 it is not always possible to convert the assignment into a series of individual
13313 component assignments. For example, consider the simple case:
13315 @smallexample @c ada
13320 This cannot be converted into:
13322 @smallexample @c ada
13328 So the aggregate has to be built first in a separate location, and then
13329 copied into the target. GNAT recognizes simple cases where this intermediate
13330 step is not required, and the assignments can be performed in place, directly
13331 into the target. The following sufficient criteria are applied:
13335 The bounds of the aggregate are static, and the associations are static.
13337 The components of the aggregate are static constants, names of
13338 simple variables that are not renamings, or expressions not involving
13339 indexed components whose operands obey these rules.
13343 If any of these conditions are violated, the aggregate will be built in
13344 a temporary (created either by the front-end or the code generator) and then
13345 that temporary will be copied onto the target.
13348 @node The Size of Discriminated Records with Default Discriminants
13349 @section The Size of Discriminated Records with Default Discriminants
13352 If a discriminated type @code{T} has discriminants with default values, it is
13353 possible to declare an object of this type without providing an explicit
13356 @smallexample @c ada
13358 type Size is range 1..100;
13360 type Rec (D : Size := 15) is record
13361 Name : String (1..D);
13369 Such an object is said to be @emph{unconstrained}.
13370 The discriminant of the object
13371 can be modified by a full assignment to the object, as long as it preserves the
13372 relation between the value of the discriminant, and the value of the components
13375 @smallexample @c ada
13377 Word := (3, "yes");
13379 Word := (5, "maybe");
13381 Word := (5, "no"); -- raises Constraint_Error
13386 In order to support this behavior efficiently, an unconstrained object is
13387 given the maximum size that any value of the type requires. In the case
13388 above, @code{Word} has storage for the discriminant and for
13389 a @code{String} of length 100.
13390 It is important to note that unconstrained objects do not require dynamic
13391 allocation. It would be an improper implementation to place on the heap those
13392 components whose size depends on discriminants. (This improper implementation
13393 was used by some Ada83 compilers, where the @code{Name} component above
13395 been stored as a pointer to a dynamic string). Following the principle that
13396 dynamic storage management should never be introduced implicitly,
13397 an Ada95 compiler should reserve the full size for an unconstrained declared
13398 object, and place it on the stack.
13400 This maximum size approach
13401 has been a source of surprise to some users, who expect the default
13402 values of the discriminants to determine the size reserved for an
13403 unconstrained object: ``If the default is 15, why should the object occupy
13405 The answer, of course, is that the discriminant may be later modified,
13406 and its full range of values must be taken into account. This is why the
13411 type Rec (D : Positive := 15) is record
13412 Name : String (1..D);
13420 is flagged by the compiler with a warning:
13421 an attempt to create @code{Too_Large} will raise @code{Storage_Error},
13422 because the required size includes @code{Positive'Last}
13423 bytes. As the first example indicates, the proper approach is to declare an
13424 index type of ``reasonable'' range so that unconstrained objects are not too
13427 One final wrinkle: if the object is declared to be @code{aliased}, or if it is
13428 created in the heap by means of an allocator, then it is @emph{not}
13430 it is constrained by the default values of the discriminants, and those values
13431 cannot be modified by full assignment. This is because in the presence of
13432 aliasing all views of the object (which may be manipulated by different tasks,
13433 say) must be consistent, so it is imperative that the object, once created,
13439 @node Project File Reference
13440 @chapter Project File Reference
13443 This chapter describes the syntax and semantics of project files.
13444 Project files specify the options to be used when building a system.
13445 Project files can specify global settings for all tools,
13446 as well as tool-specific settings.
13447 See the chapter on project files in the GNAT Users guide for examples of use.
13451 * Lexical Elements::
13453 * Empty declarations::
13454 * Typed string declarations::
13458 * Project Attributes::
13459 * Attribute References::
13460 * External Values::
13461 * Case Construction::
13463 * Package Renamings::
13465 * Project Extensions::
13466 * Project File Elaboration::
13469 @node Reserved Words
13470 @section Reserved Words
13473 All Ada95 reserved words are reserved in project files, and cannot be used
13474 as variable names or project names. In addition, the following are
13475 also reserved in project files:
13478 @item @code{extends}
13480 @item @code{external}
13482 @item @code{project}
13486 @node Lexical Elements
13487 @section Lexical Elements
13490 Rules for identifiers are the same as in Ada95. Identifiers
13491 are case-insensitive. Strings are case sensitive, except where noted.
13492 Comments have the same form as in Ada95.
13502 simple_name @{. simple_name@}
13506 @section Declarations
13509 Declarations introduce new entities that denote types, variables, attributes,
13510 and packages. Some declarations can only appear immediately within a project
13511 declaration. Others can appear within a project or within a package.
13515 declarative_item ::=
13516 simple_declarative_item |
13517 typed_string_declaration |
13518 package_declaration
13520 simple_declarative_item ::=
13521 variable_declaration |
13522 typed_variable_declaration |
13523 attribute_declaration |
13524 case_construction |
13528 @node Empty declarations
13529 @section Empty declarations
13532 empty_declaration ::=
13536 An empty declaration is allowed anywhere a declaration is allowed.
13539 @node Typed string declarations
13540 @section Typed string declarations
13543 Typed strings are sequences of string literals. Typed strings are the only
13544 named types in project files. They are used in case constructions, where they
13545 provide support for conditional attribute definitions.
13549 typed_string_declaration ::=
13550 @b{type} <typed_string_>_simple_name @b{is}
13551 ( string_literal @{, string_literal@} );
13555 A typed string declaration can only appear immediately within a project
13558 All the string literals in a typed string declaration must be distinct.
13564 Variables denote values, and appear as constituents of expressions.
13567 typed_variable_declaration ::=
13568 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
13570 variable_declaration ::=
13571 <variable_>simple_name := expression;
13575 The elaboration of a variable declaration introduces the variable and
13576 assigns to it the value of the expression. The name of the variable is
13577 available after the assignment symbol.
13580 A typed_variable can only be declare once.
13583 a non typed variable can be declared multiple times.
13586 Before the completion of its first declaration, the value of variable
13587 is the null string.
13590 @section Expressions
13593 An expression is a formula that defines a computation or retrieval of a value.
13594 In a project file the value of an expression is either a string or a list
13595 of strings. A string value in an expression is either a literal, the current
13596 value of a variable, an external value, an attribute reference, or a
13597 concatenation operation.
13610 attribute_reference
13616 ( <string_>expression @{ , <string_>expression @} )
13619 @subsection Concatenation
13621 The following concatenation functions are defined:
13623 @smallexample @c ada
13624 function "&" (X : String; Y : String) return String;
13625 function "&" (X : String_List; Y : String) return String_List;
13626 function "&" (X : String_List; Y : String_List) return String_List;
13630 @section Attributes
13633 An attribute declaration defines a property of a project or package. This
13634 property can later be queried by means of an attribute reference.
13635 Attribute values are strings or string lists.
13637 Some attributes are associative arrays. These attributes are mappings whose
13638 domain is a set of strings. These attributes are declared one association
13639 at a time, by specifying a point in the domain and the corresponding image
13640 of the attribute. They may also be declared as a full associative array,
13641 getting the same associations as the corresponding attribute in an imported
13642 or extended project.
13644 Attributes that are not associative arrays are called simple attributes.
13648 attribute_declaration ::=
13649 full_associative_array_declaration |
13650 @b{for} attribute_designator @b{use} expression ;
13652 full_associative_array_declaration ::=
13653 @b{for} <associative_array_attribute_>simple_name @b{use}
13654 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
13656 attribute_designator ::=
13657 <simple_attribute_>simple_name |
13658 <associative_array_attribute_>simple_name ( string_literal )
13662 Some attributes are project-specific, and can only appear immediately within
13663 a project declaration. Others are package-specific, and can only appear within
13664 the proper package.
13666 The expression in an attribute definition must be a string or a string_list.
13667 The string literal appearing in the attribute_designator of an associative
13668 array attribute is case-insensitive.
13670 @node Project Attributes
13671 @section Project Attributes
13674 The following attributes apply to a project. All of them are simple
13679 Expression must be a path name. The attribute defines the
13680 directory in which the object files created by the build are to be placed. If
13681 not specified, object files are placed in the project directory.
13684 Expression must be a path name. The attribute defines the
13685 directory in which the executables created by the build are to be placed.
13686 If not specified, executables are placed in the object directory.
13689 Expression must be a list of path names. The attribute
13690 defines the directories in which the source files for the project are to be
13691 found. If not specified, source files are found in the project directory.
13694 Expression must be a list of file names. The attribute
13695 defines the individual files, in the project directory, which are to be used
13696 as sources for the project. File names are path_names that contain no directory
13697 information. If the project has no sources the attribute must be declared
13698 explicitly with an empty list.
13700 @item Source_List_File
13701 Expression must a single path name. The attribute
13702 defines a text file that contains a list of source file names to be used
13703 as sources for the project
13706 Expression must be a path name. The attribute defines the
13707 directory in which a library is to be built. The directory must exist, must
13708 be distinct from the project's object directory, and must be writable.
13711 Expression must be a string that is a legal file name,
13712 without extension. The attribute defines a string that is used to generate
13713 the name of the library to be built by the project.
13716 Argument must be a string value that must be one of the
13717 following @code{"static"}, @code{"dynamic"} or @code{"relocatable"}. This
13718 string is case-insensitive. If this attribute is not specified, the library is
13719 a static library. Otherwise, the library may be dynamic or relocatable. This
13720 distinction is operating-system dependent.
13722 @item Library_Version
13723 Expression must be a string value whose interpretation
13724 is platform dependent. On UNIX, it is used only for dynamic/relocatable
13725 libraries as the internal name of the library (the @code{"soname"}). If the
13726 library file name (built from the @code{Library_Name}) is different from the
13727 @code{Library_Version}, then the library file will be a symbolic link to the
13728 actual file whose name will be @code{Library_Version}.
13730 @item Library_Interface
13731 Expression must be a string list. Each element of the string list
13732 must designate a unit of the project.
13733 If this attribute is present in a Library Project File, then the project
13734 file is a Stand-alone Library_Project_File.
13736 @item Library_Auto_Init
13737 Expression must be a single string "true" or "false", case-insensitive.
13738 If this attribute is present in a Stand-alone Library Project File,
13739 it indicates if initialization is automatic when the dynamic library
13742 @item Library_Options
13743 Expression must be a string list. Indicates additional switches that
13744 are to be used when building a shared library.
13747 Expression must be a single string. Designates an alternative to "gcc"
13748 for building shared libraries.
13750 @item Library_Src_Dir
13751 Expression must be a path name. The attribute defines the
13752 directory in which the sources of the interfaces of a Stand-alone Library will
13753 be copied. The directory must exist, must be distinct from the project's
13754 object directory and source directories, and must be writable.
13757 Expression must be a list of strings that are legal file names.
13758 These file names designate existing compilation units in the source directory
13759 that are legal main subprograms.
13761 When a project file is elaborated, as part of the execution of a gnatmake
13762 command, one or several executables are built and placed in the Exec_Dir.
13763 If the gnatmake command does not include explicit file names, the executables
13764 that are built correspond to the files specified by this attribute.
13766 @item Main_Language
13767 This is a simple attribute. Its value is a string that specifies the
13768 language of the main program.
13771 Expression must be a string list. Each string designates
13772 a programming language that is known to GNAT. The strings are case-insensitive.
13774 @item Locally_Removed_Files
13775 This attribute is legal only in a project file that extends another.
13776 Expression must be a list of strings that are legal file names.
13777 Each file name must designate a source that would normally be inherited
13778 by the current project file. It cannot designate an immediate source that is
13779 not inherited. Each of the source files in the list are not considered to
13780 be sources of the project file: they are not inherited.
13783 @node Attribute References
13784 @section Attribute References
13787 Attribute references are used to retrieve the value of previously defined
13788 attribute for a package or project.
13791 attribute_reference ::=
13792 attribute_prefix ' <simple_attribute_>simple_name [ ( string_literal ) ]
13794 attribute_prefix ::=
13796 <project_simple_name | package_identifier |
13797 <project_>simple_name . package_identifier
13801 If an attribute has not been specified for a given package or project, its
13802 value is the null string or the empty list.
13804 @node External Values
13805 @section External Values
13808 An external value is an expression whose value is obtained from the command
13809 that invoked the processing of the current project file (typically a
13815 @b{external} ( string_literal [, string_literal] )
13819 The first string_literal is the string to be used on the command line or
13820 in the environment to specify the external value. The second string_literal,
13821 if present, is the default to use if there is no specification for this
13822 external value either on the command line or in the environment.
13824 @node Case Construction
13825 @section Case Construction
13828 A case construction supports attribute declarations that depend on the value of
13829 a previously declared variable.
13833 case_construction ::=
13834 @b{case} <typed_variable_>name @b{is}
13839 @b{when} discrete_choice_list =>
13840 @{case_construction | attribute_declaration | empty_declaration@}
13842 discrete_choice_list ::=
13843 string_literal @{| string_literal@} |
13848 All choices in a choice list must be distinct. The choice lists of two
13849 distinct alternatives must be disjoint. Unlike Ada, the choice lists of all
13850 alternatives do not need to include all values of the type. An @code{others}
13851 choice must appear last in the list of alternatives.
13857 A package provides a grouping of variable declarations and attribute
13858 declarations to be used when invoking various GNAT tools. The name of
13859 the package indicates the tool(s) to which it applies.
13863 package_declaration ::=
13864 package_specification | package_renaming
13866 package_specification ::=
13867 @b{package} package_identifier @b{is}
13868 @{simple_declarative_item@}
13869 @b{end} package_identifier ;
13871 package_identifier ::=
13872 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
13873 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
13874 @code{gnatls} | @code{IDE} | @code{Pretty_Printer}
13877 @subsection Package Naming
13880 The attributes of a @code{Naming} package specifies the naming conventions
13881 that apply to the source files in a project. When invoking other GNAT tools,
13882 they will use the sources in the source directories that satisfy these
13883 naming conventions.
13885 The following attributes apply to a @code{Naming} package:
13889 This is a simple attribute whose value is a string. Legal values of this
13890 string are @code{"lowercase"}, @code{"uppercase"} or @code{"mixedcase"}.
13891 These strings are themselves case insensitive.
13894 If @code{Casing} is not specified, then the default is @code{"lowercase"}.
13896 @item Dot_Replacement
13897 This is a simple attribute whose string value satisfies the following
13901 @item It must not be empty
13902 @item It cannot start or end with an alphanumeric character
13903 @item It cannot be a single underscore
13904 @item It cannot start with an underscore followed by an alphanumeric
13905 @item It cannot contain a dot @code{'.'} if longer than one character
13909 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
13912 This is an associative array attribute, defined on language names,
13913 whose image is a string that must satisfy the following
13917 @item It must not be empty
13918 @item It cannot start with an alphanumeric character
13919 @item It cannot start with an underscore followed by an alphanumeric character
13923 For Ada, the attribute denotes the suffix used in file names that contain
13924 library unit declarations, that is to say units that are package and
13925 subprogram declarations. If @code{Spec_Suffix ("Ada")} is not
13926 specified, then the default is @code{".ads"}.
13928 For C and C++, the attribute denotes the suffix used in file names that
13929 contain prototypes.
13932 This is an associative array attribute defined on language names,
13933 whose image is a string that must satisfy the following
13937 @item It must not be empty
13938 @item It cannot start with an alphanumeric character
13939 @item It cannot start with an underscore followed by an alphanumeric character
13940 @item It cannot be a suffix of @code{Spec_Suffix}
13944 For Ada, the attribute denotes the suffix used in file names that contain
13945 library bodies, that is to say units that are package and subprogram bodies.
13946 If @code{Body_Suffix ("Ada")} is not specified, then the default is
13949 For C and C++, the attribute denotes the suffix used in file names that contain
13952 @item Separate_Suffix
13953 This is a simple attribute whose value satisfies the same conditions as
13954 @code{Body_Suffix}.
13956 This attribute is specific to Ada. It denotes the suffix used in file names
13957 that contain separate bodies. If it is not specified, then it defaults to same
13958 value as @code{Body_Suffix ("Ada")}.
13961 This is an associative array attribute, specific to Ada, defined over
13962 compilation unit names. The image is a string that is the name of the file
13963 that contains that library unit. The file name is case sensitive if the
13964 conventions of the host operating system require it.
13967 This is an associative array attribute, specific to Ada, defined over
13968 compilation unit names. The image is a string that is the name of the file
13969 that contains the library unit body for the named unit. The file name is case
13970 sensitive if the conventions of the host operating system require it.
13972 @item Specification_Exceptions
13973 This is an associative array attribute defined on language names,
13974 whose value is a list of strings.
13976 This attribute is not significant for Ada.
13978 For C and C++, each string in the list denotes the name of a file that
13979 contains prototypes, but whose suffix is not necessarily the
13980 @code{Spec_Suffix} for the language.
13982 @item Implementation_Exceptions
13983 This is an associative array attribute defined on language names,
13984 whose value is a list of strings.
13986 This attribute is not significant for Ada.
13988 For C and C++, each string in the list denotes the name of a file that
13989 contains source code, but whose suffix is not necessarily the
13990 @code{Body_Suffix} for the language.
13993 The following attributes of package @code{Naming} are obsolescent. They are
13994 kept as synonyms of other attributes for compatibility with previous versions
13995 of the Project Manager.
13998 @item Specification_Suffix
13999 This is a synonym of @code{Spec_Suffix}.
14001 @item Implementation_Suffix
14002 This is a synonym of @code{Body_Suffix}.
14004 @item Specification
14005 This is a synonym of @code{Spec}.
14007 @item Implementation
14008 This is a synonym of @code{Body}.
14011 @subsection package Compiler
14014 The attributes of the @code{Compiler} package specify the compilation options
14015 to be used by the underlying compiler.
14018 @item Default_Switches
14019 This is an associative array attribute. Its
14020 domain is a set of language names. Its range is a string list that
14021 specifies the compilation options to be used when compiling a component
14022 written in that language, for which no file-specific switches have been
14026 This is an associative array attribute. Its domain is
14027 a set of file names. Its range is a string list that specifies the
14028 compilation options to be used when compiling the named file. If a file
14029 is not specified in the Switches attribute, it is compiled with the
14030 settings specified by Default_Switches.
14032 @item Local_Configuration_Pragmas.
14033 This is a simple attribute, whose
14034 value is a path name that designates a file containing configuration pragmas
14035 to be used for all invocations of the compiler for immediate sources of the
14039 This is an associative array attribute. Its domain is
14040 a set of main source file names. Its range is a simple string that specifies
14041 the executable file name to be used when linking the specified main source.
14042 If a main source is not specified in the Executable attribute, the executable
14043 file name is deducted from the main source file name.
14046 @subsection package Builder
14049 The attributes of package @code{Builder} specify the compilation, binding, and
14050 linking options to be used when building an executable for a project. The
14051 following attributes apply to package @code{Builder}:
14054 @item Default_Switches
14060 @item Global_Configuration_Pragmas
14061 This is a simple attribute, whose
14062 value is a path name that designates a file that contains configuration pragmas
14063 to be used in every build of an executable. If both local and global
14064 configuration pragmas are specified, a compilation makes use of both sets.
14067 This is an associative array attribute, defined over
14068 compilation unit names. The image is a string that is the name of the
14069 executable file corresponding to the main source file index.
14070 This attribute has no effect if its value is the empty string.
14072 @item Executable_Suffix
14073 This is a simple attribute whose value is a suffix to be added to
14074 the executables that don't have an attribute Executable specified.
14077 @subsection package Gnatls
14080 The attributes of package @code{Gnatls} specify the tool options to be used
14081 when invoking the library browser @command{gnatls}.
14082 The following attributes apply to package @code{Gnatls}:
14089 @subsection package Binder
14092 The attributes of package @code{Binder} specify the options to be used
14093 when invoking the binder in the construction of an executable.
14094 The following attributes apply to package @code{Binder}:
14097 @item Default_Switches
14103 @subsection package Linker
14106 The attributes of package @code{Linker} specify the options to be used when
14107 invoking the linker in the construction of an executable.
14108 The following attributes apply to package @code{Linker}:
14111 @item Default_Switches
14117 @subsection package Cross_Reference
14120 The attributes of package @code{Cross_Reference} specify the tool options
14122 when invoking the library tool @command{gnatxref}.
14123 The following attributes apply to package @code{Cross_Reference}:
14126 @item Default_Switches
14132 @subsection package Finder
14135 The attributes of package @code{Finder} specify the tool options to be used
14136 when invoking the search tool @command{gnatfind}.
14137 The following attributes apply to package @code{Finder}:
14140 @item Default_Switches
14146 @subsection package Pretty_Printer
14149 The attributes of package @code{Pretty_Printer}
14150 specify the tool options to be used
14151 when invoking the formatting tool @command{gnatpp}.
14152 The following attributes apply to package @code{Pretty_Printer}:
14155 @item Default_switches
14161 @subsection package IDE
14164 The attributes of package @code{IDE} specify the options to be used when using
14165 an Integrated Development Environment such as @command{GPS}.
14169 This is a simple attribute. Its value is a string that designates the remote
14170 host in a cross-compilation environment, to be used for remote compilation and
14171 debugging. This field should not be specified when running on the local
14175 This is a simple attribute. Its value is a string that specifies the
14176 name of IP address of the embedded target in a cross-compilation environment,
14177 on which the program should execute.
14179 @item Communication_Protocol
14180 This is a simple string attribute. Its value is the name of the protocol
14181 to use to communicate with the target in a cross-compilation environment,
14182 e.g. @code{"wtx"} or @code{"vxworks"}.
14184 @item Compiler_Command
14185 This is an associative array attribute, whose domain is a language name. Its
14186 value is string that denotes the command to be used to invoke the compiler.
14187 The value of @code{Compiler_Command ("Ada")} is expected to be compatible with
14188 gnatmake, in particular in the handling of switches.
14190 @item Debugger_Command
14191 This is simple attribute, Its value is a string that specifies the name of
14192 the debugger to be used, such as gdb, powerpc-wrs-vxworks-gdb or gdb-4.
14194 @item Default_Switches
14195 This is an associative array attribute. Its indexes are the name of the
14196 external tools that the GNAT Programming System (GPS) is supporting. Its
14197 value is a list of switches to use when invoking that tool.
14200 This is a simple attribute. Its value is a string that specifies the name
14201 of the @command{gnatls} utility to be used to retrieve information about the
14202 predefined path; e.g., @code{"gnatls"}, @code{"powerpc-wrs-vxworks-gnatls"}.
14205 This is a simple atribute. Is value is a string used to specify the
14206 Version Control System (VCS) to be used for this project, e.g CVS, RCS
14207 ClearCase or Perforce.
14209 @item VCS_File_Check
14210 This is a simple attribute. Its value is a string that specifies the
14211 command used by the VCS to check the validity of a file, either
14212 when the user explicitly asks for a check, or as a sanity check before
14213 doing the check-in.
14215 @item VCS_Log_Check
14216 This is a simple attribute. Its value is a string that specifies
14217 the command used by the VCS to check the validity of a log file.
14221 @node Package Renamings
14222 @section Package Renamings
14225 A package can be defined by a renaming declaration. The new package renames
14226 a package declared in a different project file, and has the same attributes
14227 as the package it renames.
14230 package_renaming ::==
14231 @b{package} package_identifier @b{renames}
14232 <project_>simple_name.package_identifier ;
14236 The package_identifier of the renamed package must be the same as the
14237 package_identifier. The project whose name is the prefix of the renamed
14238 package must contain a package declaration with this name. This project
14239 must appear in the context_clause of the enclosing project declaration,
14240 or be the parent project of the enclosing child project.
14246 A project file specifies a set of rules for constructing a software system.
14247 A project file can be self-contained, or depend on other project files.
14248 Dependencies are expressed through a context clause that names other projects.
14254 context_clause project_declaration
14256 project_declaration ::=
14257 simple_project_declaration | project_extension
14259 simple_project_declaration ::=
14260 @b{project} <project_>simple_name @b{is}
14261 @{declarative_item@}
14262 @b{end} <project_>simple_name;
14268 [@b{limited}] @b{with} path_name @{ , path_name @} ;
14275 A path name denotes a project file. A path name can be absolute or relative.
14276 An absolute path name includes a sequence of directories, in the syntax of
14277 the host operating system, that identifies uniquely the project file in the
14278 file system. A relative path name identifies the project file, relative
14279 to the directory that contains the current project, or relative to a
14280 directory listed in the environment variable ADA_PROJECT_PATH.
14281 Path names are case sensitive if file names in the host operating system
14282 are case sensitive.
14284 The syntax of the environment variable ADA_PROJECT_PATH is a list of
14285 directory names separated by colons (semicolons on Windows).
14287 A given project name can appear only once in a context_clause.
14289 It is illegal for a project imported by a context clause to refer, directly
14290 or indirectly, to the project in which this context clause appears (the
14291 dependency graph cannot contain cycles), except when one of the with_clause
14292 in the cycle is a @code{limited with}.
14294 @node Project Extensions
14295 @section Project Extensions
14298 A project extension introduces a new project, which inherits the declarations
14299 of another project.
14303 project_extension ::=
14304 @b{project} <project_>simple_name @b{extends} path_name @b{is}
14305 @{declarative_item@}
14306 @b{end} <project_>simple_name;
14310 The project extension declares a child project. The child project inherits
14311 all the declarations and all the files of the parent project, These inherited
14312 declaration can be overridden in the child project, by means of suitable
14315 @node Project File Elaboration
14316 @section Project File Elaboration
14319 A project file is processed as part of the invocation of a gnat tool that
14320 uses the project option. Elaboration of the process file consists in the
14321 sequential elaboration of all its declarations. The computed values of
14322 attributes and variables in the project are then used to establish the
14323 environment in which the gnat tool will execute.
14325 @node Obsolescent Features
14326 @chapter Obsolescent Features
14329 This chapter describes features that are provided by GNAT, but are
14330 considered obsolescent since there are preferred ways of achieving
14331 the same effect. These features are provided solely for historical
14332 compatibility purposes.
14335 * pragma No_Run_Time::
14336 * pragma Ravenscar::
14337 * pragma Restricted_Run_Time::
14340 @node pragma No_Run_Time
14341 @section pragma No_Run_Time
14343 The pragma @code{No_Run_Time} is used to achieve an affect similar
14344 to the use of the "Zero Foot Print" configurable run time, but without
14345 requiring a specially configured run time. The result of using this
14346 pragma, which must be used for all units in a partition, is to restrict
14347 the use of any language features requiring run-time support code. The
14348 preferred usage is to use an appropriately configured run-time that
14349 includes just those features that are to be made accessible.
14351 @node pragma Ravenscar
14352 @section pragma Ravenscar
14354 The pragma @code{Ravenscar} has exactly the same effect as pragma
14355 @code{Profile (Ravenscar)}. The latter usage is preferred since it
14356 is part of the new Ada 2005 standard.
14358 @node pragma Restricted_Run_Time
14359 @section pragma Restricted_Run_Time
14361 The pragma @code{Restricted_Run_Time} has exactly the same effect as
14362 pragma @code{Profile (Restricted)}. The latter usage is
14363 preferred since the Ada 2005 pragma @code{Profile} is intended for
14364 this kind of implementation dependent addition.
14367 @c GNU Free Documentation License
14369 @node Index,,GNU Free Documentation License, Top