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
19 @settitle GNAT Reference Manual
20 @setchapternewpage odd
23 @include gcc-common.texi
25 @dircategory GNU Ada tools
27 * GNAT Reference Manual: (gnat_rm). Reference Manual for GNU Ada tools.
31 Copyright @copyright{} 1995-2004, Free Software Foundation
33 Permission is granted to copy, distribute and/or modify this document
34 under the terms of the GNU Free Documentation License, Version 1.2
35 or any later version published by the Free Software Foundation;
36 with the Invariant Sections being ``GNU Free Documentation License'',
37 with the Front-Cover Texts being ``GNAT Reference Manual'', and with
38 no Back-Cover Texts. A copy of the license is included in the section
39 entitled ``GNU Free Documentation License''.
44 @title GNAT Reference Manual
45 @subtitle GNAT, The GNU Ada 95 Compiler
46 @subtitle GCC version @value{version-GCC}
47 @author Ada Core Technologies, Inc.
50 @vskip 0pt plus 1filll
57 @node Top, About This Guide, (dir), (dir)
58 @top GNAT Reference Manual
64 GNAT, The GNU Ada 95 Compiler@*
65 GCC version @value{version-GCC}@*
68 Ada Core Technologies, Inc.
72 * Implementation Defined Pragmas::
73 * Implementation Defined Attributes::
74 * Implementation Advice::
75 * Implementation Defined Characteristics::
76 * Intrinsic Subprograms::
77 * Representation Clauses and Pragmas::
78 * Standard Library Routines::
79 * The Implementation of Standard I/O::
81 * Interfacing to Other Languages::
82 * Specialized Needs Annexes::
83 * Implementation of Specific Ada Features::
84 * Project File Reference::
85 * GNU Free Documentation License::
88 --- The Detailed Node Listing ---
92 * What This Reference Manual Contains::
93 * Related Information::
95 Implementation Defined Pragmas
97 * Pragma Abort_Defer::
103 * Pragma C_Pass_By_Copy::
105 * Pragma Common_Object::
106 * Pragma Compile_Time_Warning::
107 * Pragma Complex_Representation::
108 * Pragma Component_Alignment::
109 * Pragma Convention_Identifier::
111 * Pragma CPP_Constructor::
112 * Pragma CPP_Virtual::
113 * Pragma CPP_Vtable::
115 * Pragma Elaboration_Checks::
117 * Pragma Export_Exception::
118 * Pragma Export_Function::
119 * Pragma Export_Object::
120 * Pragma Export_Procedure::
121 * Pragma Export_Value::
122 * Pragma Export_Valued_Procedure::
123 * Pragma Extend_System::
125 * Pragma External_Name_Casing::
126 * Pragma Finalize_Storage_Only::
127 * Pragma Float_Representation::
129 * Pragma Import_Exception::
130 * Pragma Import_Function::
131 * Pragma Import_Object::
132 * Pragma Import_Procedure::
133 * Pragma Import_Valued_Procedure::
134 * Pragma Initialize_Scalars::
135 * Pragma Inline_Always::
136 * Pragma Inline_Generic::
138 * Pragma Interface_Name::
139 * Pragma Interrupt_Handler::
140 * Pragma Interrupt_State::
141 * Pragma Keep_Names::
144 * Pragma Linker_Alias::
145 * Pragma Linker_Section::
146 * Pragma Long_Float::
147 * Pragma Machine_Attribute::
148 * Pragma Main_Storage::
150 * Pragma Normalize_Scalars::
151 * Pragma Obsolescent::
154 * Pragma Propagate_Exceptions::
155 * Pragma Psect_Object::
156 * Pragma Pure_Function::
158 * Pragma Restricted_Run_Time::
159 * Pragma Restriction_Warnings::
160 * Pragma Source_File_Name::
161 * Pragma Source_File_Name_Project::
162 * Pragma Source_Reference::
163 * Pragma Stream_Convert::
164 * Pragma Style_Checks::
166 * Pragma Suppress_All::
167 * Pragma Suppress_Exception_Locations::
168 * Pragma Suppress_Initialization::
171 * Pragma Task_Storage::
172 * Pragma Thread_Body::
173 * Pragma Time_Slice::
175 * Pragma Unchecked_Union::
176 * Pragma Unimplemented_Unit::
177 * Pragma Universal_Data::
178 * Pragma Unreferenced::
179 * Pragma Unreserve_All_Interrupts::
180 * Pragma Unsuppress::
181 * Pragma Use_VADS_Size::
182 * Pragma Validity_Checks::
185 * Pragma Weak_External::
187 Implementation Defined Attributes
197 * Default_Bit_Order::
205 * Has_Discriminants::
211 * Max_Interrupt_Priority::
213 * Maximum_Alignment::
217 * Passed_By_Reference::
228 * Unconstrained_Array::
229 * Universal_Literal_String::
230 * Unrestricted_Access::
236 The Implementation of Standard I/O
238 * Standard I/O Packages::
247 * Operations on C Streams::
248 * Interfacing to C Streams::
252 * Ada.Characters.Latin_9 (a-chlat9.ads)::
253 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
254 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
255 * Ada.Command_Line.Remove (a-colire.ads)::
256 * Ada.Command_Line.Environment (a-colien.ads)::
257 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
258 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
259 * Ada.Exceptions.Traceback (a-exctra.ads)::
260 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
261 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
262 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
263 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
264 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
265 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
266 * GNAT.Array_Split (g-arrspl.ads)::
267 * GNAT.AWK (g-awk.ads)::
268 * GNAT.Bounded_Buffers (g-boubuf.ads)::
269 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
270 * GNAT.Bubble_Sort (g-bubsor.ads)::
271 * GNAT.Bubble_Sort_A (g-busora.ads)::
272 * GNAT.Bubble_Sort_G (g-busorg.ads)::
273 * GNAT.Calendar (g-calend.ads)::
274 * GNAT.Calendar.Time_IO (g-catiio.ads)::
275 * GNAT.Case_Util (g-casuti.ads)::
276 * GNAT.CGI (g-cgi.ads)::
277 * GNAT.CGI.Cookie (g-cgicoo.ads)::
278 * GNAT.CGI.Debug (g-cgideb.ads)::
279 * GNAT.Command_Line (g-comlin.ads)::
280 * GNAT.Compiler_Version (g-comver.ads)::
281 * GNAT.Ctrl_C (g-ctrl_c.ads)::
282 * GNAT.CRC32 (g-crc32.ads)::
283 * GNAT.Current_Exception (g-curexc.ads)::
284 * GNAT.Debug_Pools (g-debpoo.ads)::
285 * GNAT.Debug_Utilities (g-debuti.ads)::
286 * GNAT.Directory_Operations (g-dirope.ads)::
287 * GNAT.Dynamic_HTables (g-dynhta.ads)::
288 * GNAT.Dynamic_Tables (g-dyntab.ads)::
289 * GNAT.Exception_Actions (g-excact.ads)::
290 * GNAT.Exception_Traces (g-exctra.ads)::
291 * GNAT.Exceptions (g-except.ads)::
292 * GNAT.Expect (g-expect.ads)::
293 * GNAT.Float_Control (g-flocon.ads)::
294 * GNAT.Heap_Sort (g-heasor.ads)::
295 * GNAT.Heap_Sort_A (g-hesora.ads)::
296 * GNAT.Heap_Sort_G (g-hesorg.ads)::
297 * GNAT.HTable (g-htable.ads)::
298 * GNAT.IO (g-io.ads)::
299 * GNAT.IO_Aux (g-io_aux.ads)::
300 * GNAT.Lock_Files (g-locfil.ads)::
301 * GNAT.MD5 (g-md5.ads)::
302 * GNAT.Memory_Dump (g-memdum.ads)::
303 * GNAT.Most_Recent_Exception (g-moreex.ads)::
304 * GNAT.OS_Lib (g-os_lib.ads)::
305 * GNAT.Perfect_Hash.Generators (g-pehage.ads)::
306 * GNAT.Regexp (g-regexp.ads)::
307 * GNAT.Registry (g-regist.ads)::
308 * GNAT.Regpat (g-regpat.ads)::
309 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
310 * GNAT.Semaphores (g-semaph.ads)::
311 * GNAT.Signals (g-signal.ads)::
312 * GNAT.Sockets (g-socket.ads)::
313 * GNAT.Source_Info (g-souinf.ads)::
314 * GNAT.Spell_Checker (g-speche.ads)::
315 * GNAT.Spitbol.Patterns (g-spipat.ads)::
316 * GNAT.Spitbol (g-spitbo.ads)::
317 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
318 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
319 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
320 * GNAT.Strings (g-string.ads)::
321 * GNAT.String_Split (g-strspl.ads)::
322 * GNAT.Table (g-table.ads)::
323 * GNAT.Task_Lock (g-tasloc.ads)::
324 * GNAT.Threads (g-thread.ads)::
325 * GNAT.Traceback (g-traceb.ads)::
326 * GNAT.Traceback.Symbolic (g-trasym.ads)::
327 * GNAT.Wide_String_Split (g-wistsp.ads)::
328 * Interfaces.C.Extensions (i-cexten.ads)::
329 * Interfaces.C.Streams (i-cstrea.ads)::
330 * Interfaces.CPP (i-cpp.ads)::
331 * Interfaces.Os2lib (i-os2lib.ads)::
332 * Interfaces.Os2lib.Errors (i-os2err.ads)::
333 * Interfaces.Os2lib.Synchronization (i-os2syn.ads)::
334 * Interfaces.Os2lib.Threads (i-os2thr.ads)::
335 * Interfaces.Packed_Decimal (i-pacdec.ads)::
336 * Interfaces.VxWorks (i-vxwork.ads)::
337 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
338 * System.Address_Image (s-addima.ads)::
339 * System.Assertions (s-assert.ads)::
340 * System.Memory (s-memory.ads)::
341 * System.Partition_Interface (s-parint.ads)::
342 * System.Restrictions (s-restri.ads)::
343 * System.Rident (s-rident.ads)::
344 * System.Task_Info (s-tasinf.ads)::
345 * System.Wch_Cnv (s-wchcnv.ads)::
346 * System.Wch_Con (s-wchcon.ads)::
350 * Text_IO Stream Pointer Positioning::
351 * Text_IO Reading and Writing Non-Regular Files::
353 * Treating Text_IO Files as Streams::
354 * Text_IO Extensions::
355 * Text_IO Facilities for Unbounded Strings::
359 * Wide_Text_IO Stream Pointer Positioning::
360 * Wide_Text_IO Reading and Writing Non-Regular Files::
362 Interfacing to Other Languages
365 * Interfacing to C++::
366 * Interfacing to COBOL::
367 * Interfacing to Fortran::
368 * Interfacing to non-GNAT Ada code::
370 Specialized Needs Annexes
372 Implementation of Specific Ada Features
373 * Machine Code Insertions::
374 * GNAT Implementation of Tasking::
375 * GNAT Implementation of Shared Passive Packages::
376 * Code Generation for Array Aggregates::
378 Project File Reference
380 GNU Free Documentation License
387 @node About This Guide
388 @unnumbered About This Guide
391 This manual contains useful information in writing programs using the
392 GNAT compiler. It includes information on implementation dependent
393 characteristics of GNAT, including all the information required by Annex
396 Ada 95 is designed to be highly portable.
397 In general, a program will have the same effect even when compiled by
398 different compilers on different platforms.
399 However, since Ada 95 is designed to be used in a
400 wide variety of applications, it also contains a number of system
401 dependent features to be used in interfacing to the external world.
402 @cindex Implementation-dependent features
405 Note: Any program that makes use of implementation-dependent features
406 may be non-portable. You should follow good programming practice and
407 isolate and clearly document any sections of your program that make use
408 of these features in a non-portable manner.
411 * What This Reference Manual Contains::
413 * Related Information::
416 @node What This Reference Manual Contains
417 @unnumberedsec What This Reference Manual Contains
420 This reference manual contains the following chapters:
424 @ref{Implementation Defined Pragmas}, lists GNAT implementation-dependent
425 pragmas, which can be used to extend and enhance the functionality of the
429 @ref{Implementation Defined Attributes}, lists GNAT
430 implementation-dependent attributes which can be used to extend and
431 enhance the functionality of the compiler.
434 @ref{Implementation Advice}, provides information on generally
435 desirable behavior which are not requirements that all compilers must
436 follow since it cannot be provided on all systems, or which may be
437 undesirable on some systems.
440 @ref{Implementation Defined Characteristics}, provides a guide to
441 minimizing implementation dependent features.
444 @ref{Intrinsic Subprograms}, describes the intrinsic subprograms
445 implemented by GNAT, and how they can be imported into user
446 application programs.
449 @ref{Representation Clauses and Pragmas}, describes in detail the
450 way that GNAT represents data, and in particular the exact set
451 of representation clauses and pragmas that is accepted.
454 @ref{Standard Library Routines}, provides a listing of packages and a
455 brief description of the functionality that is provided by Ada's
456 extensive set of standard library routines as implemented by GNAT@.
459 @ref{The Implementation of Standard I/O}, details how the GNAT
460 implementation of the input-output facilities.
463 @ref{The GNAT Library}, is a catalog of packages that complement
464 the Ada predefined library.
467 @ref{Interfacing to Other Languages}, describes how programs
468 written in Ada using GNAT can be interfaced to other programming
471 @ref{Specialized Needs Annexes}, describes the GNAT implementation of all
472 of the specialized needs annexes.
475 @ref{Implementation of Specific Ada Features}, discusses issues related
476 to GNAT's implementation of machine code insertions, tasking, and several
480 @ref{Project File Reference}, presents the syntax and semantics
485 @cindex Ada 95 ISO/ANSI Standard
487 This reference manual assumes that you are familiar with Ada 95
488 language, as described in the International Standard
489 ANSI/ISO/IEC-8652:1995, Jan 1995.
492 @unnumberedsec Conventions
493 @cindex Conventions, typographical
494 @cindex Typographical conventions
497 Following are examples of the typographical and graphic conventions used
502 @code{Functions}, @code{utility program names}, @code{standard names},
509 @file{File Names}, @samp{button names}, and @samp{field names}.
518 [optional information or parameters]
521 Examples are described by text
523 and then shown this way.
528 Commands that are entered by the user are preceded in this manual by the
529 characters @samp{$ } (dollar sign followed by space). If your system uses this
530 sequence as a prompt, then the commands will appear exactly as you see them
531 in the manual. If your system uses some other prompt, then the command will
532 appear with the @samp{$} replaced by whatever prompt character you are using.
534 @node Related Information
535 @unnumberedsec Related Information
537 See the following documents for further information on GNAT:
541 @cite{GNAT User's Guide}, which provides information on how to use
542 the GNAT compiler system.
545 @cite{Ada 95 Reference Manual}, which contains all reference
546 material for the Ada 95 programming language.
549 @cite{Ada 95 Annotated Reference Manual}, which is an annotated version
550 of the standard reference manual cited above. The annotations describe
551 detailed aspects of the design decision, and in particular contain useful
552 sections on Ada 83 compatibility.
555 @cite{DEC Ada, Technical Overview and Comparison on DIGITAL Platforms},
556 which contains specific information on compatibility between GNAT and
560 @cite{DEC Ada, Language Reference Manual, part number AA-PYZAB-TK} which
561 describes in detail the pragmas and attributes provided by the DEC Ada 83
566 @node Implementation Defined Pragmas
567 @chapter Implementation Defined Pragmas
570 Ada 95 defines a set of pragmas that can be used to supply additional
571 information to the compiler. These language defined pragmas are
572 implemented in GNAT and work as described in the Ada 95 Reference
575 In addition, Ada 95 allows implementations to define additional pragmas
576 whose meaning is defined by the implementation. GNAT provides a number
577 of these implementation-dependent pragmas which can be used to extend
578 and enhance the functionality of the compiler. This section of the GNAT
579 Reference Manual describes these additional pragmas.
581 Note that any program using these pragmas may not be portable to other
582 compilers (although GNAT implements this set of pragmas on all
583 platforms). Therefore if portability to other compilers is an important
584 consideration, the use of these pragmas should be minimized.
587 * Pragma Abort_Defer::
593 * Pragma C_Pass_By_Copy::
595 * Pragma Common_Object::
596 * Pragma Compile_Time_Warning::
597 * Pragma Complex_Representation::
598 * Pragma Component_Alignment::
599 * Pragma Convention_Identifier::
601 * Pragma CPP_Constructor::
602 * Pragma CPP_Virtual::
603 * Pragma CPP_Vtable::
605 * Pragma Elaboration_Checks::
607 * Pragma Export_Exception::
608 * Pragma Export_Function::
609 * Pragma Export_Object::
610 * Pragma Export_Procedure::
611 * Pragma Export_Value::
612 * Pragma Export_Valued_Procedure::
613 * Pragma Extend_System::
615 * Pragma External_Name_Casing::
616 * Pragma Finalize_Storage_Only::
617 * Pragma Float_Representation::
619 * Pragma Import_Exception::
620 * Pragma Import_Function::
621 * Pragma Import_Object::
622 * Pragma Import_Procedure::
623 * Pragma Import_Valued_Procedure::
624 * Pragma Initialize_Scalars::
625 * Pragma Inline_Always::
626 * Pragma Inline_Generic::
628 * Pragma Interface_Name::
629 * Pragma Interrupt_Handler::
630 * Pragma Interrupt_State::
631 * Pragma Keep_Names::
634 * Pragma Linker_Alias::
635 * Pragma Linker_Section::
636 * Pragma Long_Float::
637 * Pragma Machine_Attribute::
638 * Pragma Main_Storage::
640 * Pragma Normalize_Scalars::
641 * Pragma Obsolescent::
644 * Pragma Propagate_Exceptions::
645 * Pragma Psect_Object::
646 * Pragma Pure_Function::
648 * Pragma Restricted_Run_Time::
649 * Pragma Restriction_Warnings::
650 * Pragma Source_File_Name::
651 * Pragma Source_File_Name_Project::
652 * Pragma Source_Reference::
653 * Pragma Stream_Convert::
654 * Pragma Style_Checks::
656 * Pragma Suppress_All::
657 * Pragma Suppress_Exception_Locations::
658 * Pragma Suppress_Initialization::
661 * Pragma Task_Storage::
662 * Pragma Thread_Body::
663 * Pragma Time_Slice::
665 * Pragma Unchecked_Union::
666 * Pragma Unimplemented_Unit::
667 * Pragma Universal_Data::
668 * Pragma Unreferenced::
669 * Pragma Unreserve_All_Interrupts::
670 * Pragma Unsuppress::
671 * Pragma Use_VADS_Size::
672 * Pragma Validity_Checks::
675 * Pragma Weak_External::
678 @node Pragma Abort_Defer
679 @unnumberedsec Pragma Abort_Defer
681 @cindex Deferring aborts
689 This pragma must appear at the start of the statement sequence of a
690 handled sequence of statements (right after the @code{begin}). It has
691 the effect of deferring aborts for the sequence of statements (but not
692 for the declarations or handlers, if any, associated with this statement
696 @unnumberedsec Pragma Ada_83
705 A configuration pragma that establishes Ada 83 mode for the unit to
706 which it applies, regardless of the mode set by the command line
707 switches. In Ada 83 mode, GNAT attempts to be as compatible with
708 the syntax and semantics of Ada 83, as defined in the original Ada
709 83 Reference Manual as possible. In particular, the new Ada 95
710 keywords are not recognized, optional package bodies are allowed,
711 and generics may name types with unknown discriminants without using
712 the @code{(<>)} notation. In addition, some but not all of the additional
713 restrictions of Ada 83 are enforced.
715 Ada 83 mode is intended for two purposes. Firstly, it allows existing
716 legacy Ada 83 code to be compiled and adapted to GNAT with less effort.
717 Secondly, it aids in keeping code backwards compatible with Ada 83.
718 However, there is no guarantee that code that is processed correctly
719 by GNAT in Ada 83 mode will in fact compile and execute with an Ada
720 83 compiler, since GNAT does not enforce all the additional checks
724 @unnumberedsec Pragma Ada_95
733 A configuration pragma that establishes Ada 95 mode for the unit to which
734 it applies, regardless of the mode set by the command line switches.
735 This mode is set automatically for the @code{Ada} and @code{System}
736 packages and their children, so you need not specify it in these
737 contexts. This pragma is useful when writing a reusable component that
738 itself uses Ada 95 features, but which is intended to be usable from
739 either Ada 83 or Ada 95 programs.
741 @node Pragma Annotate
742 @unnumberedsec Pragma Annotate
747 pragma Annotate (IDENTIFIER @{, ARG@});
749 ARG ::= NAME | EXPRESSION
753 This pragma is used to annotate programs. @var{identifier} identifies
754 the type of annotation. GNAT verifies this is an identifier, but does
755 not otherwise analyze it. The @var{arg} argument
756 can be either a string literal or an
757 expression. String literals are assumed to be of type
758 @code{Standard.String}. Names of entities are simply analyzed as entity
759 names. All other expressions are analyzed as expressions, and must be
762 The analyzed pragma is retained in the tree, but not otherwise processed
763 by any part of the GNAT compiler. This pragma is intended for use by
764 external tools, including ASIS@.
767 @unnumberedsec Pragma Assert
774 [, static_string_EXPRESSION]);
778 The effect of this pragma depends on whether the corresponding command
779 line switch is set to activate assertions. The pragma expands into code
780 equivalent to the following:
783 if assertions-enabled then
784 if not boolean_EXPRESSION then
785 System.Assertions.Raise_Assert_Failure
792 The string argument, if given, is the message that will be associated
793 with the exception occurrence if the exception is raised. If no second
794 argument is given, the default message is @samp{@var{file}:@var{nnn}},
795 where @var{file} is the name of the source file containing the assert,
796 and @var{nnn} is the line number of the assert. A pragma is not a
797 statement, so if a statement sequence contains nothing but a pragma
798 assert, then a null statement is required in addition, as in:
803 pragma Assert (K > 3, "Bad value for K");
809 Note that, as with the @code{if} statement to which it is equivalent, the
810 type of the expression is either @code{Standard.Boolean}, or any type derived
811 from this standard type.
813 If assertions are disabled (switch @code{-gnata} not used), then there
814 is no effect (and in particular, any side effects from the expression
815 are suppressed). More precisely it is not quite true that the pragma
816 has no effect, since the expression is analyzed, and may cause types
817 to be frozen if they are mentioned here for the first time.
819 If assertions are enabled, then the given expression is tested, and if
820 it is @code{False} then @code{System.Assertions.Raise_Assert_Failure} is called
821 which results in the raising of @code{Assert_Failure} with the given message.
823 If the boolean expression has side effects, these side effects will turn
824 on and off with the setting of the assertions mode, resulting in
825 assertions that have an effect on the program. You should generally
826 avoid side effects in the expression arguments of this pragma. However,
827 the expressions are analyzed for semantic correctness whether or not
828 assertions are enabled, so turning assertions on and off cannot affect
829 the legality of a program.
831 @node Pragma Ast_Entry
832 @unnumberedsec Pragma Ast_Entry
838 pragma AST_Entry (entry_IDENTIFIER);
842 This pragma is implemented only in the OpenVMS implementation of GNAT@. The
843 argument is the simple name of a single entry; at most one @code{AST_Entry}
844 pragma is allowed for any given entry. This pragma must be used in
845 conjunction with the @code{AST_Entry} attribute, and is only allowed after
846 the entry declaration and in the same task type specification or single task
847 as the entry to which it applies. This pragma specifies that the given entry
848 may be used to handle an OpenVMS asynchronous system trap (@code{AST})
849 resulting from an OpenVMS system service call. The pragma does not affect
850 normal use of the entry. For further details on this pragma, see the
851 DEC Ada Language Reference Manual, section 9.12a.
853 @node Pragma C_Pass_By_Copy
854 @unnumberedsec Pragma C_Pass_By_Copy
855 @cindex Passing by copy
856 @findex C_Pass_By_Copy
860 pragma C_Pass_By_Copy
861 ([Max_Size =>] static_integer_EXPRESSION);
865 Normally the default mechanism for passing C convention records to C
866 convention subprograms is to pass them by reference, as suggested by RM
867 B.3(69). Use the configuration pragma @code{C_Pass_By_Copy} to change
868 this default, by requiring that record formal parameters be passed by
869 copy if all of the following conditions are met:
873 The size of the record type does not exceed@*@var{static_integer_expression}.
875 The record type has @code{Convention C}.
877 The formal parameter has this record type, and the subprogram has a
878 foreign (non-Ada) convention.
882 If these conditions are met the argument is passed by copy, i.e.@: in a
883 manner consistent with what C expects if the corresponding formal in the
884 C prototype is a struct (rather than a pointer to a struct).
886 You can also pass records by copy by specifying the convention
887 @code{C_Pass_By_Copy} for the record type, or by using the extended
888 @code{Import} and @code{Export} pragmas, which allow specification of
889 passing mechanisms on a parameter by parameter basis.
892 @unnumberedsec Pragma Comment
898 pragma Comment (static_string_EXPRESSION);
902 This is almost identical in effect to pragma @code{Ident}. It allows the
903 placement of a comment into the object file and hence into the
904 executable file if the operating system permits such usage. The
905 difference is that @code{Comment}, unlike @code{Ident}, has
906 no limitations on placement of the pragma (it can be placed
907 anywhere in the main source unit), and if more than one pragma
908 is used, all comments are retained.
910 @node Pragma Common_Object
911 @unnumberedsec Pragma Common_Object
912 @findex Common_Object
917 pragma Common_Object (
918 [Internal =>] LOCAL_NAME,
919 [, [External =>] EXTERNAL_SYMBOL]
920 [, [Size =>] EXTERNAL_SYMBOL] );
924 | static_string_EXPRESSION
928 This pragma enables the shared use of variables stored in overlaid
929 linker areas corresponding to the use of @code{COMMON}
930 in Fortran. The single
931 object @var{local_name} is assigned to the area designated by
932 the @var{External} argument.
933 You may define a record to correspond to a series
934 of fields. The @var{size} argument
935 is syntax checked in GNAT, but otherwise ignored.
937 @code{Common_Object} is not supported on all platforms. If no
938 support is available, then the code generator will issue a message
939 indicating that the necessary attribute for implementation of this
940 pragma is not available.
942 @node Pragma Compile_Time_Warning
943 @unnumberedsec Pragma Compile_Time_Warning
944 @findex Compile_Time_Warning
949 pragma Compile_Time_Warning
950 (boolean_EXPRESSION, static_string_EXPRESSION);
954 This pragma can be used to generate additional compile time warnings. It
955 is particularly useful in generics, where warnings can be issued for
956 specific problematic instantiations. The first parameter is a boolean
957 expression. The pragma is effective only if the value of this expression
958 is known at compile time, and has the value True. The set of expressions
959 whose values are known at compile time includes all static boolean
960 expressions, and also other values which the compiler can determine
961 at compile time (e.g. the size of a record type set by an explicit
962 size representation clause, or the value of a variable which was
963 initialized to a constant and is known not to have been modified).
964 If these conditions are met, a warning message is generated using
965 the value given as the second argument. This string value may contain
966 embedded ASCII.LF characters to break the message into multiple lines.
968 @node Pragma Complex_Representation
969 @unnumberedsec Pragma Complex_Representation
970 @findex Complex_Representation
975 pragma Complex_Representation
976 ([Entity =>] LOCAL_NAME);
980 The @var{Entity} argument must be the name of a record type which has
981 two fields of the same floating-point type. The effect of this pragma is
982 to force gcc to use the special internal complex representation form for
983 this record, which may be more efficient. Note that this may result in
984 the code for this type not conforming to standard ABI (application
985 binary interface) requirements for the handling of record types. For
986 example, in some environments, there is a requirement for passing
987 records by pointer, and the use of this pragma may result in passing
988 this type in floating-point registers.
990 @node Pragma Component_Alignment
991 @unnumberedsec Pragma Component_Alignment
992 @cindex Alignments of components
993 @findex Component_Alignment
998 pragma Component_Alignment (
999 [Form =>] ALIGNMENT_CHOICE
1000 [, [Name =>] type_LOCAL_NAME]);
1002 ALIGNMENT_CHOICE ::=
1010 Specifies the alignment of components in array or record types.
1011 The meaning of the @var{Form} argument is as follows:
1014 @findex Component_Size
1015 @item Component_Size
1016 Aligns scalar components and subcomponents of the array or record type
1017 on boundaries appropriate to their inherent size (naturally
1018 aligned). For example, 1-byte components are aligned on byte boundaries,
1019 2-byte integer components are aligned on 2-byte boundaries, 4-byte
1020 integer components are aligned on 4-byte boundaries and so on. These
1021 alignment rules correspond to the normal rules for C compilers on all
1022 machines except the VAX@.
1024 @findex Component_Size_4
1025 @item Component_Size_4
1026 Naturally aligns components with a size of four or fewer
1027 bytes. Components that are larger than 4 bytes are placed on the next
1030 @findex Storage_Unit
1032 Specifies that array or record components are byte aligned, i.e.@:
1033 aligned on boundaries determined by the value of the constant
1034 @code{System.Storage_Unit}.
1038 Specifies that array or record components are aligned on default
1039 boundaries, appropriate to the underlying hardware or operating system or
1040 both. For OpenVMS VAX systems, the @code{Default} choice is the same as
1041 the @code{Storage_Unit} choice (byte alignment). For all other systems,
1042 the @code{Default} choice is the same as @code{Component_Size} (natural
1047 If the @code{Name} parameter is present, @var{type_local_name} must
1048 refer to a local record or array type, and the specified alignment
1049 choice applies to the specified type. The use of
1050 @code{Component_Alignment} together with a pragma @code{Pack} causes the
1051 @code{Component_Alignment} pragma to be ignored. The use of
1052 @code{Component_Alignment} together with a record representation clause
1053 is only effective for fields not specified by the representation clause.
1055 If the @code{Name} parameter is absent, the pragma can be used as either
1056 a configuration pragma, in which case it applies to one or more units in
1057 accordance with the normal rules for configuration pragmas, or it can be
1058 used within a declarative part, in which case it applies to types that
1059 are declared within this declarative part, or within any nested scope
1060 within this declarative part. In either case it specifies the alignment
1061 to be applied to any record or array type which has otherwise standard
1064 If the alignment for a record or array type is not specified (using
1065 pragma @code{Pack}, pragma @code{Component_Alignment}, or a record rep
1066 clause), the GNAT uses the default alignment as described previously.
1068 @node Pragma Convention_Identifier
1069 @unnumberedsec Pragma Convention_Identifier
1070 @findex Convention_Identifier
1071 @cindex Conventions, synonyms
1075 @smallexample @c ada
1076 pragma Convention_Identifier (
1077 [Name =>] IDENTIFIER,
1078 [Convention =>] convention_IDENTIFIER);
1082 This pragma provides a mechanism for supplying synonyms for existing
1083 convention identifiers. The @code{Name} identifier can subsequently
1084 be used as a synonym for the given convention in other pragmas (including
1085 for example pragma @code{Import} or another @code{Convention_Identifier}
1086 pragma). As an example of the use of this, suppose you had legacy code
1087 which used Fortran77 as the identifier for Fortran. Then the pragma:
1089 @smallexample @c ada
1090 pragma Convention_Identifier (Fortran77, Fortran);
1094 would allow the use of the convention identifier @code{Fortran77} in
1095 subsequent code, avoiding the need to modify the sources. As another
1096 example, you could use this to parametrize convention requirements
1097 according to systems. Suppose you needed to use @code{Stdcall} on
1098 windows systems, and @code{C} on some other system, then you could
1099 define a convention identifier @code{Library} and use a single
1100 @code{Convention_Identifier} pragma to specify which convention
1101 would be used system-wide.
1103 @node Pragma CPP_Class
1104 @unnumberedsec Pragma CPP_Class
1106 @cindex Interfacing with C++
1110 @smallexample @c ada
1111 pragma CPP_Class ([Entity =>] LOCAL_NAME);
1115 The argument denotes an entity in the current declarative region
1116 that is declared as a tagged or untagged record type. It indicates that
1117 the type corresponds to an externally declared C++ class type, and is to
1118 be laid out the same way that C++ would lay out the type.
1120 If (and only if) the type is tagged, at least one component in the
1121 record must be of type @code{Interfaces.CPP.Vtable_Ptr}, corresponding
1122 to the C++ Vtable (or Vtables in the case of multiple inheritance) used
1125 Types for which @code{CPP_Class} is specified do not have assignment or
1126 equality operators defined (such operations can be imported or declared
1127 as subprograms as required). Initialization is allowed only by
1128 constructor functions (see pragma @code{CPP_Constructor}).
1130 Pragma @code{CPP_Class} is intended primarily for automatic generation
1131 using an automatic binding generator tool.
1132 See @ref{Interfacing to C++} for related information.
1134 @node Pragma CPP_Constructor
1135 @unnumberedsec Pragma CPP_Constructor
1136 @cindex Interfacing with C++
1137 @findex CPP_Constructor
1141 @smallexample @c ada
1142 pragma CPP_Constructor ([Entity =>] LOCAL_NAME);
1146 This pragma identifies an imported function (imported in the usual way
1147 with pragma @code{Import}) as corresponding to a C++
1148 constructor. The argument is a name that must have been
1149 previously mentioned in a pragma @code{Import}
1150 with @code{Convention} = @code{CPP}, and must be of one of the following
1155 @code{function @var{Fname} return @var{T}'Class}
1158 @code{function @var{Fname} (@dots{}) return @var{T}'Class}
1162 where @var{T} is a tagged type to which the pragma @code{CPP_Class} applies.
1164 The first form is the default constructor, used when an object of type
1165 @var{T} is created on the Ada side with no explicit constructor. Other
1166 constructors (including the copy constructor, which is simply a special
1167 case of the second form in which the one and only argument is of type
1168 @var{T}), can only appear in two contexts:
1172 On the right side of an initialization of an object of type @var{T}.
1174 In an extension aggregate for an object of a type derived from @var{T}.
1178 Although the constructor is described as a function that returns a value
1179 on the Ada side, it is typically a procedure with an extra implicit
1180 argument (the object being initialized) at the implementation
1181 level. GNAT issues the appropriate call, whatever it is, to get the
1182 object properly initialized.
1184 In the case of derived objects, you may use one of two possible forms
1185 for declaring and creating an object:
1188 @item @code{New_Object : Derived_T}
1189 @item @code{New_Object : Derived_T := (@var{constructor-call with} @dots{})}
1193 In the first case the default constructor is called and extension fields
1194 if any are initialized according to the default initialization
1195 expressions in the Ada declaration. In the second case, the given
1196 constructor is called and the extension aggregate indicates the explicit
1197 values of the extension fields.
1199 If no constructors are imported, it is impossible to create any objects
1200 on the Ada side. If no default constructor is imported, only the
1201 initialization forms using an explicit call to a constructor are
1204 Pragma @code{CPP_Constructor} is intended primarily for automatic generation
1205 using an automatic binding generator tool.
1206 See @ref{Interfacing to C++} for more related information.
1208 @node Pragma CPP_Virtual
1209 @unnumberedsec Pragma CPP_Virtual
1210 @cindex Interfacing to C++
1215 @smallexample @c ada
1218 [, [Vtable_Ptr =>] vtable_ENTITY,]
1219 [, [Position =>] static_integer_EXPRESSION]);
1223 This pragma serves the same function as pragma @code{Import} in that
1224 case of a virtual function imported from C++. The @var{Entity} argument
1226 primitive subprogram of a tagged type to which pragma @code{CPP_Class}
1227 applies. The @var{Vtable_Ptr} argument specifies
1228 the Vtable_Ptr component which contains the
1229 entry for this virtual function. The @var{Position} argument
1230 is the sequential number
1231 counting virtual functions for this Vtable starting at 1.
1233 The @code{Vtable_Ptr} and @code{Position} arguments may be omitted if
1234 there is one Vtable_Ptr present (single inheritance case) and all
1235 virtual functions are imported. In that case the compiler can deduce both
1238 No @code{External_Name} or @code{Link_Name} arguments are required for a
1239 virtual function, since it is always accessed indirectly via the
1240 appropriate Vtable entry.
1242 Pragma @code{CPP_Virtual} is intended primarily for automatic generation
1243 using an automatic binding generator tool.
1244 See @ref{Interfacing to C++} for related information.
1246 @node Pragma CPP_Vtable
1247 @unnumberedsec Pragma CPP_Vtable
1248 @cindex Interfacing with C++
1253 @smallexample @c ada
1256 [Vtable_Ptr =>] vtable_ENTITY,
1257 [Entry_Count =>] static_integer_EXPRESSION);
1261 Given a record to which the pragma @code{CPP_Class} applies,
1262 this pragma can be specified for each component of type
1263 @code{CPP.Interfaces.Vtable_Ptr}.
1264 @var{Entity} is the tagged type, @var{Vtable_Ptr}
1265 is the record field of type @code{Vtable_Ptr}, and @var{Entry_Count} is
1266 the number of virtual functions on the C++ side. Not all of these
1267 functions need to be imported on the Ada side.
1269 You may omit the @code{CPP_Vtable} pragma if there is only one
1270 @code{Vtable_Ptr} component in the record and all virtual functions are
1271 imported on the Ada side (the default value for the entry count in this
1272 case is simply the total number of virtual functions).
1274 Pragma @code{CPP_Vtable} is intended primarily for automatic generation
1275 using an automatic binding generator tool.
1276 See @ref{Interfacing to C++} for related information.
1279 @unnumberedsec Pragma Debug
1284 @smallexample @c ada
1285 pragma Debug (PROCEDURE_CALL_WITHOUT_SEMICOLON);
1287 PROCEDURE_CALL_WITHOUT_SEMICOLON ::=
1289 | PROCEDURE_PREFIX ACTUAL_PARAMETER_PART
1293 The argument has the syntactic form of an expression, meeting the
1294 syntactic requirements for pragmas.
1296 If assertions are not enabled on the command line, this pragma has no
1297 effect. If asserts are enabled, the semantics of the pragma is exactly
1298 equivalent to the procedure call statement corresponding to the argument
1299 with a terminating semicolon. Pragmas are permitted in sequences of
1300 declarations, so you can use pragma @code{Debug} to intersperse calls to
1301 debug procedures in the middle of declarations.
1303 @node Pragma Elaboration_Checks
1304 @unnumberedsec Pragma Elaboration_Checks
1305 @cindex Elaboration control
1306 @findex Elaboration_Checks
1310 @smallexample @c ada
1311 pragma Elaboration_Checks (Dynamic | Static);
1315 This is a configuration pragma that provides control over the
1316 elaboration model used by the compilation affected by the
1317 pragma. If the parameter is @code{Dynamic},
1318 then the dynamic elaboration
1319 model described in the Ada Reference Manual is used, as though
1320 the @code{-gnatE} switch had been specified on the command
1321 line. If the parameter is @code{Static}, then the default GNAT static
1322 model is used. This configuration pragma overrides the setting
1323 of the command line. For full details on the elaboration models
1324 used by the GNAT compiler, see section ``Elaboration Order
1325 Handling in GNAT'' in the @cite{GNAT User's Guide}.
1327 @node Pragma Eliminate
1328 @unnumberedsec Pragma Eliminate
1329 @cindex Elimination of unused subprograms
1334 @smallexample @c ada
1336 [Unit_Name =>] IDENTIFIER |
1337 SELECTED_COMPONENT);
1340 [Unit_Name =>] IDENTIFIER |
1342 [Entity =>] IDENTIFIER |
1343 SELECTED_COMPONENT |
1345 [,OVERLOADING_RESOLUTION]);
1347 OVERLOADING_RESOLUTION ::= PARAMETER_AND_RESULT_TYPE_PROFILE |
1350 PARAMETER_AND_RESULT_TYPE_PROFILE ::= PROCEDURE_PROFILE |
1353 PROCEDURE_PROFILE ::= Parameter_Types => PARAMETER_TYPES
1355 FUNCTION_PROFILE ::= [Parameter_Types => PARAMETER_TYPES,]
1356 Result_Type => result_SUBTYPE_NAME]
1358 PARAMETER_TYPES ::= (SUBTYPE_NAME @{, SUBTYPE_NAME@})
1359 SUBTYPE_NAME ::= STRING_VALUE
1361 SOURCE_LOCATION ::= Source_Location => SOURCE_TRACE
1362 SOURCE_TRACE ::= STRING_VALUE
1364 STRING_VALUE ::= STRING_LITERAL @{& STRING_LITERAL@}
1368 This pragma indicates that the given entity is not used outside the
1369 compilation unit it is defined in. The entity must be an explicitly declared
1370 subprogram; this includes generic subprogram instances and
1371 subprograms declared in generic package instances.
1373 If the entity to be eliminated is a library level subprogram, then
1374 the first form of pragma @code{Eliminate} is used with only a single argument.
1375 In this form, the @code{Unit_Name} argument specifies the name of the
1376 library level unit to be eliminated.
1378 In all other cases, both @code{Unit_Name} and @code{Entity} arguments
1379 are required. If item is an entity of a library package, then the first
1380 argument specifies the unit name, and the second argument specifies
1381 the particular entity. If the second argument is in string form, it must
1382 correspond to the internal manner in which GNAT stores entity names (see
1383 compilation unit Namet in the compiler sources for details).
1385 The remaining parameters (OVERLOADING_RESOLUTION) are optionally used
1386 to distinguish between overloaded subprograms. If a pragma does not contain
1387 the OVERLOADING_RESOLUTION parameter(s), it is applied to all the overloaded
1388 subprograms denoted by the first two parameters.
1390 Use PARAMETER_AND_RESULT_TYPE_PROFILE to specify the profile of the subprogram
1391 to be eliminated in a manner similar to that used for the extended
1392 @code{Import} and @code{Export} pragmas, except that the subtype names are
1393 always given as strings. At the moment, this form of distinguishing
1394 overloaded subprograms is implemented only partially, so we do not recommend
1395 using it for practical subprogram elimination.
1397 Note, that in case of a parameterless procedure its profile is represented
1398 as @code{Parameter_Types => ("")}
1400 Alternatively, the @code{Source_Location} parameter is used to specify
1401 which overloaded alternative is to be eliminated by pointing to the
1402 location of the DEFINING_PROGRAM_UNIT_NAME of this subprogram in the
1403 source text. The string literal (or concatenation of string literals)
1404 given as SOURCE_TRACE must have the following format:
1406 @smallexample @c ada
1407 SOURCE_TRACE ::= SOURCE_LOCATION@{LBRACKET SOURCE_LOCATION RBRACKET@}
1412 SOURCE_LOCATION ::= FILE_NAME:LINE_NUMBER
1413 FILE_NAME ::= STRING_LITERAL
1414 LINE_NUMBER ::= DIGIT @{DIGIT@}
1417 SOURCE_TRACE should be the short name of the source file (with no directory
1418 information), and LINE_NUMBER is supposed to point to the line where the
1419 defining name of the subprogram is located.
1421 For the subprograms that are not a part of generic instantiations, only one
1422 SOURCE_LOCATION is used. If a subprogram is declared in a package
1423 instantiation, SOURCE_TRACE contains two SOURCE_LOCATIONs, the first one is
1424 the location of the (DEFINING_PROGRAM_UNIT_NAME of the) instantiation, and the
1425 second one denotes the declaration of the corresponding subprogram in the
1426 generic package. This approach is recursively used to create SOURCE_LOCATIONs
1427 in case of nested instantiations.
1429 The effect of the pragma is to allow the compiler to eliminate
1430 the code or data associated with the named entity. Any reference to
1431 an eliminated entity outside the compilation unit it is defined in,
1432 causes a compile time or link time error.
1434 The intention of pragma @code{Eliminate} is to allow a program to be compiled
1435 in a system independent manner, with unused entities eliminated, without
1436 the requirement of modifying the source text. Normally the required set
1437 of @code{Eliminate} pragmas is constructed automatically using the gnatelim
1438 tool. Elimination of unused entities local to a compilation unit is
1439 automatic, without requiring the use of pragma @code{Eliminate}.
1441 Note that the reason this pragma takes string literals where names might
1442 be expected is that a pragma @code{Eliminate} can appear in a context where the
1443 relevant names are not visible.
1445 Note that any change in the source files that includes removing, splitting of
1446 adding lines may make the set of Eliminate pragmas using SOURCE_LOCATION
1449 @node Pragma Export_Exception
1450 @unnumberedsec Pragma Export_Exception
1452 @findex Export_Exception
1456 @smallexample @c ada
1457 pragma Export_Exception (
1458 [Internal =>] LOCAL_NAME,
1459 [, [External =>] EXTERNAL_SYMBOL,]
1460 [, [Form =>] Ada | VMS]
1461 [, [Code =>] static_integer_EXPRESSION]);
1465 | static_string_EXPRESSION
1469 This pragma is implemented only in the OpenVMS implementation of GNAT@. It
1470 causes the specified exception to be propagated outside of the Ada program,
1471 so that it can be handled by programs written in other OpenVMS languages.
1472 This pragma establishes an external name for an Ada exception and makes the
1473 name available to the OpenVMS Linker as a global symbol. For further details
1474 on this pragma, see the
1475 DEC Ada Language Reference Manual, section 13.9a3.2.
1477 @node Pragma Export_Function
1478 @unnumberedsec Pragma Export_Function
1479 @cindex Argument passing mechanisms
1480 @findex Export_Function
1485 @smallexample @c ada
1486 pragma Export_Function (
1487 [Internal =>] LOCAL_NAME,
1488 [, [External =>] EXTERNAL_SYMBOL]
1489 [, [Parameter_Types =>] PARAMETER_TYPES]
1490 [, [Result_Type =>] result_SUBTYPE_MARK]
1491 [, [Mechanism =>] MECHANISM]
1492 [, [Result_Mechanism =>] MECHANISM_NAME]);
1496 | static_string_EXPRESSION
1501 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1505 | subtype_Name ' Access
1509 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1511 MECHANISM_ASSOCIATION ::=
1512 [formal_parameter_NAME =>] MECHANISM_NAME
1520 Use this pragma to make a function externally callable and optionally
1521 provide information on mechanisms to be used for passing parameter and
1522 result values. We recommend, for the purposes of improving portability,
1523 this pragma always be used in conjunction with a separate pragma
1524 @code{Export}, which must precede the pragma @code{Export_Function}.
1525 GNAT does not require a separate pragma @code{Export}, but if none is
1526 present, @code{Convention Ada} is assumed, which is usually
1527 not what is wanted, so it is usually appropriate to use this
1528 pragma in conjunction with a @code{Export} or @code{Convention}
1529 pragma that specifies the desired foreign convention.
1530 Pragma @code{Export_Function}
1531 (and @code{Export}, if present) must appear in the same declarative
1532 region as the function to which they apply.
1534 @var{internal_name} must uniquely designate the function to which the
1535 pragma applies. If more than one function name exists of this name in
1536 the declarative part you must use the @code{Parameter_Types} and
1537 @code{Result_Type} parameters is mandatory to achieve the required
1538 unique designation. @var{subtype_ mark}s in these parameters must
1539 exactly match the subtypes in the corresponding function specification,
1540 using positional notation to match parameters with subtype marks.
1541 The form with an @code{'Access} attribute can be used to match an
1542 anonymous access parameter.
1545 @cindex Passing by descriptor
1546 Note that passing by descriptor is not supported, even on the OpenVMS
1549 @cindex Suppressing external name
1550 Special treatment is given if the EXTERNAL is an explicit null
1551 string or a static string expressions that evaluates to the null
1552 string. In this case, no external name is generated. This form
1553 still allows the specification of parameter mechanisms.
1555 @node Pragma Export_Object
1556 @unnumberedsec Pragma Export_Object
1557 @findex Export_Object
1561 @smallexample @c ada
1562 pragma Export_Object
1563 [Internal =>] LOCAL_NAME,
1564 [, [External =>] EXTERNAL_SYMBOL]
1565 [, [Size =>] EXTERNAL_SYMBOL]
1569 | static_string_EXPRESSION
1573 This pragma designates an object as exported, and apart from the
1574 extended rules for external symbols, is identical in effect to the use of
1575 the normal @code{Export} pragma applied to an object. You may use a
1576 separate Export pragma (and you probably should from the point of view
1577 of portability), but it is not required. @var{Size} is syntax checked,
1578 but otherwise ignored by GNAT@.
1580 @node Pragma Export_Procedure
1581 @unnumberedsec Pragma Export_Procedure
1582 @findex Export_Procedure
1586 @smallexample @c ada
1587 pragma Export_Procedure (
1588 [Internal =>] LOCAL_NAME
1589 [, [External =>] EXTERNAL_SYMBOL]
1590 [, [Parameter_Types =>] PARAMETER_TYPES]
1591 [, [Mechanism =>] MECHANISM]);
1595 | static_string_EXPRESSION
1600 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1604 | subtype_Name ' Access
1608 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1610 MECHANISM_ASSOCIATION ::=
1611 [formal_parameter_NAME =>] MECHANISM_NAME
1619 This pragma is identical to @code{Export_Function} except that it
1620 applies to a procedure rather than a function and the parameters
1621 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
1622 GNAT does not require a separate pragma @code{Export}, but if none is
1623 present, @code{Convention Ada} is assumed, which is usually
1624 not what is wanted, so it is usually appropriate to use this
1625 pragma in conjunction with a @code{Export} or @code{Convention}
1626 pragma that specifies the desired foreign convention.
1629 @cindex Passing by descriptor
1630 Note that passing by descriptor is not supported, even on the OpenVMS
1633 @cindex Suppressing external name
1634 Special treatment is given if the EXTERNAL is an explicit null
1635 string or a static string expressions that evaluates to the null
1636 string. In this case, no external name is generated. This form
1637 still allows the specification of parameter mechanisms.
1639 @node Pragma Export_Value
1640 @unnumberedsec Pragma Export_Value
1641 @findex Export_Value
1645 @smallexample @c ada
1646 pragma Export_Value (
1647 [Value =>] static_integer_EXPRESSION,
1648 [Link_Name =>] static_string_EXPRESSION);
1652 This pragma serves to export a static integer value for external use.
1653 The first argument specifies the value to be exported. The Link_Name
1654 argument specifies the symbolic name to be associated with the integer
1655 value. This pragma is useful for defining a named static value in Ada
1656 that can be referenced in assembly language units to be linked with
1657 the application. This pragma is currently supported only for the
1658 AAMP target and is ignored for other targets.
1660 @node Pragma Export_Valued_Procedure
1661 @unnumberedsec Pragma Export_Valued_Procedure
1662 @findex Export_Valued_Procedure
1666 @smallexample @c ada
1667 pragma Export_Valued_Procedure (
1668 [Internal =>] LOCAL_NAME
1669 [, [External =>] EXTERNAL_SYMBOL]
1670 [, [Parameter_Types =>] PARAMETER_TYPES]
1671 [, [Mechanism =>] MECHANISM]);
1675 | static_string_EXPRESSION
1680 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1684 | subtype_Name ' Access
1688 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
1690 MECHANISM_ASSOCIATION ::=
1691 [formal_parameter_NAME =>] MECHANISM_NAME
1699 This pragma is identical to @code{Export_Procedure} except that the
1700 first parameter of @var{local_name}, which must be present, must be of
1701 mode @code{OUT}, and externally the subprogram is treated as a function
1702 with this parameter as the result of the function. GNAT provides for
1703 this capability to allow the use of @code{OUT} and @code{IN OUT}
1704 parameters in interfacing to external functions (which are not permitted
1706 GNAT does not require a separate pragma @code{Export}, but if none is
1707 present, @code{Convention Ada} is assumed, which is almost certainly
1708 not what is wanted since the whole point of this pragma is to interface
1709 with foreign language functions, so it is usually appropriate to use this
1710 pragma in conjunction with a @code{Export} or @code{Convention}
1711 pragma that specifies the desired foreign convention.
1714 @cindex Passing by descriptor
1715 Note that passing by descriptor is not supported, even on the OpenVMS
1718 @cindex Suppressing external name
1719 Special treatment is given if the EXTERNAL is an explicit null
1720 string or a static string expressions that evaluates to the null
1721 string. In this case, no external name is generated. This form
1722 still allows the specification of parameter mechanisms.
1724 @node Pragma Extend_System
1725 @unnumberedsec Pragma Extend_System
1726 @cindex @code{system}, extending
1728 @findex Extend_System
1732 @smallexample @c ada
1733 pragma Extend_System ([Name =>] IDENTIFIER);
1737 This pragma is used to provide backwards compatibility with other
1738 implementations that extend the facilities of package @code{System}. In
1739 GNAT, @code{System} contains only the definitions that are present in
1740 the Ada 95 RM@. However, other implementations, notably the DEC Ada 83
1741 implementation, provide many extensions to package @code{System}.
1743 For each such implementation accommodated by this pragma, GNAT provides a
1744 package @code{Aux_@var{xxx}}, e.g.@: @code{Aux_DEC} for the DEC Ada 83
1745 implementation, which provides the required additional definitions. You
1746 can use this package in two ways. You can @code{with} it in the normal
1747 way and access entities either by selection or using a @code{use}
1748 clause. In this case no special processing is required.
1750 However, if existing code contains references such as
1751 @code{System.@var{xxx}} where @var{xxx} is an entity in the extended
1752 definitions provided in package @code{System}, you may use this pragma
1753 to extend visibility in @code{System} in a non-standard way that
1754 provides greater compatibility with the existing code. Pragma
1755 @code{Extend_System} is a configuration pragma whose single argument is
1756 the name of the package containing the extended definition
1757 (e.g.@: @code{Aux_DEC} for the DEC Ada case). A unit compiled under
1758 control of this pragma will be processed using special visibility
1759 processing that looks in package @code{System.Aux_@var{xxx}} where
1760 @code{Aux_@var{xxx}} is the pragma argument for any entity referenced in
1761 package @code{System}, but not found in package @code{System}.
1763 You can use this pragma either to access a predefined @code{System}
1764 extension supplied with the compiler, for example @code{Aux_DEC} or
1765 you can construct your own extension unit following the above
1766 definition. Note that such a package is a child of @code{System}
1767 and thus is considered part of the implementation. To compile
1768 it you will have to use the appropriate switch for compiling
1769 system units. See the GNAT User's Guide for details.
1771 @node Pragma External
1772 @unnumberedsec Pragma External
1777 @smallexample @c ada
1779 [ Convention =>] convention_IDENTIFIER,
1780 [ Entity =>] local_NAME
1781 [, [External_Name =>] static_string_EXPRESSION ]
1782 [, [Link_Name =>] static_string_EXPRESSION ]);
1786 This pragma is identical in syntax and semantics to pragma
1787 @code{Export} as defined in the Ada Reference Manual. It is
1788 provided for compatibility with some Ada 83 compilers that
1789 used this pragma for exactly the same purposes as pragma
1790 @code{Export} before the latter was standardized.
1792 @node Pragma External_Name_Casing
1793 @unnumberedsec Pragma External_Name_Casing
1794 @cindex Dec Ada 83 casing compatibility
1795 @cindex External Names, casing
1796 @cindex Casing of External names
1797 @findex External_Name_Casing
1801 @smallexample @c ada
1802 pragma External_Name_Casing (
1803 Uppercase | Lowercase
1804 [, Uppercase | Lowercase | As_Is]);
1808 This pragma provides control over the casing of external names associated
1809 with Import and Export pragmas. There are two cases to consider:
1812 @item Implicit external names
1813 Implicit external names are derived from identifiers. The most common case
1814 arises when a standard Ada 95 Import or Export pragma is used with only two
1817 @smallexample @c ada
1818 pragma Import (C, C_Routine);
1822 Since Ada is a case insensitive language, the spelling of the identifier in
1823 the Ada source program does not provide any information on the desired
1824 casing of the external name, and so a convention is needed. In GNAT the
1825 default treatment is that such names are converted to all lower case
1826 letters. This corresponds to the normal C style in many environments.
1827 The first argument of pragma @code{External_Name_Casing} can be used to
1828 control this treatment. If @code{Uppercase} is specified, then the name
1829 will be forced to all uppercase letters. If @code{Lowercase} is specified,
1830 then the normal default of all lower case letters will be used.
1832 This same implicit treatment is also used in the case of extended DEC Ada 83
1833 compatible Import and Export pragmas where an external name is explicitly
1834 specified using an identifier rather than a string.
1836 @item Explicit external names
1837 Explicit external names are given as string literals. The most common case
1838 arises when a standard Ada 95 Import or Export pragma is used with three
1841 @smallexample @c ada
1842 pragma Import (C, C_Routine, "C_routine");
1846 In this case, the string literal normally provides the exact casing required
1847 for the external name. The second argument of pragma
1848 @code{External_Name_Casing} may be used to modify this behavior.
1849 If @code{Uppercase} is specified, then the name
1850 will be forced to all uppercase letters. If @code{Lowercase} is specified,
1851 then the name will be forced to all lowercase letters. A specification of
1852 @code{As_Is} provides the normal default behavior in which the casing is
1853 taken from the string provided.
1857 This pragma may appear anywhere that a pragma is valid. In particular, it
1858 can be used as a configuration pragma in the @file{gnat.adc} file, in which
1859 case it applies to all subsequent compilations, or it can be used as a program
1860 unit pragma, in which case it only applies to the current unit, or it can
1861 be used more locally to control individual Import/Export pragmas.
1863 It is primarily intended for use with OpenVMS systems, where many
1864 compilers convert all symbols to upper case by default. For interfacing to
1865 such compilers (e.g.@: the DEC C compiler), it may be convenient to use
1868 @smallexample @c ada
1869 pragma External_Name_Casing (Uppercase, Uppercase);
1873 to enforce the upper casing of all external symbols.
1875 @node Pragma Finalize_Storage_Only
1876 @unnumberedsec Pragma Finalize_Storage_Only
1877 @findex Finalize_Storage_Only
1881 @smallexample @c ada
1882 pragma Finalize_Storage_Only (first_subtype_LOCAL_NAME);
1886 This pragma allows the compiler not to emit a Finalize call for objects
1887 defined at the library level. This is mostly useful for types where
1888 finalization is only used to deal with storage reclamation since in most
1889 environments it is not necessary to reclaim memory just before terminating
1890 execution, hence the name.
1892 @node Pragma Float_Representation
1893 @unnumberedsec Pragma Float_Representation
1895 @findex Float_Representation
1899 @smallexample @c ada
1900 pragma Float_Representation (FLOAT_REP);
1902 FLOAT_REP ::= VAX_Float | IEEE_Float
1907 allows control over the internal representation chosen for the predefined
1908 floating point types declared in the packages @code{Standard} and
1909 @code{System}. On all systems other than OpenVMS, the argument must
1910 be @code{IEEE_Float} and the pragma has no effect. On OpenVMS, the
1911 argument may be @code{VAX_Float} to specify the use of the VAX float
1912 format for the floating-point types in Standard. This requires that
1913 the standard runtime libraries be recompiled. See the
1914 description of the @code{GNAT LIBRARY} command in the OpenVMS version
1915 of the GNAT Users Guide for details on the use of this command.
1918 @unnumberedsec Pragma Ident
1923 @smallexample @c ada
1924 pragma Ident (static_string_EXPRESSION);
1928 This pragma provides a string identification in the generated object file,
1929 if the system supports the concept of this kind of identification string.
1930 This pragma is allowed only in the outermost declarative part or
1931 declarative items of a compilation unit. If more than one @code{Ident}
1932 pragma is given, only the last one processed is effective.
1934 On OpenVMS systems, the effect of the pragma is identical to the effect of
1935 the DEC Ada 83 pragma of the same name. Note that in DEC Ada 83, the
1936 maximum allowed length is 31 characters, so if it is important to
1937 maintain compatibility with this compiler, you should obey this length
1940 @node Pragma Import_Exception
1941 @unnumberedsec Pragma Import_Exception
1943 @findex Import_Exception
1947 @smallexample @c ada
1948 pragma Import_Exception (
1949 [Internal =>] LOCAL_NAME,
1950 [, [External =>] EXTERNAL_SYMBOL,]
1951 [, [Form =>] Ada | VMS]
1952 [, [Code =>] static_integer_EXPRESSION]);
1956 | static_string_EXPRESSION
1960 This pragma is implemented only in the OpenVMS implementation of GNAT@.
1961 It allows OpenVMS conditions (for example, from OpenVMS system services or
1962 other OpenVMS languages) to be propagated to Ada programs as Ada exceptions.
1963 The pragma specifies that the exception associated with an exception
1964 declaration in an Ada program be defined externally (in non-Ada code).
1965 For further details on this pragma, see the
1966 DEC Ada Language Reference Manual, section 13.9a.3.1.
1968 @node Pragma Import_Function
1969 @unnumberedsec Pragma Import_Function
1970 @findex Import_Function
1974 @smallexample @c ada
1975 pragma Import_Function (
1976 [Internal =>] LOCAL_NAME,
1977 [, [External =>] EXTERNAL_SYMBOL]
1978 [, [Parameter_Types =>] PARAMETER_TYPES]
1979 [, [Result_Type =>] SUBTYPE_MARK]
1980 [, [Mechanism =>] MECHANISM]
1981 [, [Result_Mechanism =>] MECHANISM_NAME]
1982 [, [First_Optional_Parameter =>] IDENTIFIER]);
1986 | static_string_EXPRESSION
1990 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
1994 | subtype_Name ' Access
1998 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2000 MECHANISM_ASSOCIATION ::=
2001 [formal_parameter_NAME =>] MECHANISM_NAME
2006 | Descriptor [([Class =>] CLASS_NAME)]
2008 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2012 This pragma is used in conjunction with a pragma @code{Import} to
2013 specify additional information for an imported function. The pragma
2014 @code{Import} (or equivalent pragma @code{Interface}) must precede the
2015 @code{Import_Function} pragma and both must appear in the same
2016 declarative part as the function specification.
2018 The @var{Internal} argument must uniquely designate
2019 the function to which the
2020 pragma applies. If more than one function name exists of this name in
2021 the declarative part you must use the @code{Parameter_Types} and
2022 @var{Result_Type} parameters to achieve the required unique
2023 designation. Subtype marks in these parameters must exactly match the
2024 subtypes in the corresponding function specification, using positional
2025 notation to match parameters with subtype marks.
2026 The form with an @code{'Access} attribute can be used to match an
2027 anonymous access parameter.
2029 You may optionally use the @var{Mechanism} and @var{Result_Mechanism}
2030 parameters to specify passing mechanisms for the
2031 parameters and result. If you specify a single mechanism name, it
2032 applies to all parameters. Otherwise you may specify a mechanism on a
2033 parameter by parameter basis using either positional or named
2034 notation. If the mechanism is not specified, the default mechanism
2038 @cindex Passing by descriptor
2039 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
2041 @code{First_Optional_Parameter} applies only to OpenVMS ports of GNAT@.
2042 It specifies that the designated parameter and all following parameters
2043 are optional, meaning that they are not passed at the generated code
2044 level (this is distinct from the notion of optional parameters in Ada
2045 where the parameters are passed anyway with the designated optional
2046 parameters). All optional parameters must be of mode @code{IN} and have
2047 default parameter values that are either known at compile time
2048 expressions, or uses of the @code{'Null_Parameter} attribute.
2050 @node Pragma Import_Object
2051 @unnumberedsec Pragma Import_Object
2052 @findex Import_Object
2056 @smallexample @c ada
2057 pragma Import_Object
2058 [Internal =>] LOCAL_NAME,
2059 [, [External =>] EXTERNAL_SYMBOL],
2060 [, [Size =>] EXTERNAL_SYMBOL]);
2064 | static_string_EXPRESSION
2068 This pragma designates an object as imported, and apart from the
2069 extended rules for external symbols, is identical in effect to the use of
2070 the normal @code{Import} pragma applied to an object. Unlike the
2071 subprogram case, you need not use a separate @code{Import} pragma,
2072 although you may do so (and probably should do so from a portability
2073 point of view). @var{size} is syntax checked, but otherwise ignored by
2076 @node Pragma Import_Procedure
2077 @unnumberedsec Pragma Import_Procedure
2078 @findex Import_Procedure
2082 @smallexample @c ada
2083 pragma Import_Procedure (
2084 [Internal =>] LOCAL_NAME,
2085 [, [External =>] EXTERNAL_SYMBOL]
2086 [, [Parameter_Types =>] PARAMETER_TYPES]
2087 [, [Mechanism =>] MECHANISM]
2088 [, [First_Optional_Parameter =>] IDENTIFIER]);
2092 | static_string_EXPRESSION
2096 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2100 | subtype_Name ' Access
2104 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2106 MECHANISM_ASSOCIATION ::=
2107 [formal_parameter_NAME =>] MECHANISM_NAME
2112 | Descriptor [([Class =>] CLASS_NAME)]
2114 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2118 This pragma is identical to @code{Import_Function} except that it
2119 applies to a procedure rather than a function and the parameters
2120 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
2122 @node Pragma Import_Valued_Procedure
2123 @unnumberedsec Pragma Import_Valued_Procedure
2124 @findex Import_Valued_Procedure
2128 @smallexample @c ada
2129 pragma Import_Valued_Procedure (
2130 [Internal =>] LOCAL_NAME,
2131 [, [External =>] EXTERNAL_SYMBOL]
2132 [, [Parameter_Types =>] PARAMETER_TYPES]
2133 [, [Mechanism =>] MECHANISM]
2134 [, [First_Optional_Parameter =>] IDENTIFIER]);
2138 | static_string_EXPRESSION
2142 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2146 | subtype_Name ' Access
2150 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2152 MECHANISM_ASSOCIATION ::=
2153 [formal_parameter_NAME =>] MECHANISM_NAME
2158 | Descriptor [([Class =>] CLASS_NAME)]
2160 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
2164 This pragma is identical to @code{Import_Procedure} except that the
2165 first parameter of @var{local_name}, which must be present, must be of
2166 mode @code{OUT}, and externally the subprogram is treated as a function
2167 with this parameter as the result of the function. The purpose of this
2168 capability is to allow the use of @code{OUT} and @code{IN OUT}
2169 parameters in interfacing to external functions (which are not permitted
2170 in Ada functions). You may optionally use the @code{Mechanism}
2171 parameters to specify passing mechanisms for the parameters.
2172 If you specify a single mechanism name, it applies to all parameters.
2173 Otherwise you may specify a mechanism on a parameter by parameter
2174 basis using either positional or named notation. If the mechanism is not
2175 specified, the default mechanism is used.
2177 Note that it is important to use this pragma in conjunction with a separate
2178 pragma Import that specifies the desired convention, since otherwise the
2179 default convention is Ada, which is almost certainly not what is required.
2181 @node Pragma Initialize_Scalars
2182 @unnumberedsec Pragma Initialize_Scalars
2183 @findex Initialize_Scalars
2184 @cindex debugging with Initialize_Scalars
2188 @smallexample @c ada
2189 pragma Initialize_Scalars;
2193 This pragma is similar to @code{Normalize_Scalars} conceptually but has
2194 two important differences. First, there is no requirement for the pragma
2195 to be used uniformly in all units of a partition, in particular, it is fine
2196 to use this just for some or all of the application units of a partition,
2197 without needing to recompile the run-time library.
2199 In the case where some units are compiled with the pragma, and some without,
2200 then a declaration of a variable where the type is defined in package
2201 Standard or is locally declared will always be subject to initialization,
2202 as will any declaration of a scalar variable. For composite variables,
2203 whether the variable is initialized may also depend on whether the package
2204 in which the type of the variable is declared is compiled with the pragma.
2206 The other important difference is that there is control over the value used
2207 for initializing scalar objects. At bind time, you can select whether to
2208 initialize with invalid values (like Normalize_Scalars), or with high or
2209 low values, or with a specified bit pattern. See the users guide for binder
2210 options for specifying these cases.
2212 This means that you can compile a program, and then without having to
2213 recompile the program, you can run it with different values being used
2214 for initializing otherwise uninitialized values, to test if your program
2215 behavior depends on the choice. Of course the behavior should not change,
2216 and if it does, then most likely you have an erroneous reference to an
2217 uninitialized value.
2219 Note that pragma @code{Initialize_Scalars} is particularly useful in
2220 conjunction with the enhanced validity checking that is now provided
2221 in GNAT, which checks for invalid values under more conditions.
2222 Using this feature (see description of the @code{-gnatV} flag in the
2223 users guide) in conjunction with pragma @code{Initialize_Scalars}
2224 provides a powerful new tool to assist in the detection of problems
2225 caused by uninitialized variables.
2227 @node Pragma Inline_Always
2228 @unnumberedsec Pragma Inline_Always
2229 @findex Inline_Always
2233 @smallexample @c ada
2234 pragma Inline_Always (NAME [, NAME]);
2238 Similar to pragma @code{Inline} except that inlining is not subject to
2239 the use of option @code{-gnatn} and the inlining happens regardless of
2240 whether this option is used.
2242 @node Pragma Inline_Generic
2243 @unnumberedsec Pragma Inline_Generic
2244 @findex Inline_Generic
2248 @smallexample @c ada
2249 pragma Inline_Generic (generic_package_NAME);
2253 This is implemented for compatibility with DEC Ada 83 and is recognized,
2254 but otherwise ignored, by GNAT@. All generic instantiations are inlined
2255 by default when using GNAT@.
2257 @node Pragma Interface
2258 @unnumberedsec Pragma Interface
2263 @smallexample @c ada
2265 [Convention =>] convention_identifier,
2266 [Entity =>] local_name
2267 [, [External_Name =>] static_string_expression],
2268 [, [Link_Name =>] static_string_expression]);
2272 This pragma is identical in syntax and semantics to
2273 the standard Ada 95 pragma @code{Import}. It is provided for compatibility
2274 with Ada 83. The definition is upwards compatible both with pragma
2275 @code{Interface} as defined in the Ada 83 Reference Manual, and also
2276 with some extended implementations of this pragma in certain Ada 83
2279 @node Pragma Interface_Name
2280 @unnumberedsec Pragma Interface_Name
2281 @findex Interface_Name
2285 @smallexample @c ada
2286 pragma Interface_Name (
2287 [Entity =>] LOCAL_NAME
2288 [, [External_Name =>] static_string_EXPRESSION]
2289 [, [Link_Name =>] static_string_EXPRESSION]);
2293 This pragma provides an alternative way of specifying the interface name
2294 for an interfaced subprogram, and is provided for compatibility with Ada
2295 83 compilers that use the pragma for this purpose. You must provide at
2296 least one of @var{External_Name} or @var{Link_Name}.
2298 @node Pragma Interrupt_Handler
2299 @unnumberedsec Pragma Interrupt_Handler
2300 @findex Interrupt_Handler
2304 @smallexample @c ada
2305 pragma Interrupt_Handler (procedure_LOCAL_NAME);
2309 This program unit pragma is supported for parameterless protected procedures
2310 as described in Annex C of the Ada Reference Manual. On the AAMP target
2311 the pragma can also be specified for nonprotected parameterless procedures
2312 that are declared at the library level (which includes procedures
2313 declared at the top level of a library package). In the case of AAMP,
2314 when this pragma is applied to a nonprotected procedure, the instruction
2315 @code{IERET} is generated for returns from the procedure, enabling
2316 maskable interrupts, in place of the normal return instruction.
2318 @node Pragma Interrupt_State
2319 @unnumberedsec Pragma Interrupt_State
2320 @findex Interrupt_State
2324 @smallexample @c ada
2325 pragma Interrupt_State (Name => value, State => SYSTEM | RUNTIME | USER);
2329 Normally certain interrupts are reserved to the implementation. Any attempt
2330 to attach an interrupt causes Program_Error to be raised, as described in
2331 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
2332 many systems for an @kbd{Ctrl-C} interrupt. Normally this interrupt is
2333 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
2334 interrupt execution. Additionally, signals such as @code{SIGSEGV},
2335 @code{SIGABRT}, @code{SIGFPE} and @code{SIGILL} are often mapped to specific
2336 Ada exceptions, or used to implement run-time functions such as the
2337 @code{abort} statement and stack overflow checking.
2339 Pragma @code{Interrupt_State} provides a general mechanism for overriding
2340 such uses of interrupts. It subsumes the functionality of pragma
2341 @code{Unreserve_All_Interrupts}. Pragma @code{Interrupt_State} is not
2342 available on OS/2, Windows or VMS. On all other platforms than VxWorks,
2343 it applies to signals; on VxWorks, it applies to vectored hardware interrupts
2344 and may be used to mark interrupts required by the board support package
2347 Interrupts can be in one of three states:
2351 The interrupt is reserved (no Ada handler can be installed), and the
2352 Ada run-time may not install a handler. As a result you are guaranteed
2353 standard system default action if this interrupt is raised.
2357 The interrupt is reserved (no Ada handler can be installed). The run time
2358 is allowed to install a handler for internal control purposes, but is
2359 not required to do so.
2363 The interrupt is unreserved. The user may install a handler to provide
2368 These states are the allowed values of the @code{State} parameter of the
2369 pragma. The @code{Name} parameter is a value of the type
2370 @code{Ada.Interrupts.Interrupt_ID}. Typically, it is a name declared in
2371 @code{Ada.Interrupts.Names}.
2373 This is a configuration pragma, and the binder will check that there
2374 are no inconsistencies between different units in a partition in how a
2375 given interrupt is specified. It may appear anywhere a pragma is legal.
2377 The effect is to move the interrupt to the specified state.
2379 By declaring interrupts to be SYSTEM, you guarantee the standard system
2380 action, such as a core dump.
2382 By declaring interrupts to be USER, you guarantee that you can install
2385 Note that certain signals on many operating systems cannot be caught and
2386 handled by applications. In such cases, the pragma is ignored. See the
2387 operating system documentation, or the value of the array @code{Reserved}
2388 declared in the specification of package @code{System.OS_Interface}.
2390 Overriding the default state of signals used by the Ada runtime may interfere
2391 with an application's runtime behavior in the cases of the synchronous signals,
2392 and in the case of the signal used to implement the @code{abort} statement.
2394 @node Pragma Keep_Names
2395 @unnumberedsec Pragma Keep_Names
2400 @smallexample @c ada
2401 pragma Keep_Names ([On =>] enumeration_first_subtype_LOCAL_NAME);
2405 The @var{LOCAL_NAME} argument
2406 must refer to an enumeration first subtype
2407 in the current declarative part. The effect is to retain the enumeration
2408 literal names for use by @code{Image} and @code{Value} even if a global
2409 @code{Discard_Names} pragma applies. This is useful when you want to
2410 generally suppress enumeration literal names and for example you therefore
2411 use a @code{Discard_Names} pragma in the @file{gnat.adc} file, but you
2412 want to retain the names for specific enumeration types.
2414 @node Pragma License
2415 @unnumberedsec Pragma License
2417 @cindex License checking
2421 @smallexample @c ada
2422 pragma License (Unrestricted | GPL | Modified_GPL | Restricted);
2426 This pragma is provided to allow automated checking for appropriate license
2427 conditions with respect to the standard and modified GPL@. A pragma
2428 @code{License}, which is a configuration pragma that typically appears at
2429 the start of a source file or in a separate @file{gnat.adc} file, specifies
2430 the licensing conditions of a unit as follows:
2434 This is used for a unit that can be freely used with no license restrictions.
2435 Examples of such units are public domain units, and units from the Ada
2439 This is used for a unit that is licensed under the unmodified GPL, and which
2440 therefore cannot be @code{with}'ed by a restricted unit.
2443 This is used for a unit licensed under the GNAT modified GPL that includes
2444 a special exception paragraph that specifically permits the inclusion of
2445 the unit in programs without requiring the entire program to be released
2446 under the GPL@. This is the license used for the GNAT run-time which ensures
2447 that the run-time can be used freely in any program without GPL concerns.
2450 This is used for a unit that is restricted in that it is not permitted to
2451 depend on units that are licensed under the GPL@. Typical examples are
2452 proprietary code that is to be released under more restrictive license
2453 conditions. Note that restricted units are permitted to @code{with} units
2454 which are licensed under the modified GPL (this is the whole point of the
2460 Normally a unit with no @code{License} pragma is considered to have an
2461 unknown license, and no checking is done. However, standard GNAT headers
2462 are recognized, and license information is derived from them as follows.
2466 A GNAT license header starts with a line containing 78 hyphens. The following
2467 comment text is searched for the appearance of any of the following strings.
2469 If the string ``GNU General Public License'' is found, then the unit is assumed
2470 to have GPL license, unless the string ``As a special exception'' follows, in
2471 which case the license is assumed to be modified GPL@.
2473 If one of the strings
2474 ``This specification is adapted from the Ada Semantic Interface'' or
2475 ``This specification is derived from the Ada Reference Manual'' is found
2476 then the unit is assumed to be unrestricted.
2480 These default actions means that a program with a restricted license pragma
2481 will automatically get warnings if a GPL unit is inappropriately
2482 @code{with}'ed. For example, the program:
2484 @smallexample @c ada
2487 procedure Secret_Stuff is
2493 if compiled with pragma @code{License} (@code{Restricted}) in a
2494 @file{gnat.adc} file will generate the warning:
2499 >>> license of withed unit "Sem_Ch3" is incompatible
2501 2. with GNAT.Sockets;
2502 3. procedure Secret_Stuff is
2506 Here we get a warning on @code{Sem_Ch3} since it is part of the GNAT
2507 compiler and is licensed under the
2508 GPL, but no warning for @code{GNAT.Sockets} which is part of the GNAT
2509 run time, and is therefore licensed under the modified GPL@.
2511 @node Pragma Link_With
2512 @unnumberedsec Pragma Link_With
2517 @smallexample @c ada
2518 pragma Link_With (static_string_EXPRESSION @{,static_string_EXPRESSION@});
2522 This pragma is provided for compatibility with certain Ada 83 compilers.
2523 It has exactly the same effect as pragma @code{Linker_Options} except
2524 that spaces occurring within one of the string expressions are treated
2525 as separators. For example, in the following case:
2527 @smallexample @c ada
2528 pragma Link_With ("-labc -ldef");
2532 results in passing the strings @code{-labc} and @code{-ldef} as two
2533 separate arguments to the linker. In addition pragma Link_With allows
2534 multiple arguments, with the same effect as successive pragmas.
2536 @node Pragma Linker_Alias
2537 @unnumberedsec Pragma Linker_Alias
2538 @findex Linker_Alias
2542 @smallexample @c ada
2543 pragma Linker_Alias (
2544 [Entity =>] LOCAL_NAME
2545 [Alias =>] static_string_EXPRESSION);
2549 This pragma establishes a linker alias for the given named entity. For
2550 further details on the exact effect, consult the GCC manual.
2552 @node Pragma Linker_Section
2553 @unnumberedsec Pragma Linker_Section
2554 @findex Linker_Section
2558 @smallexample @c ada
2559 pragma Linker_Section (
2560 [Entity =>] LOCAL_NAME
2561 [Section =>] static_string_EXPRESSION);
2565 This pragma specifies the name of the linker section for the given entity.
2566 For further details on the exact effect, consult the GCC manual.
2568 @node Pragma Long_Float
2569 @unnumberedsec Pragma Long_Float
2575 @smallexample @c ada
2576 pragma Long_Float (FLOAT_FORMAT);
2578 FLOAT_FORMAT ::= D_Float | G_Float
2582 This pragma is implemented only in the OpenVMS implementation of GNAT@.
2583 It allows control over the internal representation chosen for the predefined
2584 type @code{Long_Float} and for floating point type representations with
2585 @code{digits} specified in the range 7 through 15.
2586 For further details on this pragma, see the
2587 @cite{DEC Ada Language Reference Manual}, section 3.5.7b. Note that to use
2588 this pragma, the standard runtime libraries must be recompiled. See the
2589 description of the @code{GNAT LIBRARY} command in the OpenVMS version
2590 of the GNAT User's Guide for details on the use of this command.
2592 @node Pragma Machine_Attribute
2593 @unnumberedsec Pragma Machine_Attribute
2594 @findex Machine_Attribute
2598 @smallexample @c ada
2599 pragma Machine_Attribute (
2600 [Attribute_Name =>] string_EXPRESSION,
2601 [Entity =>] LOCAL_NAME);
2605 Machine dependent attributes can be specified for types and/or
2606 declarations. Currently only subprogram entities are supported. This
2607 pragma is semantically equivalent to
2608 @code{__attribute__((@var{string_expression}))} in GNU C,
2609 where @code{@var{string_expression}} is
2610 recognized by the GNU C macros @code{VALID_MACHINE_TYPE_ATTRIBUTE} and
2611 @code{VALID_MACHINE_DECL_ATTRIBUTE} which are defined in the
2612 configuration header file @file{tm.h} for each machine. See the GCC
2613 manual for further information.
2615 @node Pragma Main_Storage
2616 @unnumberedsec Pragma Main_Storage
2618 @findex Main_Storage
2622 @smallexample @c ada
2624 (MAIN_STORAGE_OPTION [, MAIN_STORAGE_OPTION]);
2626 MAIN_STORAGE_OPTION ::=
2627 [WORKING_STORAGE =>] static_SIMPLE_EXPRESSION
2628 | [TOP_GUARD =>] static_SIMPLE_EXPRESSION
2633 This pragma is provided for compatibility with OpenVMS VAX Systems. It has
2634 no effect in GNAT, other than being syntax checked. Note that the pragma
2635 also has no effect in DEC Ada 83 for OpenVMS Alpha Systems.
2637 @node Pragma No_Return
2638 @unnumberedsec Pragma No_Return
2643 @smallexample @c ada
2644 pragma No_Return (procedure_LOCAL_NAME);
2648 @var{procedure_local_NAME} must refer to one or more procedure
2649 declarations in the current declarative part. A procedure to which this
2650 pragma is applied may not contain any explicit @code{return} statements,
2651 and also may not contain any implicit return statements from falling off
2652 the end of a statement sequence. One use of this pragma is to identify
2653 procedures whose only purpose is to raise an exception.
2655 Another use of this pragma is to suppress incorrect warnings about
2656 missing returns in functions, where the last statement of a function
2657 statement sequence is a call to such a procedure.
2659 @node Pragma Normalize_Scalars
2660 @unnumberedsec Pragma Normalize_Scalars
2661 @findex Normalize_Scalars
2665 @smallexample @c ada
2666 pragma Normalize_Scalars;
2670 This is a language defined pragma which is fully implemented in GNAT@. The
2671 effect is to cause all scalar objects that are not otherwise initialized
2672 to be initialized. The initial values are implementation dependent and
2676 @item Standard.Character
2678 Objects whose root type is Standard.Character are initialized to
2679 Character'Last. This will be out of range of the subtype only if
2680 the subtype range excludes this value.
2682 @item Standard.Wide_Character
2684 Objects whose root type is Standard.Wide_Character are initialized to
2685 Wide_Character'Last. This will be out of range of the subtype only if
2686 the subtype range excludes this value.
2690 Objects of an integer type are initialized to base_type'First, where
2691 base_type is the base type of the object type. This will be out of range
2692 of the subtype only if the subtype range excludes this value. For example,
2693 if you declare the subtype:
2695 @smallexample @c ada
2696 subtype Ityp is integer range 1 .. 10;
2700 then objects of type x will be initialized to Integer'First, a negative
2701 number that is certainly outside the range of subtype @code{Ityp}.
2704 Objects of all real types (fixed and floating) are initialized to
2705 base_type'First, where base_Type is the base type of the object type.
2706 This will be out of range of the subtype only if the subtype range
2707 excludes this value.
2710 Objects of a modular type are initialized to typ'Last. This will be out
2711 of range of the subtype only if the subtype excludes this value.
2713 @item Enumeration types
2714 Objects of an enumeration type are initialized to all one-bits, i.e.@: to
2715 the value @code{2 ** typ'Size - 1}. This will be out of range of the
2716 enumeration subtype in all cases except where the subtype contains
2717 exactly 2**8, 2**16, or 2**32 elements.
2721 @node Pragma Obsolescent
2722 @unnumberedsec Pragma Obsolescent
2727 @smallexample @c ada
2728 pragma Obsolescent [(static_string_EXPRESSION)];
2732 This pragma must occur immediately following a subprogram
2733 declaration. It indicates that the associated function or procedure
2734 is considered obsolescent and should not be used. Typically this is
2735 used when an API must be modified by eventually removing or modifying
2736 existing subprograms. The pragma can be used at an intermediate stage
2737 when the subprogram is still present, but will be removed later.
2739 The effect of this pragma is to output a warning message that the
2740 subprogram is obsolescent if the appropriate warning option in the
2741 compiler is activated. If a parameter is present, then a second
2742 warning message is given containing this text.
2744 @node Pragma Passive
2745 @unnumberedsec Pragma Passive
2750 @smallexample @c ada
2751 pragma Passive ([Semaphore | No]);
2755 Syntax checked, but otherwise ignored by GNAT@. This is recognized for
2756 compatibility with DEC Ada 83 implementations, where it is used within a
2757 task definition to request that a task be made passive. If the argument
2758 @code{Semaphore} is present, or the argument is omitted, then DEC Ada 83
2759 treats the pragma as an assertion that the containing task is passive
2760 and that optimization of context switch with this task is permitted and
2761 desired. If the argument @code{No} is present, the task must not be
2762 optimized. GNAT does not attempt to optimize any tasks in this manner
2763 (since protected objects are available in place of passive tasks).
2765 @node Pragma Polling
2766 @unnumberedsec Pragma Polling
2771 @smallexample @c ada
2772 pragma Polling (ON | OFF);
2776 This pragma controls the generation of polling code. This is normally off.
2777 If @code{pragma Polling (ON)} is used then periodic calls are generated to
2778 the routine @code{Ada.Exceptions.Poll}. This routine is a separate unit in the
2779 runtime library, and can be found in file @file{a-excpol.adb}.
2781 Pragma @code{Polling} can appear as a configuration pragma (for example it
2782 can be placed in the @file{gnat.adc} file) to enable polling globally, or it
2783 can be used in the statement or declaration sequence to control polling
2786 A call to the polling routine is generated at the start of every loop and
2787 at the start of every subprogram call. This guarantees that the @code{Poll}
2788 routine is called frequently, and places an upper bound (determined by
2789 the complexity of the code) on the period between two @code{Poll} calls.
2791 The primary purpose of the polling interface is to enable asynchronous
2792 aborts on targets that cannot otherwise support it (for example Windows
2793 NT), but it may be used for any other purpose requiring periodic polling.
2794 The standard version is null, and can be replaced by a user program. This
2795 will require re-compilation of the @code{Ada.Exceptions} package that can
2796 be found in files @file{a-except.ads} and @file{a-except.adb}.
2798 A standard alternative unit (in file @file{4wexcpol.adb} in the standard GNAT
2799 distribution) is used to enable the asynchronous abort capability on
2800 targets that do not normally support the capability. The version of
2801 @code{Poll} in this file makes a call to the appropriate runtime routine
2802 to test for an abort condition.
2804 Note that polling can also be enabled by use of the @code{-gnatP} switch. See
2805 the @cite{GNAT User's Guide} for details.
2807 @node Pragma Propagate_Exceptions
2808 @unnumberedsec Pragma Propagate_Exceptions
2809 @findex Propagate_Exceptions
2810 @cindex Zero Cost Exceptions
2814 @smallexample @c ada
2815 pragma Propagate_Exceptions (subprogram_LOCAL_NAME);
2819 This pragma indicates that the given entity, which is the name of an
2820 imported foreign-language subprogram may receive an Ada exception,
2821 and that the exception should be propagated. It is relevant only if
2822 zero cost exception handling is in use, and is thus never needed if
2823 the alternative @code{longjmp} / @code{setjmp} implementation of
2824 exceptions is used (although it is harmless to use it in such cases).
2826 The implementation of fast exceptions always properly propagates
2827 exceptions through Ada code, as described in the Ada Reference Manual.
2828 However, this manual is silent about the propagation of exceptions
2829 through foreign code. For example, consider the
2830 situation where @code{P1} calls
2831 @code{P2}, and @code{P2} calls @code{P3}, where
2832 @code{P1} and @code{P3} are in Ada, but @code{P2} is in C@.
2833 @code{P3} raises an Ada exception. The question is whether or not
2834 it will be propagated through @code{P2} and can be handled in
2837 For the @code{longjmp} / @code{setjmp} implementation of exceptions,
2838 the answer is always yes. For some targets on which zero cost exception
2839 handling is implemented, the answer is also always yes. However, there
2840 are some targets, notably in the current version all x86 architecture
2841 targets, in which the answer is that such propagation does not
2842 happen automatically. If such propagation is required on these
2843 targets, it is mandatory to use @code{Propagate_Exceptions} to
2844 name all foreign language routines through which Ada exceptions
2847 @node Pragma Psect_Object
2848 @unnumberedsec Pragma Psect_Object
2849 @findex Psect_Object
2853 @smallexample @c ada
2854 pragma Psect_Object (
2855 [Internal =>] LOCAL_NAME,
2856 [, [External =>] EXTERNAL_SYMBOL]
2857 [, [Size =>] EXTERNAL_SYMBOL]);
2861 | static_string_EXPRESSION
2865 This pragma is identical in effect to pragma @code{Common_Object}.
2867 @node Pragma Pure_Function
2868 @unnumberedsec Pragma Pure_Function
2869 @findex Pure_Function
2873 @smallexample @c ada
2874 pragma Pure_Function ([Entity =>] function_LOCAL_NAME);
2878 This pragma appears in the same declarative part as a function
2879 declaration (or a set of function declarations if more than one
2880 overloaded declaration exists, in which case the pragma applies
2881 to all entities). It specifies that the function @code{Entity} is
2882 to be considered pure for the purposes of code generation. This means
2883 that the compiler can assume that there are no side effects, and
2884 in particular that two calls with identical arguments produce the
2885 same result. It also means that the function can be used in an
2888 Note that, quite deliberately, there are no static checks to try
2889 to ensure that this promise is met, so @code{Pure_Function} can be used
2890 with functions that are conceptually pure, even if they do modify
2891 global variables. For example, a square root function that is
2892 instrumented to count the number of times it is called is still
2893 conceptually pure, and can still be optimized, even though it
2894 modifies a global variable (the count). Memo functions are another
2895 example (where a table of previous calls is kept and consulted to
2896 avoid re-computation).
2899 Note: Most functions in a @code{Pure} package are automatically pure, and
2900 there is no need to use pragma @code{Pure_Function} for such functions. One
2901 exception is any function that has at least one formal of type
2902 @code{System.Address} or a type derived from it. Such functions are not
2903 considered pure by default, since the compiler assumes that the
2904 @code{Address} parameter may be functioning as a pointer and that the
2905 referenced data may change even if the address value does not.
2906 Similarly, imported functions are not considered to be pure by default,
2907 since there is no way of checking that they are in fact pure. The use
2908 of pragma @code{Pure_Function} for such a function will override these default
2909 assumption, and cause the compiler to treat a designated subprogram as pure
2912 Note: If pragma @code{Pure_Function} is applied to a renamed function, it
2913 applies to the underlying renamed function. This can be used to
2914 disambiguate cases of overloading where some but not all functions
2915 in a set of overloaded functions are to be designated as pure.
2917 @node Pragma Ravenscar
2918 @unnumberedsec Pragma Ravenscar
2923 @smallexample @c ada
2928 A configuration pragma that establishes the following set of restrictions:
2931 @item No_Abort_Statements
2932 [RM D.7] There are no abort_statements, and there are
2933 no calls to Task_Identification.Abort_Task.
2935 @item No_Select_Statements
2936 There are no select_statements.
2938 @item No_Task_Hierarchy
2939 [RM D.7] All (non-environment) tasks depend
2940 directly on the environment task of the partition.
2942 @item No_Task_Allocators
2943 [RM D.7] There are no allocators for task types
2944 or types containing task subcomponents.
2946 @item No_Dynamic_Priorities
2947 [RM D.7] There are no semantic dependencies on the package Dynamic_Priorities.
2949 @item No_Terminate_Alternatives
2950 [RM D.7] There are no selective_accepts with terminate_alternatives
2952 @item No_Dynamic_Interrupts
2953 There are no semantic dependencies on Ada.Interrupts.
2955 @item No_Implicit_Heap_Allocations
2956 [RM D.7] No constructs are allowed to cause implicit heap allocation
2958 @item No_Protected_Type_Allocators
2959 There are no allocators for protected types or
2960 types containing protected subcomponents.
2962 @item No_Local_Protected_Objects
2963 Protected objects and access types that designate
2964 such objects shall be declared only at library level.
2966 @item No_Requeue_Statements
2967 Requeue statements are not allowed.
2970 There are no semantic dependencies on the package Ada.Calendar.
2972 @item No_Relative_Delay
2973 There are no delay_relative_statements.
2975 @item No_Task_Attributes_Package
2976 There are no semantic dependencies on the Ada.Task_Attributes package.
2978 @item Simple_Barriers
2979 Entry barrier condition expressions shall be either static
2980 boolean expressions or boolean objects which are declared in
2981 the protected type which contains the entry.
2983 @item Max_Asynchronous_Select_Nesting = 0
2984 [RM D.7] Specifies the maximum dynamic nesting level of asynchronous_selects.
2985 A value of zero prevents the use of any asynchronous_select.
2987 @item Max_Task_Entries = 0
2988 [RM D.7] Specifies the maximum number of entries
2989 per task. The bounds of every entry family
2990 of a task unit shall be static, or shall be
2991 defined by a discriminant of a subtype whose
2992 corresponding bound is static. A value of zero
2993 indicates that no rendezvous are possible. For
2994 the Ravenscar pragma, the value of Max_Task_Entries is always
2997 @item Max_Protected_Entries = 1
2998 [RM D.7] Specifies the maximum number of entries per
2999 protected type. The bounds of every entry family of
3000 a protected unit shall be static, or shall be defined
3001 by a discriminant of a subtype whose corresponding
3002 bound is static. For the Ravenscar pragma the value of
3003 Max_Protected_Entries is always 1.
3005 @item Max_Select_Alternatives = 0
3006 [RM D.7] Specifies the maximum number of alternatives in a selective_accept.
3007 For the Ravenscar pragma the value is always 0.
3009 @item No_Task_Termination
3010 Tasks which terminate are erroneous.
3012 @item No_Entry_Queue
3013 No task can be queued on a protected entry. Note that this restrictions is
3014 checked at run time. The violation of this restriction generates a
3015 Program_Error exception.
3019 This set of restrictions corresponds to the definition of the ``Ravenscar
3020 Profile'' for limited tasking, devised and published by the
3021 @cite{International Real-Time Ada Workshop}, 1997,
3022 and whose most recent description is available at
3023 @url{ftp://ftp.openravenscar.org/openravenscar/ravenscar00.pdf}.
3025 The above set is a superset of the restrictions provided by pragma
3026 @code{Restricted_Run_Time}, it includes five additional restrictions
3027 (@code{Simple_Barriers}, @code{No_Select_Statements},
3029 @code{No_Relative_Delay} and @code{No_Task_Termination}). This means
3030 that pragma @code{Ravenscar}, like the pragma @code{Restricted_Run_Time},
3031 automatically causes the use of a simplified, more efficient version
3032 of the tasking run-time system.
3034 @node Pragma Restricted_Run_Time
3035 @unnumberedsec Pragma Restricted_Run_Time
3036 @findex Restricted_Run_Time
3040 @smallexample @c ada
3041 pragma Restricted_Run_Time;
3045 A configuration pragma that establishes the following set of restrictions:
3048 @item No_Abort_Statements
3049 @item No_Entry_Queue
3050 @item No_Task_Hierarchy
3051 @item No_Task_Allocators
3052 @item No_Dynamic_Priorities
3053 @item No_Terminate_Alternatives
3054 @item No_Dynamic_Interrupts
3055 @item No_Protected_Type_Allocators
3056 @item No_Local_Protected_Objects
3057 @item No_Requeue_Statements
3058 @item No_Task_Attributes_Package
3059 @item Max_Asynchronous_Select_Nesting = 0
3060 @item Max_Task_Entries = 0
3061 @item Max_Protected_Entries = 1
3062 @item Max_Select_Alternatives = 0
3066 This set of restrictions causes the automatic selection of a simplified
3067 version of the run time that provides improved performance for the
3068 limited set of tasking functionality permitted by this set of restrictions.
3070 @node Pragma Restriction_Warnings
3071 @unnumberedsec Pragma Restriction_Warnings
3072 @findex Restriction_Warnings
3076 @smallexample @c ada
3077 pragma Restriction_Warnings
3078 (restriction_IDENTIFIER @{, restriction_IDENTIFIER@});
3082 This pragma allows a series of restriction identifiers to be
3083 specified (the list of allowed identifiers is the same as for
3084 pragma @code{Restrictions}). For each of these identifiers
3085 the compiler checks for violations of the restriction, but
3086 generates a warning message rather than an error message
3087 if the restriction is violated.
3089 @node Pragma Source_File_Name
3090 @unnumberedsec Pragma Source_File_Name
3091 @findex Source_File_Name
3095 @smallexample @c ada
3096 pragma Source_File_Name (
3097 [Unit_Name =>] unit_NAME,
3098 Spec_File_Name => STRING_LITERAL);
3100 pragma Source_File_Name (
3101 [Unit_Name =>] unit_NAME,
3102 Body_File_Name => STRING_LITERAL);
3106 Use this to override the normal naming convention. It is a configuration
3107 pragma, and so has the usual applicability of configuration pragmas
3108 (i.e.@: it applies to either an entire partition, or to all units in a
3109 compilation, or to a single unit, depending on how it is used.
3110 @var{unit_name} is mapped to @var{file_name_literal}. The identifier for
3111 the second argument is required, and indicates whether this is the file
3112 name for the spec or for the body.
3114 Another form of the @code{Source_File_Name} pragma allows
3115 the specification of patterns defining alternative file naming schemes
3116 to apply to all files.
3118 @smallexample @c ada
3119 pragma Source_File_Name
3120 (Spec_File_Name => STRING_LITERAL
3121 [,Casing => CASING_SPEC]
3122 [,Dot_Replacement => STRING_LITERAL]);
3124 pragma Source_File_Name
3125 (Body_File_Name => STRING_LITERAL
3126 [,Casing => CASING_SPEC]
3127 [,Dot_Replacement => STRING_LITERAL]);
3129 pragma Source_File_Name
3130 (Subunit_File_Name => STRING_LITERAL
3131 [,Casing => CASING_SPEC]
3132 [,Dot_Replacement => STRING_LITERAL]);
3134 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
3138 The first argument is a pattern that contains a single asterisk indicating
3139 the point at which the unit name is to be inserted in the pattern string
3140 to form the file name. The second argument is optional. If present it
3141 specifies the casing of the unit name in the resulting file name string.
3142 The default is lower case. Finally the third argument allows for systematic
3143 replacement of any dots in the unit name by the specified string literal.
3145 A pragma Source_File_Name cannot appear after a
3146 @ref{Pragma Source_File_Name_Project}.
3148 For more details on the use of the @code{Source_File_Name} pragma,
3149 see the sections ``Using Other File Names'' and
3150 ``Alternative File Naming Schemes'' in the @cite{GNAT User's Guide}.
3152 @node Pragma Source_File_Name_Project
3153 @unnumberedsec Pragma Source_File_Name_Project
3154 @findex Source_File_Name_Project
3157 This pragma has the same syntax and semantics as pragma Source_File_Name.
3158 It is only allowed as a stand alone configuration pragma.
3159 It cannot appear after a @ref{Pragma Source_File_Name}, and
3160 most importantly, once pragma Source_File_Name_Project appears,
3161 no further Source_File_Name pragmas are allowed.
3163 The intention is that Source_File_Name_Project pragmas are always
3164 generated by the Project Manager in a manner consistent with the naming
3165 specified in a project file, and when naming is controlled in this manner,
3166 it is not permissible to attempt to modify this naming scheme using
3167 Source_File_Name pragmas (which would not be known to the project manager).
3169 @node Pragma Source_Reference
3170 @unnumberedsec Pragma Source_Reference
3171 @findex Source_Reference
3175 @smallexample @c ada
3176 pragma Source_Reference (INTEGER_LITERAL, STRING_LITERAL);
3180 This pragma must appear as the first line of a source file.
3181 @var{integer_literal} is the logical line number of the line following
3182 the pragma line (for use in error messages and debugging
3183 information). @var{string_literal} is a static string constant that
3184 specifies the file name to be used in error messages and debugging
3185 information. This is most notably used for the output of @code{gnatchop}
3186 with the @code{-r} switch, to make sure that the original unchopped
3187 source file is the one referred to.
3189 The second argument must be a string literal, it cannot be a static
3190 string expression other than a string literal. This is because its value
3191 is needed for error messages issued by all phases of the compiler.
3193 @node Pragma Stream_Convert
3194 @unnumberedsec Pragma Stream_Convert
3195 @findex Stream_Convert
3199 @smallexample @c ada
3200 pragma Stream_Convert (
3201 [Entity =>] type_LOCAL_NAME,
3202 [Read =>] function_NAME,
3203 [Write =>] function_NAME);
3207 This pragma provides an efficient way of providing stream functions for
3208 types defined in packages. Not only is it simpler to use than declaring
3209 the necessary functions with attribute representation clauses, but more
3210 significantly, it allows the declaration to made in such a way that the
3211 stream packages are not loaded unless they are needed. The use of
3212 the Stream_Convert pragma adds no overhead at all, unless the stream
3213 attributes are actually used on the designated type.
3215 The first argument specifies the type for which stream functions are
3216 provided. The second parameter provides a function used to read values
3217 of this type. It must name a function whose argument type may be any
3218 subtype, and whose returned type must be the type given as the first
3219 argument to the pragma.
3221 The meaning of the @var{Read}
3222 parameter is that if a stream attribute directly
3223 or indirectly specifies reading of the type given as the first parameter,
3224 then a value of the type given as the argument to the Read function is
3225 read from the stream, and then the Read function is used to convert this
3226 to the required target type.
3228 Similarly the @var{Write} parameter specifies how to treat write attributes
3229 that directly or indirectly apply to the type given as the first parameter.
3230 It must have an input parameter of the type specified by the first parameter,
3231 and the return type must be the same as the input type of the Read function.
3232 The effect is to first call the Write function to convert to the given stream
3233 type, and then write the result type to the stream.
3235 The Read and Write functions must not be overloaded subprograms. If necessary
3236 renamings can be supplied to meet this requirement.
3237 The usage of this attribute is best illustrated by a simple example, taken
3238 from the GNAT implementation of package Ada.Strings.Unbounded:
3240 @smallexample @c ada
3241 function To_Unbounded (S : String)
3242 return Unbounded_String
3243 renames To_Unbounded_String;
3245 pragma Stream_Convert
3246 (Unbounded_String, To_Unbounded, To_String);
3250 The specifications of the referenced functions, as given in the Ada 95
3251 Reference Manual are:
3253 @smallexample @c ada
3254 function To_Unbounded_String (Source : String)
3255 return Unbounded_String;
3257 function To_String (Source : Unbounded_String)
3262 The effect is that if the value of an unbounded string is written to a
3263 stream, then the representation of the item in the stream is in the same
3264 format used for @code{Standard.String}, and this same representation is
3265 expected when a value of this type is read from the stream.
3267 @node Pragma Style_Checks
3268 @unnumberedsec Pragma Style_Checks
3269 @findex Style_Checks
3273 @smallexample @c ada
3274 pragma Style_Checks (string_LITERAL | ALL_CHECKS |
3275 On | Off [, LOCAL_NAME]);
3279 This pragma is used in conjunction with compiler switches to control the
3280 built in style checking provided by GNAT@. The compiler switches, if set,
3281 provide an initial setting for the switches, and this pragma may be used
3282 to modify these settings, or the settings may be provided entirely by
3283 the use of the pragma. This pragma can be used anywhere that a pragma
3284 is legal, including use as a configuration pragma (including use in
3285 the @file{gnat.adc} file).
3287 The form with a string literal specifies which style options are to be
3288 activated. These are additive, so they apply in addition to any previously
3289 set style check options. The codes for the options are the same as those
3290 used in the @code{-gnaty} switch to @code{gcc} or @code{gnatmake}.
3291 For example the following two methods can be used to enable
3296 @smallexample @c ada
3297 pragma Style_Checks ("l");
3302 gcc -c -gnatyl @dots{}
3307 The form ALL_CHECKS activates all standard checks (its use is equivalent
3308 to the use of the @code{gnaty} switch with no options. See GNAT User's
3311 The forms with @code{Off} and @code{On}
3312 can be used to temporarily disable style checks
3313 as shown in the following example:
3315 @smallexample @c ada
3319 pragma Style_Checks ("k"); -- requires keywords in lower case
3320 pragma Style_Checks (Off); -- turn off style checks
3321 NULL; -- this will not generate an error message
3322 pragma Style_Checks (On); -- turn style checks back on
3323 NULL; -- this will generate an error message
3327 Finally the two argument form is allowed only if the first argument is
3328 @code{On} or @code{Off}. The effect is to turn of semantic style checks
3329 for the specified entity, as shown in the following example:
3331 @smallexample @c ada
3335 pragma Style_Checks ("r"); -- require consistency of identifier casing
3337 Rf1 : Integer := ARG; -- incorrect, wrong case
3338 pragma Style_Checks (Off, Arg);
3339 Rf2 : Integer := ARG; -- OK, no error
3342 @node Pragma Subtitle
3343 @unnumberedsec Pragma Subtitle
3348 @smallexample @c ada
3349 pragma Subtitle ([Subtitle =>] STRING_LITERAL);
3353 This pragma is recognized for compatibility with other Ada compilers
3354 but is ignored by GNAT@.
3356 @node Pragma Suppress_All
3357 @unnumberedsec Pragma Suppress_All
3358 @findex Suppress_All
3362 @smallexample @c ada
3363 pragma Suppress_All;
3367 This pragma can only appear immediately following a compilation
3368 unit. The effect is to apply @code{Suppress (All_Checks)} to the unit
3369 which it follows. This pragma is implemented for compatibility with DEC
3370 Ada 83 usage. The use of pragma @code{Suppress (All_Checks)} as a normal
3371 configuration pragma is the preferred usage in GNAT@.
3373 @node Pragma Suppress_Exception_Locations
3374 @unnumberedsec Pragma Suppress_Exception_Locations
3375 @findex Suppress_Exception_Locations
3379 @smallexample @c ada
3380 pragma Suppress_Exception_Locations;
3384 In normal mode, a raise statement for an exception by default generates
3385 an exception message giving the file name and line number for the location
3386 of the raise. This is useful for debugging and logging purposes, but this
3387 entails extra space for the strings for the messages. The configuration
3388 pragma @code{Suppress_Exception_Locations} can be used to suppress the
3389 generation of these strings, with the result that space is saved, but the
3390 exception message for such raises is null. This configuration pragma may
3391 appear in a global configuration pragma file, or in a specific unit as
3392 usual. It is not required that this pragma be used consistently within
3393 a partition, so it is fine to have some units within a partition compiled
3394 with this pragma and others compiled in normal mode without it.
3396 @node Pragma Suppress_Initialization
3397 @unnumberedsec Pragma Suppress_Initialization
3398 @findex Suppress_Initialization
3399 @cindex Suppressing initialization
3400 @cindex Initialization, suppression of
3404 @smallexample @c ada
3405 pragma Suppress_Initialization ([Entity =>] type_Name);
3409 This pragma suppresses any implicit or explicit initialization
3410 associated with the given type name for all variables of this type.
3412 @node Pragma Task_Info
3413 @unnumberedsec Pragma Task_Info
3418 @smallexample @c ada
3419 pragma Task_Info (EXPRESSION);
3423 This pragma appears within a task definition (like pragma
3424 @code{Priority}) and applies to the task in which it appears. The
3425 argument must be of type @code{System.Task_Info.Task_Info_Type}.
3426 The @code{Task_Info} pragma provides system dependent control over
3427 aspects of tasking implementation, for example, the ability to map
3428 tasks to specific processors. For details on the facilities available
3429 for the version of GNAT that you are using, see the documentation
3430 in the specification of package System.Task_Info in the runtime
3433 @node Pragma Task_Name
3434 @unnumberedsec Pragma Task_Name
3439 @smallexample @c ada
3440 pragma Task_Name (string_EXPRESSION);
3444 This pragma appears within a task definition (like pragma
3445 @code{Priority}) and applies to the task in which it appears. The
3446 argument must be of type String, and provides a name to be used for
3447 the task instance when the task is created. Note that this expression
3448 is not required to be static, and in particular, it can contain
3449 references to task discriminants. This facility can be used to
3450 provide different names for different tasks as they are created,
3451 as illustrated in the example below.
3453 The task name is recorded internally in the run-time structures
3454 and is accessible to tools like the debugger. In addition the
3455 routine @code{Ada.Task_Identification.Image} will return this
3456 string, with a unique task address appended.
3458 @smallexample @c ada
3459 -- Example of the use of pragma Task_Name
3461 with Ada.Task_Identification;
3462 use Ada.Task_Identification;
3463 with Text_IO; use Text_IO;
3466 type Astring is access String;
3468 task type Task_Typ (Name : access String) is
3469 pragma Task_Name (Name.all);
3472 task body Task_Typ is
3473 Nam : constant String := Image (Current_Task);
3475 Put_Line ("-->" & Nam (1 .. 14) & "<--");
3478 type Ptr_Task is access Task_Typ;
3479 Task_Var : Ptr_Task;
3483 new Task_Typ (new String'("This is task 1"));
3485 new Task_Typ (new String'("This is task 2"));
3489 @node Pragma Task_Storage
3490 @unnumberedsec Pragma Task_Storage
3491 @findex Task_Storage
3494 @smallexample @c ada
3495 pragma Task_Storage (
3496 [Task_Type =>] LOCAL_NAME,
3497 [Top_Guard =>] static_integer_EXPRESSION);
3501 This pragma specifies the length of the guard area for tasks. The guard
3502 area is an additional storage area allocated to a task. A value of zero
3503 means that either no guard area is created or a minimal guard area is
3504 created, depending on the target. This pragma can appear anywhere a
3505 @code{Storage_Size} attribute definition clause is allowed for a task
3508 @node Pragma Thread_Body
3509 @unnumberedsec Pragma Thread_Body
3513 @smallexample @c ada
3514 pragma Thread_Body (
3515 [Entity =>] LOCAL_NAME,
3516 [[Secondary_Stack_Size =>] static_integer_EXPRESSION)];
3520 This pragma specifies that the subprogram whose name is given as the
3521 @code{Entity} argument is a thread body, which will be activated
3522 by being called via its Address from foreign code. The purpose is
3523 to allow execution and registration of the foreign thread within the
3524 Ada run-time system.
3526 See the library unit @code{System.Threads} for details on the expansion of
3527 a thread body subprogram, including the calls made to subprograms
3528 within System.Threads to register the task. This unit also lists the
3529 targets and runtime systems for which this pragma is supported.
3531 A thread body subprogram may not be called directly from Ada code, and
3532 it is not permitted to apply the Access (or Unrestricted_Access) attributes
3533 to such a subprogram. The only legitimate way of calling such a subprogram
3534 is to pass its Address to foreign code and then make the call from the
3537 A thread body subprogram may have any parameters, and it may be a function
3538 returning a result. The convention of the thread body subprogram may be
3539 set in the usual manner using @code{pragma Convention}.
3541 The secondary stack size parameter, if given, is used to set the size
3542 of secondary stack for the thread. The secondary stack is allocated as
3543 a local variable of the expanded thread body subprogram, and thus is
3544 allocated out of the main thread stack size. If no secondary stack
3545 size parameter is present, the default size (from the declaration in
3546 @code{System.Secondary_Stack} is used.
3548 @node Pragma Time_Slice
3549 @unnumberedsec Pragma Time_Slice
3554 @smallexample @c ada
3555 pragma Time_Slice (static_duration_EXPRESSION);
3559 For implementations of GNAT on operating systems where it is possible
3560 to supply a time slice value, this pragma may be used for this purpose.
3561 It is ignored if it is used in a system that does not allow this control,
3562 or if it appears in other than the main program unit.
3564 Note that the effect of this pragma is identical to the effect of the
3565 DEC Ada 83 pragma of the same name when operating under OpenVMS systems.
3568 @unnumberedsec Pragma Title
3573 @smallexample @c ada
3574 pragma Title (TITLING_OPTION [, TITLING OPTION]);
3577 [Title =>] STRING_LITERAL,
3578 | [Subtitle =>] STRING_LITERAL
3582 Syntax checked but otherwise ignored by GNAT@. This is a listing control
3583 pragma used in DEC Ada 83 implementations to provide a title and/or
3584 subtitle for the program listing. The program listing generated by GNAT
3585 does not have titles or subtitles.
3587 Unlike other pragmas, the full flexibility of named notation is allowed
3588 for this pragma, i.e.@: the parameters may be given in any order if named
3589 notation is used, and named and positional notation can be mixed
3590 following the normal rules for procedure calls in Ada.
3592 @node Pragma Unchecked_Union
3593 @unnumberedsec Pragma Unchecked_Union
3595 @findex Unchecked_Union
3599 @smallexample @c ada
3600 pragma Unchecked_Union (first_subtype_LOCAL_NAME);
3604 This pragma is used to declare that the specified type should be represented
3606 equivalent to a C union type, and is intended only for use in
3607 interfacing with C code that uses union types. In Ada terms, the named
3608 type must obey the following rules:
3612 It is a non-tagged non-limited record type.
3614 It has a single discrete discriminant with a default value.
3616 The component list consists of a single variant part.
3618 Each variant has a component list with a single component.
3620 No nested variants are allowed.
3622 No component has an explicit default value.
3624 No component has a non-static constraint.
3628 In addition, given a type that meets the above requirements, the
3629 following restrictions apply to its use throughout the program:
3633 The discriminant name can be mentioned only in an aggregate.
3635 No subtypes may be created of this type.
3637 The type may not be constrained by giving a discriminant value.
3639 The type cannot be passed as the actual for a generic formal with a
3644 Equality and inequality operations on @code{unchecked_unions} are not
3645 available, since there is no discriminant to compare and the compiler
3646 does not even know how many bits to compare. It is implementation
3647 dependent whether this is detected at compile time as an illegality or
3648 whether it is undetected and considered to be an erroneous construct. In
3649 GNAT, a direct comparison is illegal, but GNAT does not attempt to catch
3650 the composite case (where two composites are compared that contain an
3651 unchecked union component), so such comparisons are simply considered
3654 The layout of the resulting type corresponds exactly to a C union, where
3655 each branch of the union corresponds to a single variant in the Ada
3656 record. The semantics of the Ada program is not changed in any way by
3657 the pragma, i.e.@: provided the above restrictions are followed, and no
3658 erroneous incorrect references to fields or erroneous comparisons occur,
3659 the semantics is exactly as described by the Ada reference manual.
3660 Pragma @code{Suppress (Discriminant_Check)} applies implicitly to the
3661 type and the default convention is C.
3663 @node Pragma Unimplemented_Unit
3664 @unnumberedsec Pragma Unimplemented_Unit
3665 @findex Unimplemented_Unit
3669 @smallexample @c ada
3670 pragma Unimplemented_Unit;
3674 If this pragma occurs in a unit that is processed by the compiler, GNAT
3675 aborts with the message @samp{@var{xxx} not implemented}, where
3676 @var{xxx} is the name of the current compilation unit. This pragma is
3677 intended to allow the compiler to handle unimplemented library units in
3680 The abort only happens if code is being generated. Thus you can use
3681 specs of unimplemented packages in syntax or semantic checking mode.
3683 @node Pragma Universal_Data
3684 @unnumberedsec Pragma Universal_Data
3685 @findex Universal_Data
3689 @smallexample @c ada
3690 pragma Universal_Data [(library_unit_Name)];
3694 This pragma is supported only for the AAMP target and is ignored for
3695 other targets. The pragma specifies that all library-level objects
3696 (Counter 0 data) associated with the library unit are to be accessed
3697 and updated using universal addressing (24-bit addresses for AAMP5)
3698 rather than the default of 16-bit Data Environment (DENV) addressing.
3699 Use of this pragma will generally result in less efficient code for
3700 references to global data associated with the library unit, but
3701 allows such data to be located anywhere in memory. This pragma is
3702 a library unit pragma, but can also be used as a configuration pragma
3703 (including use in the @file{gnat.adc} file). The functionality
3704 of this pragma is also available by applying the -univ switch on the
3705 compilations of units where universal addressing of the data is desired.
3707 @node Pragma Unreferenced
3708 @unnumberedsec Pragma Unreferenced
3709 @findex Unreferenced
3710 @cindex Warnings, unreferenced
3714 @smallexample @c ada
3715 pragma Unreferenced (local_Name @{, local_Name@});
3719 This pragma signals that the entities whose names are listed are
3720 deliberately not referenced in the current source unit. This
3721 suppresses warnings about the
3722 entities being unreferenced, and in addition a warning will be
3723 generated if one of these entities is in fact referenced in the
3724 same unit as the pragma (or in the corresponding body, or one
3727 This is particularly useful for clearly signaling that a particular
3728 parameter is not referenced in some particular subprogram implementation
3729 and that this is deliberate. It can also be useful in the case of
3730 objects declared only for their initialization or finalization side
3733 If @code{local_Name} identifies more than one matching homonym in the
3734 current scope, then the entity most recently declared is the one to which
3737 The left hand side of an assignment does not count as a reference for the
3738 purpose of this pragma. Thus it is fine to assign to an entity for which
3739 pragma Unreferenced is given.
3741 @node Pragma Unreserve_All_Interrupts
3742 @unnumberedsec Pragma Unreserve_All_Interrupts
3743 @findex Unreserve_All_Interrupts
3747 @smallexample @c ada
3748 pragma Unreserve_All_Interrupts;
3752 Normally certain interrupts are reserved to the implementation. Any attempt
3753 to attach an interrupt causes Program_Error to be raised, as described in
3754 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
3755 many systems for a @kbd{Ctrl-C} interrupt. Normally this interrupt is
3756 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
3757 interrupt execution.
3759 If the pragma @code{Unreserve_All_Interrupts} appears anywhere in any unit in
3760 a program, then all such interrupts are unreserved. This allows the
3761 program to handle these interrupts, but disables their standard
3762 functions. For example, if this pragma is used, then pressing
3763 @kbd{Ctrl-C} will not automatically interrupt execution. However,
3764 a program can then handle the @code{SIGINT} interrupt as it chooses.
3766 For a full list of the interrupts handled in a specific implementation,
3767 see the source code for the specification of @code{Ada.Interrupts.Names} in
3768 file @file{a-intnam.ads}. This is a target dependent file that contains the
3769 list of interrupts recognized for a given target. The documentation in
3770 this file also specifies what interrupts are affected by the use of
3771 the @code{Unreserve_All_Interrupts} pragma.
3773 For a more general facility for controlling what interrupts can be
3774 handled, see pragma @code{Interrupt_State}, which subsumes the functionality
3775 of the @code{Unreserve_All_Interrupts} pragma.
3777 @node Pragma Unsuppress
3778 @unnumberedsec Pragma Unsuppress
3783 @smallexample @c ada
3784 pragma Unsuppress (IDENTIFIER [, [On =>] NAME]);
3788 This pragma undoes the effect of a previous pragma @code{Suppress}. If
3789 there is no corresponding pragma @code{Suppress} in effect, it has no
3790 effect. The range of the effect is the same as for pragma
3791 @code{Suppress}. The meaning of the arguments is identical to that used
3792 in pragma @code{Suppress}.
3794 One important application is to ensure that checks are on in cases where
3795 code depends on the checks for its correct functioning, so that the code
3796 will compile correctly even if the compiler switches are set to suppress
3799 @node Pragma Use_VADS_Size
3800 @unnumberedsec Pragma Use_VADS_Size
3801 @cindex @code{Size}, VADS compatibility
3802 @findex Use_VADS_Size
3806 @smallexample @c ada
3807 pragma Use_VADS_Size;
3811 This is a configuration pragma. In a unit to which it applies, any use
3812 of the 'Size attribute is automatically interpreted as a use of the
3813 'VADS_Size attribute. Note that this may result in incorrect semantic
3814 processing of valid Ada 95 programs. This is intended to aid in the
3815 handling of legacy code which depends on the interpretation of Size
3816 as implemented in the VADS compiler. See description of the VADS_Size
3817 attribute for further details.
3819 @node Pragma Validity_Checks
3820 @unnumberedsec Pragma Validity_Checks
3821 @findex Validity_Checks
3825 @smallexample @c ada
3826 pragma Validity_Checks (string_LITERAL | ALL_CHECKS | On | Off);
3830 This pragma is used in conjunction with compiler switches to control the
3831 built-in validity checking provided by GNAT@. The compiler switches, if set
3832 provide an initial setting for the switches, and this pragma may be used
3833 to modify these settings, or the settings may be provided entirely by
3834 the use of the pragma. This pragma can be used anywhere that a pragma
3835 is legal, including use as a configuration pragma (including use in
3836 the @file{gnat.adc} file).
3838 The form with a string literal specifies which validity options are to be
3839 activated. The validity checks are first set to include only the default
3840 reference manual settings, and then a string of letters in the string
3841 specifies the exact set of options required. The form of this string
3842 is exactly as described for the @code{-gnatVx} compiler switch (see the
3843 GNAT users guide for details). For example the following two methods
3844 can be used to enable validity checking for mode @code{in} and
3845 @code{in out} subprogram parameters:
3849 @smallexample @c ada
3850 pragma Validity_Checks ("im");
3855 gcc -c -gnatVim @dots{}
3860 The form ALL_CHECKS activates all standard checks (its use is equivalent
3861 to the use of the @code{gnatva} switch.
3863 The forms with @code{Off} and @code{On}
3864 can be used to temporarily disable validity checks
3865 as shown in the following example:
3867 @smallexample @c ada
3871 pragma Validity_Checks ("c"); -- validity checks for copies
3872 pragma Validity_Checks (Off); -- turn off validity checks
3873 A := B; -- B will not be validity checked
3874 pragma Validity_Checks (On); -- turn validity checks back on
3875 A := C; -- C will be validity checked
3878 @node Pragma Volatile
3879 @unnumberedsec Pragma Volatile
3884 @smallexample @c ada
3885 pragma Volatile (local_NAME);
3889 This pragma is defined by the Ada 95 Reference Manual, and the GNAT
3890 implementation is fully conformant with this definition. The reason it
3891 is mentioned in this section is that a pragma of the same name was supplied
3892 in some Ada 83 compilers, including DEC Ada 83. The Ada 95 implementation
3893 of pragma Volatile is upwards compatible with the implementation in
3896 @node Pragma Warnings
3897 @unnumberedsec Pragma Warnings
3902 @smallexample @c ada
3903 pragma Warnings (On | Off [, LOCAL_NAME]);
3907 Normally warnings are enabled, with the output being controlled by
3908 the command line switch. Warnings (@code{Off}) turns off generation of
3909 warnings until a Warnings (@code{On}) is encountered or the end of the
3910 current unit. If generation of warnings is turned off using this
3911 pragma, then no warning messages are output, regardless of the
3912 setting of the command line switches.
3914 The form with a single argument is a configuration pragma.
3916 If the @var{local_name} parameter is present, warnings are suppressed for
3917 the specified entity. This suppression is effective from the point where
3918 it occurs till the end of the extended scope of the variable (similar to
3919 the scope of @code{Suppress}).
3921 @node Pragma Weak_External
3922 @unnumberedsec Pragma Weak_External
3923 @findex Weak_External
3927 @smallexample @c ada
3928 pragma Weak_External ([Entity =>] LOCAL_NAME);
3932 This pragma specifies that the given entity should be marked as a weak
3933 external (one that does not have to be resolved) for the linker. For
3934 further details, consult the GCC manual.
3936 @node Implementation Defined Attributes
3937 @chapter Implementation Defined Attributes
3938 Ada 95 defines (throughout the Ada 95 reference manual,
3939 summarized in annex K),
3940 a set of attributes that provide useful additional functionality in all
3941 areas of the language. These language defined attributes are implemented
3942 in GNAT and work as described in the Ada 95 Reference Manual.
3944 In addition, Ada 95 allows implementations to define additional
3945 attributes whose meaning is defined by the implementation. GNAT provides
3946 a number of these implementation-dependent attributes which can be used
3947 to extend and enhance the functionality of the compiler. This section of
3948 the GNAT reference manual describes these additional attributes.
3950 Note that any program using these attributes may not be portable to
3951 other compilers (although GNAT implements this set of attributes on all
3952 platforms). Therefore if portability to other compilers is an important
3953 consideration, you should minimize the use of these attributes.
3964 * Default_Bit_Order::
3972 * Has_Discriminants::
3978 * Max_Interrupt_Priority::
3980 * Maximum_Alignment::
3984 * Passed_By_Reference::
3995 * Unconstrained_Array::
3996 * Universal_Literal_String::
3997 * Unrestricted_Access::
4005 @unnumberedsec Abort_Signal
4006 @findex Abort_Signal
4008 @code{Standard'Abort_Signal} (@code{Standard} is the only allowed
4009 prefix) provides the entity for the special exception used to signal
4010 task abort or asynchronous transfer of control. Normally this attribute
4011 should only be used in the tasking runtime (it is highly peculiar, and
4012 completely outside the normal semantics of Ada, for a user program to
4013 intercept the abort exception).
4016 @unnumberedsec Address_Size
4017 @cindex Size of @code{Address}
4018 @findex Address_Size
4020 @code{Standard'Address_Size} (@code{Standard} is the only allowed
4021 prefix) is a static constant giving the number of bits in an
4022 @code{Address}. It is the same value as System.Address'Size,
4023 but has the advantage of being static, while a direct
4024 reference to System.Address'Size is non-static because Address
4028 @unnumberedsec Asm_Input
4031 The @code{Asm_Input} attribute denotes a function that takes two
4032 parameters. The first is a string, the second is an expression of the
4033 type designated by the prefix. The first (string) argument is required
4034 to be a static expression, and is the constraint for the parameter,
4035 (e.g.@: what kind of register is required). The second argument is the
4036 value to be used as the input argument. The possible values for the
4037 constant are the same as those used in the RTL, and are dependent on
4038 the configuration file used to built the GCC back end.
4039 @ref{Machine Code Insertions}
4042 @unnumberedsec Asm_Output
4045 The @code{Asm_Output} attribute denotes a function that takes two
4046 parameters. The first is a string, the second is the name of a variable
4047 of the type designated by the attribute prefix. The first (string)
4048 argument is required to be a static expression and designates the
4049 constraint for the parameter (e.g.@: what kind of register is
4050 required). The second argument is the variable to be updated with the
4051 result. The possible values for constraint are the same as those used in
4052 the RTL, and are dependent on the configuration file used to build the
4053 GCC back end. If there are no output operands, then this argument may
4054 either be omitted, or explicitly given as @code{No_Output_Operands}.
4055 @ref{Machine Code Insertions}
4058 @unnumberedsec AST_Entry
4062 This attribute is implemented only in OpenVMS versions of GNAT@. Applied to
4063 the name of an entry, it yields a value of the predefined type AST_Handler
4064 (declared in the predefined package System, as extended by the use of
4065 pragma @code{Extend_System (Aux_DEC)}). This value enables the given entry to
4066 be called when an AST occurs. For further details, refer to the @cite{DEC Ada
4067 Language Reference Manual}, section 9.12a.
4072 @code{@var{obj}'Bit}, where @var{obj} is any object, yields the bit
4073 offset within the storage unit (byte) that contains the first bit of
4074 storage allocated for the object. The value of this attribute is of the
4075 type @code{Universal_Integer}, and is always a non-negative number not
4076 exceeding the value of @code{System.Storage_Unit}.
4078 For an object that is a variable or a constant allocated in a register,
4079 the value is zero. (The use of this attribute does not force the
4080 allocation of a variable to memory).
4082 For an object that is a formal parameter, this attribute applies
4083 to either the matching actual parameter or to a copy of the
4084 matching actual parameter.
4086 For an access object the value is zero. Note that
4087 @code{@var{obj}.all'Bit} is subject to an @code{Access_Check} for the
4088 designated object. Similarly for a record component
4089 @code{@var{X}.@var{C}'Bit} is subject to a discriminant check and
4090 @code{@var{X}(@var{I}).Bit} and @code{@var{X}(@var{I1}..@var{I2})'Bit}
4091 are subject to index checks.
4093 This attribute is designed to be compatible with the DEC Ada 83 definition
4094 and implementation of the @code{Bit} attribute.
4097 @unnumberedsec Bit_Position
4098 @findex Bit_Position
4100 @code{@var{R.C}'Bit}, where @var{R} is a record object and C is one
4101 of the fields of the record type, yields the bit
4102 offset within the record contains the first bit of
4103 storage allocated for the object. The value of this attribute is of the
4104 type @code{Universal_Integer}. The value depends only on the field
4105 @var{C} and is independent of the alignment of
4106 the containing record @var{R}.
4109 @unnumberedsec Code_Address
4110 @findex Code_Address
4111 @cindex Subprogram address
4112 @cindex Address of subprogram code
4115 attribute may be applied to subprograms in Ada 95, but the
4116 intended effect from the Ada 95 reference manual seems to be to provide
4117 an address value which can be used to call the subprogram by means of
4118 an address clause as in the following example:
4120 @smallexample @c ada
4121 procedure K is @dots{}
4124 for L'Address use K'Address;
4125 pragma Import (Ada, L);
4129 A call to @code{L} is then expected to result in a call to @code{K}@.
4130 In Ada 83, where there were no access-to-subprogram values, this was
4131 a common work around for getting the effect of an indirect call.
4132 GNAT implements the above use of @code{Address} and the technique
4133 illustrated by the example code works correctly.
4135 However, for some purposes, it is useful to have the address of the start
4136 of the generated code for the subprogram. On some architectures, this is
4137 not necessarily the same as the @code{Address} value described above.
4138 For example, the @code{Address} value may reference a subprogram
4139 descriptor rather than the subprogram itself.
4141 The @code{'Code_Address} attribute, which can only be applied to
4142 subprogram entities, always returns the address of the start of the
4143 generated code of the specified subprogram, which may or may not be
4144 the same value as is returned by the corresponding @code{'Address}
4147 @node Default_Bit_Order
4148 @unnumberedsec Default_Bit_Order
4150 @cindex Little endian
4151 @findex Default_Bit_Order
4153 @code{Standard'Default_Bit_Order} (@code{Standard} is the only
4154 permissible prefix), provides the value @code{System.Default_Bit_Order}
4155 as a @code{Pos} value (0 for @code{High_Order_First}, 1 for
4156 @code{Low_Order_First}). This is used to construct the definition of
4157 @code{Default_Bit_Order} in package @code{System}.
4160 @unnumberedsec Elaborated
4163 The prefix of the @code{'Elaborated} attribute must be a unit name. The
4164 value is a Boolean which indicates whether or not the given unit has been
4165 elaborated. This attribute is primarily intended for internal use by the
4166 generated code for dynamic elaboration checking, but it can also be used
4167 in user programs. The value will always be True once elaboration of all
4168 units has been completed. An exception is for units which need no
4169 elaboration, the value is always False for such units.
4172 @unnumberedsec Elab_Body
4175 This attribute can only be applied to a program unit name. It returns
4176 the entity for the corresponding elaboration procedure for elaborating
4177 the body of the referenced unit. This is used in the main generated
4178 elaboration procedure by the binder and is not normally used in any
4179 other context. However, there may be specialized situations in which it
4180 is useful to be able to call this elaboration procedure from Ada code,
4181 e.g.@: if it is necessary to do selective re-elaboration to fix some
4185 @unnumberedsec Elab_Spec
4188 This attribute can only be applied to a program unit name. It returns
4189 the entity for the corresponding elaboration procedure for elaborating
4190 the specification of the referenced unit. This is used in the main
4191 generated elaboration procedure by the binder and is not normally used
4192 in any other context. However, there may be specialized situations in
4193 which it is useful to be able to call this elaboration procedure from
4194 Ada code, e.g.@: if it is necessary to do selective re-elaboration to fix
4199 @cindex Ada 83 attributes
4202 The @code{Emax} attribute is provided for compatibility with Ada 83. See
4203 the Ada 83 reference manual for an exact description of the semantics of
4207 @unnumberedsec Enum_Rep
4208 @cindex Representation of enums
4211 For every enumeration subtype @var{S}, @code{@var{S}'Enum_Rep} denotes a
4212 function with the following spec:
4214 @smallexample @c ada
4215 function @var{S}'Enum_Rep (Arg : @var{S}'Base)
4216 return @i{Universal_Integer};
4220 It is also allowable to apply @code{Enum_Rep} directly to an object of an
4221 enumeration type or to a non-overloaded enumeration
4222 literal. In this case @code{@var{S}'Enum_Rep} is equivalent to
4223 @code{@var{typ}'Enum_Rep(@var{S})} where @var{typ} is the type of the
4224 enumeration literal or object.
4226 The function returns the representation value for the given enumeration
4227 value. This will be equal to value of the @code{Pos} attribute in the
4228 absence of an enumeration representation clause. This is a static
4229 attribute (i.e.@: the result is static if the argument is static).
4231 @code{@var{S}'Enum_Rep} can also be used with integer types and objects,
4232 in which case it simply returns the integer value. The reason for this
4233 is to allow it to be used for @code{(<>)} discrete formal arguments in
4234 a generic unit that can be instantiated with either enumeration types
4235 or integer types. Note that if @code{Enum_Rep} is used on a modular
4236 type whose upper bound exceeds the upper bound of the largest signed
4237 integer type, and the argument is a variable, so that the universal
4238 integer calculation is done at run-time, then the call to @code{Enum_Rep}
4239 may raise @code{Constraint_Error}.
4242 @unnumberedsec Epsilon
4243 @cindex Ada 83 attributes
4246 The @code{Epsilon} attribute is provided for compatibility with Ada 83. See
4247 the Ada 83 reference manual for an exact description of the semantics of
4251 @unnumberedsec Fixed_Value
4254 For every fixed-point type @var{S}, @code{@var{S}'Fixed_Value} denotes a
4255 function with the following specification:
4257 @smallexample @c ada
4258 function @var{S}'Fixed_Value (Arg : @i{Universal_Integer})
4263 The value returned is the fixed-point value @var{V} such that
4265 @smallexample @c ada
4266 @var{V} = Arg * @var{S}'Small
4270 The effect is thus similar to first converting the argument to the
4271 integer type used to represent @var{S}, and then doing an unchecked
4272 conversion to the fixed-point type. The difference is
4273 that there are full range checks, to ensure that the result is in range.
4274 This attribute is primarily intended for use in implementation of the
4275 input-output functions for fixed-point values.
4277 @node Has_Discriminants
4278 @unnumberedsec Has_Discriminants
4279 @cindex Discriminants, testing for
4280 @findex Has_Discriminants
4282 The prefix of the @code{Has_Discriminants} attribute is a type. The result
4283 is a Boolean value which is True if the type has discriminants, and False
4284 otherwise. The intended use of this attribute is in conjunction with generic
4285 definitions. If the attribute is applied to a generic private type, it
4286 indicates whether or not the corresponding actual type has discriminants.
4292 The @code{Img} attribute differs from @code{Image} in that it may be
4293 applied to objects as well as types, in which case it gives the
4294 @code{Image} for the subtype of the object. This is convenient for
4297 @smallexample @c ada
4298 Put_Line ("X = " & X'Img);
4302 has the same meaning as the more verbose:
4304 @smallexample @c ada
4305 Put_Line ("X = " & @var{T}'Image (X));
4309 where @var{T} is the (sub)type of the object @code{X}.
4312 @unnumberedsec Integer_Value
4313 @findex Integer_Value
4315 For every integer type @var{S}, @code{@var{S}'Integer_Value} denotes a
4316 function with the following spec:
4318 @smallexample @c ada
4319 function @var{S}'Integer_Value (Arg : @i{Universal_Fixed})
4324 The value returned is the integer value @var{V}, such that
4326 @smallexample @c ada
4327 Arg = @var{V} * @var{T}'Small
4331 where @var{T} is the type of @code{Arg}.
4332 The effect is thus similar to first doing an unchecked conversion from
4333 the fixed-point type to its corresponding implementation type, and then
4334 converting the result to the target integer type. The difference is
4335 that there are full range checks, to ensure that the result is in range.
4336 This attribute is primarily intended for use in implementation of the
4337 standard input-output functions for fixed-point values.
4340 @unnumberedsec Large
4341 @cindex Ada 83 attributes
4344 The @code{Large} attribute is provided for compatibility with Ada 83. See
4345 the Ada 83 reference manual for an exact description of the semantics of
4349 @unnumberedsec Machine_Size
4350 @findex Machine_Size
4352 This attribute is identical to the @code{Object_Size} attribute. It is
4353 provided for compatibility with the DEC Ada 83 attribute of this name.
4356 @unnumberedsec Mantissa
4357 @cindex Ada 83 attributes
4360 The @code{Mantissa} attribute is provided for compatibility with Ada 83. See
4361 the Ada 83 reference manual for an exact description of the semantics of
4364 @node Max_Interrupt_Priority
4365 @unnumberedsec Max_Interrupt_Priority
4366 @cindex Interrupt priority, maximum
4367 @findex Max_Interrupt_Priority
4369 @code{Standard'Max_Interrupt_Priority} (@code{Standard} is the only
4370 permissible prefix), provides the same value as
4371 @code{System.Max_Interrupt_Priority}.
4374 @unnumberedsec Max_Priority
4375 @cindex Priority, maximum
4376 @findex Max_Priority
4378 @code{Standard'Max_Priority} (@code{Standard} is the only permissible
4379 prefix) provides the same value as @code{System.Max_Priority}.
4381 @node Maximum_Alignment
4382 @unnumberedsec Maximum_Alignment
4383 @cindex Alignment, maximum
4384 @findex Maximum_Alignment
4386 @code{Standard'Maximum_Alignment} (@code{Standard} is the only
4387 permissible prefix) provides the maximum useful alignment value for the
4388 target. This is a static value that can be used to specify the alignment
4389 for an object, guaranteeing that it is properly aligned in all
4392 @node Mechanism_Code
4393 @unnumberedsec Mechanism_Code
4394 @cindex Return values, passing mechanism
4395 @cindex Parameters, passing mechanism
4396 @findex Mechanism_Code
4398 @code{@var{function}'Mechanism_Code} yields an integer code for the
4399 mechanism used for the result of function, and
4400 @code{@var{subprogram}'Mechanism_Code (@var{n})} yields the mechanism
4401 used for formal parameter number @var{n} (a static integer value with 1
4402 meaning the first parameter) of @var{subprogram}. The code returned is:
4410 by descriptor (default descriptor class)
4412 by descriptor (UBS: unaligned bit string)
4414 by descriptor (UBSB: aligned bit string with arbitrary bounds)
4416 by descriptor (UBA: unaligned bit array)
4418 by descriptor (S: string, also scalar access type parameter)
4420 by descriptor (SB: string with arbitrary bounds)
4422 by descriptor (A: contiguous array)
4424 by descriptor (NCA: non-contiguous array)
4428 Values from 3 through 10 are only relevant to Digital OpenVMS implementations.
4431 @node Null_Parameter
4432 @unnumberedsec Null_Parameter
4433 @cindex Zero address, passing
4434 @findex Null_Parameter
4436 A reference @code{@var{T}'Null_Parameter} denotes an imaginary object of
4437 type or subtype @var{T} allocated at machine address zero. The attribute
4438 is allowed only as the default expression of a formal parameter, or as
4439 an actual expression of a subprogram call. In either case, the
4440 subprogram must be imported.
4442 The identity of the object is represented by the address zero in the
4443 argument list, independent of the passing mechanism (explicit or
4446 This capability is needed to specify that a zero address should be
4447 passed for a record or other composite object passed by reference.
4448 There is no way of indicating this without the @code{Null_Parameter}
4452 @unnumberedsec Object_Size
4453 @cindex Size, used for objects
4456 The size of an object is not necessarily the same as the size of the type
4457 of an object. This is because by default object sizes are increased to be
4458 a multiple of the alignment of the object. For example,
4459 @code{Natural'Size} is
4460 31, but by default objects of type @code{Natural} will have a size of 32 bits.
4461 Similarly, a record containing an integer and a character:
4463 @smallexample @c ada
4471 will have a size of 40 (that is @code{Rec'Size} will be 40. The
4472 alignment will be 4, because of the
4473 integer field, and so the default size of record objects for this type
4474 will be 64 (8 bytes).
4476 The @code{@var{type}'Object_Size} attribute
4477 has been added to GNAT to allow the
4478 default object size of a type to be easily determined. For example,
4479 @code{Natural'Object_Size} is 32, and
4480 @code{Rec'Object_Size} (for the record type in the above example) will be
4481 64. Note also that, unlike the situation with the
4482 @code{Size} attribute as defined in the Ada RM, the
4483 @code{Object_Size} attribute can be specified individually
4484 for different subtypes. For example:
4486 @smallexample @c ada
4487 type R is new Integer;
4488 subtype R1 is R range 1 .. 10;
4489 subtype R2 is R range 1 .. 10;
4490 for R2'Object_Size use 8;
4494 In this example, @code{R'Object_Size} and @code{R1'Object_Size} are both
4495 32 since the default object size for a subtype is the same as the object size
4496 for the parent subtype. This means that objects of type @code{R}
4498 by default be 32 bits (four bytes). But objects of type
4499 @code{R2} will be only
4500 8 bits (one byte), since @code{R2'Object_Size} has been set to 8.
4502 @node Passed_By_Reference
4503 @unnumberedsec Passed_By_Reference
4504 @cindex Parameters, when passed by reference
4505 @findex Passed_By_Reference
4507 @code{@var{type}'Passed_By_Reference} for any subtype @var{type} returns
4508 a value of type @code{Boolean} value that is @code{True} if the type is
4509 normally passed by reference and @code{False} if the type is normally
4510 passed by copy in calls. For scalar types, the result is always @code{False}
4511 and is static. For non-scalar types, the result is non-static.
4514 @unnumberedsec Range_Length
4515 @findex Range_Length
4517 @code{@var{type}'Range_Length} for any discrete type @var{type} yields
4518 the number of values represented by the subtype (zero for a null
4519 range). The result is static for static subtypes. @code{Range_Length}
4520 applied to the index subtype of a one dimensional array always gives the
4521 same result as @code{Range} applied to the array itself.
4524 @unnumberedsec Safe_Emax
4525 @cindex Ada 83 attributes
4528 The @code{Safe_Emax} attribute is provided for compatibility with Ada 83. See
4529 the Ada 83 reference manual for an exact description of the semantics of
4533 @unnumberedsec Safe_Large
4534 @cindex Ada 83 attributes
4537 The @code{Safe_Large} attribute is provided for compatibility with Ada 83. See
4538 the Ada 83 reference manual for an exact description of the semantics of
4542 @unnumberedsec Small
4543 @cindex Ada 83 attributes
4546 The @code{Small} attribute is defined in Ada 95 only for fixed-point types.
4547 GNAT also allows this attribute to be applied to floating-point types
4548 for compatibility with Ada 83. See
4549 the Ada 83 reference manual for an exact description of the semantics of
4550 this attribute when applied to floating-point types.
4553 @unnumberedsec Storage_Unit
4554 @findex Storage_Unit
4556 @code{Standard'Storage_Unit} (@code{Standard} is the only permissible
4557 prefix) provides the same value as @code{System.Storage_Unit}.
4560 @unnumberedsec Target_Name
4563 @code{Standard'Target_Name} (@code{Standard} is the only permissible
4564 prefix) provides a static string value that identifies the target
4565 for the current compilation. For GCC implementations, this is the
4566 standard gcc target name without the terminating slash (for
4567 example, GNAT 5.0 on windows yields "i586-pc-mingw32msv").
4573 @code{Standard'Tick} (@code{Standard} is the only permissible prefix)
4574 provides the same value as @code{System.Tick},
4577 @unnumberedsec To_Address
4580 The @code{System'To_Address}
4581 (@code{System} is the only permissible prefix)
4582 denotes a function identical to
4583 @code{System.Storage_Elements.To_Address} except that
4584 it is a static attribute. This means that if its argument is
4585 a static expression, then the result of the attribute is a
4586 static expression. The result is that such an expression can be
4587 used in contexts (e.g.@: preelaborable packages) which require a
4588 static expression and where the function call could not be used
4589 (since the function call is always non-static, even if its
4590 argument is static).
4593 @unnumberedsec Type_Class
4596 @code{@var{type}'Type_Class} for any type or subtype @var{type} yields
4597 the value of the type class for the full type of @var{type}. If
4598 @var{type} is a generic formal type, the value is the value for the
4599 corresponding actual subtype. The value of this attribute is of type
4600 @code{System.Aux_DEC.Type_Class}, which has the following definition:
4602 @smallexample @c ada
4604 (Type_Class_Enumeration,
4606 Type_Class_Fixed_Point,
4607 Type_Class_Floating_Point,
4612 Type_Class_Address);
4616 Protected types yield the value @code{Type_Class_Task}, which thus
4617 applies to all concurrent types. This attribute is designed to
4618 be compatible with the DEC Ada 83 attribute of the same name.
4621 @unnumberedsec UET_Address
4624 The @code{UET_Address} attribute can only be used for a prefix which
4625 denotes a library package. It yields the address of the unit exception
4626 table when zero cost exception handling is used. This attribute is
4627 intended only for use within the GNAT implementation. See the unit
4628 @code{Ada.Exceptions} in files @file{a-except.ads} and @file{a-except.adb}
4629 for details on how this attribute is used in the implementation.
4631 @node Unconstrained_Array
4632 @unnumberedsec Unconstrained_Array
4633 @findex Unconstrained_Array
4635 The @code{Unconstrained_Array} attribute can be used with a prefix that
4636 denotes any type or subtype. It is a static attribute that yields
4637 @code{True} if the prefix designates an unconstrained array,
4638 and @code{False} otherwise. In a generic instance, the result is
4639 still static, and yields the result of applying this test to the
4642 @node Universal_Literal_String
4643 @unnumberedsec Universal_Literal_String
4644 @cindex Named numbers, representation of
4645 @findex Universal_Literal_String
4647 The prefix of @code{Universal_Literal_String} must be a named
4648 number. The static result is the string consisting of the characters of
4649 the number as defined in the original source. This allows the user
4650 program to access the actual text of named numbers without intermediate
4651 conversions and without the need to enclose the strings in quotes (which
4652 would preclude their use as numbers). This is used internally for the
4653 construction of values of the floating-point attributes from the file
4654 @file{ttypef.ads}, but may also be used by user programs.
4656 @node Unrestricted_Access
4657 @unnumberedsec Unrestricted_Access
4658 @cindex @code{Access}, unrestricted
4659 @findex Unrestricted_Access
4661 The @code{Unrestricted_Access} attribute is similar to @code{Access}
4662 except that all accessibility and aliased view checks are omitted. This
4663 is a user-beware attribute. It is similar to
4664 @code{Address}, for which it is a desirable replacement where the value
4665 desired is an access type. In other words, its effect is identical to
4666 first applying the @code{Address} attribute and then doing an unchecked
4667 conversion to a desired access type. In GNAT, but not necessarily in
4668 other implementations, the use of static chains for inner level
4669 subprograms means that @code{Unrestricted_Access} applied to a
4670 subprogram yields a value that can be called as long as the subprogram
4671 is in scope (normal Ada 95 accessibility rules restrict this usage).
4673 It is possible to use @code{Unrestricted_Access} for any type, but care
4674 must be excercised if it is used to create pointers to unconstrained
4675 objects. In this case, the resulting pointer has the same scope as the
4676 context of the attribute, and may not be returned to some enclosing
4677 scope. For instance, a function cannot use @code{Unrestricted_Access}
4678 to create a unconstrained pointer and then return that value to the
4682 @unnumberedsec VADS_Size
4683 @cindex @code{Size}, VADS compatibility
4686 The @code{'VADS_Size} attribute is intended to make it easier to port
4687 legacy code which relies on the semantics of @code{'Size} as implemented
4688 by the VADS Ada 83 compiler. GNAT makes a best effort at duplicating the
4689 same semantic interpretation. In particular, @code{'VADS_Size} applied
4690 to a predefined or other primitive type with no Size clause yields the
4691 Object_Size (for example, @code{Natural'Size} is 32 rather than 31 on
4692 typical machines). In addition @code{'VADS_Size} applied to an object
4693 gives the result that would be obtained by applying the attribute to
4694 the corresponding type.
4697 @unnumberedsec Value_Size
4698 @cindex @code{Size}, setting for not-first subtype
4700 @code{@var{type}'Value_Size} is the number of bits required to represent
4701 a value of the given subtype. It is the same as @code{@var{type}'Size},
4702 but, unlike @code{Size}, may be set for non-first subtypes.
4705 @unnumberedsec Wchar_T_Size
4706 @findex Wchar_T_Size
4707 @code{Standard'Wchar_T_Size} (@code{Standard} is the only permissible
4708 prefix) provides the size in bits of the C @code{wchar_t} type
4709 primarily for constructing the definition of this type in
4710 package @code{Interfaces.C}.
4713 @unnumberedsec Word_Size
4715 @code{Standard'Word_Size} (@code{Standard} is the only permissible
4716 prefix) provides the value @code{System.Word_Size}.
4718 @c ------------------------
4719 @node Implementation Advice
4720 @chapter Implementation Advice
4722 The main text of the Ada 95 Reference Manual describes the required
4723 behavior of all Ada 95 compilers, and the GNAT compiler conforms to
4726 In addition, there are sections throughout the Ada 95
4727 reference manual headed
4728 by the phrase ``implementation advice''. These sections are not normative,
4729 i.e.@: they do not specify requirements that all compilers must
4730 follow. Rather they provide advice on generally desirable behavior. You
4731 may wonder why they are not requirements. The most typical answer is
4732 that they describe behavior that seems generally desirable, but cannot
4733 be provided on all systems, or which may be undesirable on some systems.
4735 As far as practical, GNAT follows the implementation advice sections in
4736 the Ada 95 Reference Manual. This chapter contains a table giving the
4737 reference manual section number, paragraph number and several keywords
4738 for each advice. Each entry consists of the text of the advice followed
4739 by the GNAT interpretation of this advice. Most often, this simply says
4740 ``followed'', which means that GNAT follows the advice. However, in a
4741 number of cases, GNAT deliberately deviates from this advice, in which
4742 case the text describes what GNAT does and why.
4744 @cindex Error detection
4745 @unnumberedsec 1.1.3(20): Error Detection
4748 If an implementation detects the use of an unsupported Specialized Needs
4749 Annex feature at run time, it should raise @code{Program_Error} if
4752 Not relevant. All specialized needs annex features are either supported,
4753 or diagnosed at compile time.
4756 @unnumberedsec 1.1.3(31): Child Units
4759 If an implementation wishes to provide implementation-defined
4760 extensions to the functionality of a language-defined library unit, it
4761 should normally do so by adding children to the library unit.
4765 @cindex Bounded errors
4766 @unnumberedsec 1.1.5(12): Bounded Errors
4769 If an implementation detects a bounded error or erroneous
4770 execution, it should raise @code{Program_Error}.
4772 Followed in all cases in which the implementation detects a bounded
4773 error or erroneous execution. Not all such situations are detected at
4777 @unnumberedsec 2.8(16): Pragmas
4780 Normally, implementation-defined pragmas should have no semantic effect
4781 for error-free programs; that is, if the implementation-defined pragmas
4782 are removed from a working program, the program should still be legal,
4783 and should still have the same semantics.
4785 The following implementation defined pragmas are exceptions to this
4797 @item CPP_Constructor
4805 @item Interface_Name
4807 @item Machine_Attribute
4809 @item Unimplemented_Unit
4811 @item Unchecked_Union
4816 In each of the above cases, it is essential to the purpose of the pragma
4817 that this advice not be followed. For details see the separate section
4818 on implementation defined pragmas.
4820 @unnumberedsec 2.8(17-19): Pragmas
4823 Normally, an implementation should not define pragmas that can
4824 make an illegal program legal, except as follows:
4828 A pragma used to complete a declaration, such as a pragma @code{Import};
4832 A pragma used to configure the environment by adding, removing, or
4833 replacing @code{library_items}.
4835 See response to paragraph 16 of this same section.
4837 @cindex Character Sets
4838 @cindex Alternative Character Sets
4839 @unnumberedsec 3.5.2(5): Alternative Character Sets
4842 If an implementation supports a mode with alternative interpretations
4843 for @code{Character} and @code{Wide_Character}, the set of graphic
4844 characters of @code{Character} should nevertheless remain a proper
4845 subset of the set of graphic characters of @code{Wide_Character}. Any
4846 character set ``localizations'' should be reflected in the results of
4847 the subprograms defined in the language-defined package
4848 @code{Characters.Handling} (see A.3) available in such a mode. In a mode with
4849 an alternative interpretation of @code{Character}, the implementation should
4850 also support a corresponding change in what is a legal
4851 @code{identifier_letter}.
4853 Not all wide character modes follow this advice, in particular the JIS
4854 and IEC modes reflect standard usage in Japan, and in these encoding,
4855 the upper half of the Latin-1 set is not part of the wide-character
4856 subset, since the most significant bit is used for wide character
4857 encoding. However, this only applies to the external forms. Internally
4858 there is no such restriction.
4860 @cindex Integer types
4861 @unnumberedsec 3.5.4(28): Integer Types
4865 An implementation should support @code{Long_Integer} in addition to
4866 @code{Integer} if the target machine supports 32-bit (or longer)
4867 arithmetic. No other named integer subtypes are recommended for package
4868 @code{Standard}. Instead, appropriate named integer subtypes should be
4869 provided in the library package @code{Interfaces} (see B.2).
4871 @code{Long_Integer} is supported. Other standard integer types are supported
4872 so this advice is not fully followed. These types
4873 are supported for convenient interface to C, and so that all hardware
4874 types of the machine are easily available.
4875 @unnumberedsec 3.5.4(29): Integer Types
4879 An implementation for a two's complement machine should support
4880 modular types with a binary modulus up to @code{System.Max_Int*2+2}. An
4881 implementation should support a non-binary modules up to @code{Integer'Last}.
4885 @cindex Enumeration values
4886 @unnumberedsec 3.5.5(8): Enumeration Values
4889 For the evaluation of a call on @code{@var{S}'Pos} for an enumeration
4890 subtype, if the value of the operand does not correspond to the internal
4891 code for any enumeration literal of its type (perhaps due to an
4892 un-initialized variable), then the implementation should raise
4893 @code{Program_Error}. This is particularly important for enumeration
4894 types with noncontiguous internal codes specified by an
4895 enumeration_representation_clause.
4900 @unnumberedsec 3.5.7(17): Float Types
4903 An implementation should support @code{Long_Float} in addition to
4904 @code{Float} if the target machine supports 11 or more digits of
4905 precision. No other named floating point subtypes are recommended for
4906 package @code{Standard}. Instead, appropriate named floating point subtypes
4907 should be provided in the library package @code{Interfaces} (see B.2).
4909 @code{Short_Float} and @code{Long_Long_Float} are also provided. The
4910 former provides improved compatibility with other implementations
4911 supporting this type. The latter corresponds to the highest precision
4912 floating-point type supported by the hardware. On most machines, this
4913 will be the same as @code{Long_Float}, but on some machines, it will
4914 correspond to the IEEE extended form. The notable case is all ia32
4915 (x86) implementations, where @code{Long_Long_Float} corresponds to
4916 the 80-bit extended precision format supported in hardware on this
4917 processor. Note that the 128-bit format on SPARC is not supported,
4918 since this is a software rather than a hardware format.
4920 @cindex Multidimensional arrays
4921 @cindex Arrays, multidimensional
4922 @unnumberedsec 3.6.2(11): Multidimensional Arrays
4925 An implementation should normally represent multidimensional arrays in
4926 row-major order, consistent with the notation used for multidimensional
4927 array aggregates (see 4.3.3). However, if a pragma @code{Convention}
4928 (@code{Fortran}, @dots{}) applies to a multidimensional array type, then
4929 column-major order should be used instead (see B.5, ``Interfacing with
4934 @findex Duration'Small
4935 @unnumberedsec 9.6(30-31): Duration'Small
4938 Whenever possible in an implementation, the value of @code{Duration'Small}
4939 should be no greater than 100 microseconds.
4941 Followed. (@code{Duration'Small} = 10**(@minus{}9)).
4945 The time base for @code{delay_relative_statements} should be monotonic;
4946 it need not be the same time base as used for @code{Calendar.Clock}.
4950 @unnumberedsec 10.2.1(12): Consistent Representation
4953 In an implementation, a type declared in a pre-elaborated package should
4954 have the same representation in every elaboration of a given version of
4955 the package, whether the elaborations occur in distinct executions of
4956 the same program, or in executions of distinct programs or partitions
4957 that include the given version.
4959 Followed, except in the case of tagged types. Tagged types involve
4960 implicit pointers to a local copy of a dispatch table, and these pointers
4961 have representations which thus depend on a particular elaboration of the
4962 package. It is not easy to see how it would be possible to follow this
4963 advice without severely impacting efficiency of execution.
4965 @cindex Exception information
4966 @unnumberedsec 11.4.1(19): Exception Information
4969 @code{Exception_Message} by default and @code{Exception_Information}
4970 should produce information useful for
4971 debugging. @code{Exception_Message} should be short, about one
4972 line. @code{Exception_Information} can be long. @code{Exception_Message}
4973 should not include the
4974 @code{Exception_Name}. @code{Exception_Information} should include both
4975 the @code{Exception_Name} and the @code{Exception_Message}.
4977 Followed. For each exception that doesn't have a specified
4978 @code{Exception_Message}, the compiler generates one containing the location
4979 of the raise statement. This location has the form ``file:line'', where
4980 file is the short file name (without path information) and line is the line
4981 number in the file. Note that in the case of the Zero Cost Exception
4982 mechanism, these messages become redundant with the Exception_Information that
4983 contains a full backtrace of the calling sequence, so they are disabled.
4984 To disable explicitly the generation of the source location message, use the
4985 Pragma @code{Discard_Names}.
4987 @cindex Suppression of checks
4988 @cindex Checks, suppression of
4989 @unnumberedsec 11.5(28): Suppression of Checks
4992 The implementation should minimize the code executed for checks that
4993 have been suppressed.
4997 @cindex Representation clauses
4998 @unnumberedsec 13.1 (21-24): Representation Clauses
5001 The recommended level of support for all representation items is
5002 qualified as follows:
5006 An implementation need not support representation items containing
5007 non-static expressions, except that an implementation should support a
5008 representation item for a given entity if each non-static expression in
5009 the representation item is a name that statically denotes a constant
5010 declared before the entity.
5012 Followed. GNAT does not support non-static expressions in representation
5013 clauses unless they are constants declared before the entity. For
5016 @smallexample @c ada
5018 for X'Address use To_address (16#2000#);
5022 will be rejected, since the To_Address expression is non-static. Instead
5025 @smallexample @c ada
5026 X_Address : constant Address : = To_Address (16#2000#);
5028 for X'Address use X_Address;
5033 An implementation need not support a specification for the @code{Size}
5034 for a given composite subtype, nor the size or storage place for an
5035 object (including a component) of a given composite subtype, unless the
5036 constraints on the subtype and its composite subcomponents (if any) are
5037 all static constraints.
5039 Followed. Size Clauses are not permitted on non-static components, as
5044 An aliased component, or a component whose type is by-reference, should
5045 always be allocated at an addressable location.
5049 @cindex Packed types
5050 @unnumberedsec 13.2(6-8): Packed Types
5053 If a type is packed, then the implementation should try to minimize
5054 storage allocated to objects of the type, possibly at the expense of
5055 speed of accessing components, subject to reasonable complexity in
5056 addressing calculations.
5060 The recommended level of support pragma @code{Pack} is:
5062 For a packed record type, the components should be packed as tightly as
5063 possible subject to the Sizes of the component subtypes, and subject to
5064 any @code{record_representation_clause} that applies to the type; the
5065 implementation may, but need not, reorder components or cross aligned
5066 word boundaries to improve the packing. A component whose @code{Size} is
5067 greater than the word size may be allocated an integral number of words.
5069 Followed. Tight packing of arrays is supported for all component sizes
5070 up to 64-bits. If the array component size is 1 (that is to say, if
5071 the component is a boolean type or an enumeration type with two values)
5072 then values of the type are implicitly initialized to zero. This
5073 happens both for objects of the packed type, and for objects that have a
5074 subcomponent of the packed type.
5078 An implementation should support Address clauses for imported
5082 @cindex @code{Address} clauses
5083 @unnumberedsec 13.3(14-19): Address Clauses
5087 For an array @var{X}, @code{@var{X}'Address} should point at the first
5088 component of the array, and not at the array bounds.
5094 The recommended level of support for the @code{Address} attribute is:
5096 @code{@var{X}'Address} should produce a useful result if @var{X} is an
5097 object that is aliased or of a by-reference type, or is an entity whose
5098 @code{Address} has been specified.
5100 Followed. A valid address will be produced even if none of those
5101 conditions have been met. If necessary, the object is forced into
5102 memory to ensure the address is valid.
5106 An implementation should support @code{Address} clauses for imported
5113 Objects (including subcomponents) that are aliased or of a by-reference
5114 type should be allocated on storage element boundaries.
5120 If the @code{Address} of an object is specified, or it is imported or exported,
5121 then the implementation should not perform optimizations based on
5122 assumptions of no aliases.
5126 @cindex @code{Alignment} clauses
5127 @unnumberedsec 13.3(29-35): Alignment Clauses
5130 The recommended level of support for the @code{Alignment} attribute for
5133 An implementation should support specified Alignments that are factors
5134 and multiples of the number of storage elements per word, subject to the
5141 An implementation need not support specified @code{Alignment}s for
5142 combinations of @code{Size}s and @code{Alignment}s that cannot be easily
5143 loaded and stored by available machine instructions.
5149 An implementation need not support specified @code{Alignment}s that are
5150 greater than the maximum @code{Alignment} the implementation ever returns by
5157 The recommended level of support for the @code{Alignment} attribute for
5160 Same as above, for subtypes, but in addition:
5166 For stand-alone library-level objects of statically constrained
5167 subtypes, the implementation should support all @code{Alignment}s
5168 supported by the target linker. For example, page alignment is likely to
5169 be supported for such objects, but not for subtypes.
5173 @cindex @code{Size} clauses
5174 @unnumberedsec 13.3(42-43): Size Clauses
5177 The recommended level of support for the @code{Size} attribute of
5180 A @code{Size} clause should be supported for an object if the specified
5181 @code{Size} is at least as large as its subtype's @code{Size}, and
5182 corresponds to a size in storage elements that is a multiple of the
5183 object's @code{Alignment} (if the @code{Alignment} is nonzero).
5187 @unnumberedsec 13.3(50-56): Size Clauses
5190 If the @code{Size} of a subtype is specified, and allows for efficient
5191 independent addressability (see 9.10) on the target architecture, then
5192 the @code{Size} of the following objects of the subtype should equal the
5193 @code{Size} of the subtype:
5195 Aliased objects (including components).
5201 @code{Size} clause on a composite subtype should not affect the
5202 internal layout of components.
5208 The recommended level of support for the @code{Size} attribute of subtypes is:
5212 The @code{Size} (if not specified) of a static discrete or fixed point
5213 subtype should be the number of bits needed to represent each value
5214 belonging to the subtype using an unbiased representation, leaving space
5215 for a sign bit only if the subtype contains negative values. If such a
5216 subtype is a first subtype, then an implementation should support a
5217 specified @code{Size} for it that reflects this representation.
5223 For a subtype implemented with levels of indirection, the @code{Size}
5224 should include the size of the pointers, but not the size of what they
5229 @cindex @code{Component_Size} clauses
5230 @unnumberedsec 13.3(71-73): Component Size Clauses
5233 The recommended level of support for the @code{Component_Size}
5238 An implementation need not support specified @code{Component_Sizes} that are
5239 less than the @code{Size} of the component subtype.
5245 An implementation should support specified @code{Component_Size}s that
5246 are factors and multiples of the word size. For such
5247 @code{Component_Size}s, the array should contain no gaps between
5248 components. For other @code{Component_Size}s (if supported), the array
5249 should contain no gaps between components when packing is also
5250 specified; the implementation should forbid this combination in cases
5251 where it cannot support a no-gaps representation.
5255 @cindex Enumeration representation clauses
5256 @cindex Representation clauses, enumeration
5257 @unnumberedsec 13.4(9-10): Enumeration Representation Clauses
5260 The recommended level of support for enumeration representation clauses
5263 An implementation need not support enumeration representation clauses
5264 for boolean types, but should at minimum support the internal codes in
5265 the range @code{System.Min_Int.System.Max_Int}.
5269 @cindex Record representation clauses
5270 @cindex Representation clauses, records
5271 @unnumberedsec 13.5.1(17-22): Record Representation Clauses
5274 The recommended level of support for
5275 @*@code{record_representation_clauses} is:
5277 An implementation should support storage places that can be extracted
5278 with a load, mask, shift sequence of machine code, and set with a load,
5279 shift, mask, store sequence, given the available machine instructions
5286 A storage place should be supported if its size is equal to the
5287 @code{Size} of the component subtype, and it starts and ends on a
5288 boundary that obeys the @code{Alignment} of the component subtype.
5294 If the default bit ordering applies to the declaration of a given type,
5295 then for a component whose subtype's @code{Size} is less than the word
5296 size, any storage place that does not cross an aligned word boundary
5297 should be supported.
5303 An implementation may reserve a storage place for the tag field of a
5304 tagged type, and disallow other components from overlapping that place.
5306 Followed. The storage place for the tag field is the beginning of the tagged
5307 record, and its size is Address'Size. GNAT will reject an explicit component
5308 clause for the tag field.
5312 An implementation need not support a @code{component_clause} for a
5313 component of an extension part if the storage place is not after the
5314 storage places of all components of the parent type, whether or not
5315 those storage places had been specified.
5317 Followed. The above advice on record representation clauses is followed,
5318 and all mentioned features are implemented.
5320 @cindex Storage place attributes
5321 @unnumberedsec 13.5.2(5): Storage Place Attributes
5324 If a component is represented using some form of pointer (such as an
5325 offset) to the actual data of the component, and this data is contiguous
5326 with the rest of the object, then the storage place attributes should
5327 reflect the place of the actual data, not the pointer. If a component is
5328 allocated discontinuously from the rest of the object, then a warning
5329 should be generated upon reference to one of its storage place
5332 Followed. There are no such components in GNAT@.
5334 @cindex Bit ordering
5335 @unnumberedsec 13.5.3(7-8): Bit Ordering
5338 The recommended level of support for the non-default bit ordering is:
5342 If @code{Word_Size} = @code{Storage_Unit}, then the implementation
5343 should support the non-default bit ordering in addition to the default
5346 Followed. Word size does not equal storage size in this implementation.
5347 Thus non-default bit ordering is not supported.
5349 @cindex @code{Address}, as private type
5350 @unnumberedsec 13.7(37): Address as Private
5353 @code{Address} should be of a private type.
5357 @cindex Operations, on @code{Address}
5358 @cindex @code{Address}, operations of
5359 @unnumberedsec 13.7.1(16): Address Operations
5362 Operations in @code{System} and its children should reflect the target
5363 environment semantics as closely as is reasonable. For example, on most
5364 machines, it makes sense for address arithmetic to ``wrap around''.
5365 Operations that do not make sense should raise @code{Program_Error}.
5367 Followed. Address arithmetic is modular arithmetic that wraps around. No
5368 operation raises @code{Program_Error}, since all operations make sense.
5370 @cindex Unchecked conversion
5371 @unnumberedsec 13.9(14-17): Unchecked Conversion
5374 The @code{Size} of an array object should not include its bounds; hence,
5375 the bounds should not be part of the converted data.
5381 The implementation should not generate unnecessary run-time checks to
5382 ensure that the representation of @var{S} is a representation of the
5383 target type. It should take advantage of the permission to return by
5384 reference when possible. Restrictions on unchecked conversions should be
5385 avoided unless required by the target environment.
5387 Followed. There are no restrictions on unchecked conversion. A warning is
5388 generated if the source and target types do not have the same size since
5389 the semantics in this case may be target dependent.
5393 The recommended level of support for unchecked conversions is:
5397 Unchecked conversions should be supported and should be reversible in
5398 the cases where this clause defines the result. To enable meaningful use
5399 of unchecked conversion, a contiguous representation should be used for
5400 elementary subtypes, for statically constrained array subtypes whose
5401 component subtype is one of the subtypes described in this paragraph,
5402 and for record subtypes without discriminants whose component subtypes
5403 are described in this paragraph.
5407 @cindex Heap usage, implicit
5408 @unnumberedsec 13.11(23-25): Implicit Heap Usage
5411 An implementation should document any cases in which it dynamically
5412 allocates heap storage for a purpose other than the evaluation of an
5415 Followed, the only other points at which heap storage is dynamically
5416 allocated are as follows:
5420 At initial elaboration time, to allocate dynamically sized global
5424 To allocate space for a task when a task is created.
5427 To extend the secondary stack dynamically when needed. The secondary
5428 stack is used for returning variable length results.
5433 A default (implementation-provided) storage pool for an
5434 access-to-constant type should not have overhead to support deallocation of
5441 A storage pool for an anonymous access type should be created at the
5442 point of an allocator for the type, and be reclaimed when the designated
5443 object becomes inaccessible.
5447 @cindex Unchecked deallocation
5448 @unnumberedsec 13.11.2(17): Unchecked De-allocation
5451 For a standard storage pool, @code{Free} should actually reclaim the
5456 @cindex Stream oriented attributes
5457 @unnumberedsec 13.13.2(17): Stream Oriented Attributes
5460 If a stream element is the same size as a storage element, then the
5461 normal in-memory representation should be used by @code{Read} and
5462 @code{Write} for scalar objects. Otherwise, @code{Read} and @code{Write}
5463 should use the smallest number of stream elements needed to represent
5464 all values in the base range of the scalar type.
5467 Followed. By default, GNAT uses the interpretation suggested by AI-195,
5468 which specifies using the size of the first subtype.
5469 However, such an implementation is based on direct binary
5470 representations and is therefore target- and endianness-dependent.
5471 To address this issue, GNAT also supplies an alternate implementation
5472 of the stream attributes @code{Read} and @code{Write},
5473 which uses the target-independent XDR standard representation
5475 @cindex XDR representation
5476 @cindex @code{Read} attribute
5477 @cindex @code{Write} attribute
5478 @cindex Stream oriented attributes
5479 The XDR implementation is provided as an alternative body of the
5480 @code{System.Stream_Attributes} package, in the file
5481 @file{s-strxdr.adb} in the GNAT library.
5482 There is no @file{s-strxdr.ads} file.
5483 In order to install the XDR implementation, do the following:
5485 @item Replace the default implementation of the
5486 @code{System.Stream_Attributes} package with the XDR implementation.
5487 For example on a Unix platform issue the commands:
5489 $ mv s-stratt.adb s-strold.adb
5490 $ mv s-strxdr.adb s-stratt.adb
5494 Rebuild the GNAT run-time library as documented in the
5495 @cite{GNAT User's Guide}
5498 @unnumberedsec A.1(52): Names of Predefined Numeric Types
5501 If an implementation provides additional named predefined integer types,
5502 then the names should end with @samp{Integer} as in
5503 @samp{Long_Integer}. If an implementation provides additional named
5504 predefined floating point types, then the names should end with
5505 @samp{Float} as in @samp{Long_Float}.
5509 @findex Ada.Characters.Handling
5510 @unnumberedsec A.3.2(49): @code{Ada.Characters.Handling}
5513 If an implementation provides a localized definition of @code{Character}
5514 or @code{Wide_Character}, then the effects of the subprograms in
5515 @code{Characters.Handling} should reflect the localizations. See also
5518 Followed. GNAT provides no such localized definitions.
5520 @cindex Bounded-length strings
5521 @unnumberedsec A.4.4(106): Bounded-Length String Handling
5524 Bounded string objects should not be implemented by implicit pointers
5525 and dynamic allocation.
5527 Followed. No implicit pointers or dynamic allocation are used.
5529 @cindex Random number generation
5530 @unnumberedsec A.5.2(46-47): Random Number Generation
5533 Any storage associated with an object of type @code{Generator} should be
5534 reclaimed on exit from the scope of the object.
5540 If the generator period is sufficiently long in relation to the number
5541 of distinct initiator values, then each possible value of
5542 @code{Initiator} passed to @code{Reset} should initiate a sequence of
5543 random numbers that does not, in a practical sense, overlap the sequence
5544 initiated by any other value. If this is not possible, then the mapping
5545 between initiator values and generator states should be a rapidly
5546 varying function of the initiator value.
5548 Followed. The generator period is sufficiently long for the first
5549 condition here to hold true.
5551 @findex Get_Immediate
5552 @unnumberedsec A.10.7(23): @code{Get_Immediate}
5555 The @code{Get_Immediate} procedures should be implemented with
5556 unbuffered input. For a device such as a keyboard, input should be
5557 @dfn{available} if a key has already been typed, whereas for a disk
5558 file, input should always be available except at end of file. For a file
5559 associated with a keyboard-like device, any line-editing features of the
5560 underlying operating system should be disabled during the execution of
5561 @code{Get_Immediate}.
5563 Followed on all targets except VxWorks. For VxWorks, there is no way to
5564 provide this functionality that does not result in the input buffer being
5565 flushed before the @code{Get_Immediate} call. A special unit
5566 @code{Interfaces.Vxworks.IO} is provided that contains routines to enable
5570 @unnumberedsec B.1(39-41): Pragma @code{Export}
5573 If an implementation supports pragma @code{Export} to a given language,
5574 then it should also allow the main subprogram to be written in that
5575 language. It should support some mechanism for invoking the elaboration
5576 of the Ada library units included in the system, and for invoking the
5577 finalization of the environment task. On typical systems, the
5578 recommended mechanism is to provide two subprograms whose link names are
5579 @code{adainit} and @code{adafinal}. @code{adainit} should contain the
5580 elaboration code for library units. @code{adafinal} should contain the
5581 finalization code. These subprograms should have no effect the second
5582 and subsequent time they are called.
5588 Automatic elaboration of pre-elaborated packages should be
5589 provided when pragma @code{Export} is supported.
5591 Followed when the main program is in Ada. If the main program is in a
5592 foreign language, then
5593 @code{adainit} must be called to elaborate pre-elaborated
5598 For each supported convention @var{L} other than @code{Intrinsic}, an
5599 implementation should support @code{Import} and @code{Export} pragmas
5600 for objects of @var{L}-compatible types and for subprograms, and pragma
5601 @code{Convention} for @var{L}-eligible types and for subprograms,
5602 presuming the other language has corresponding features. Pragma
5603 @code{Convention} need not be supported for scalar types.
5607 @cindex Package @code{Interfaces}
5609 @unnumberedsec B.2(12-13): Package @code{Interfaces}
5612 For each implementation-defined convention identifier, there should be a
5613 child package of package Interfaces with the corresponding name. This
5614 package should contain any declarations that would be useful for
5615 interfacing to the language (implementation) represented by the
5616 convention. Any declarations useful for interfacing to any language on
5617 the given hardware architecture should be provided directly in
5620 Followed. An additional package not defined
5621 in the Ada 95 Reference Manual is @code{Interfaces.CPP}, used
5622 for interfacing to C++.
5626 An implementation supporting an interface to C, COBOL, or Fortran should
5627 provide the corresponding package or packages described in the following
5630 Followed. GNAT provides all the packages described in this section.
5632 @cindex C, interfacing with
5633 @unnumberedsec B.3(63-71): Interfacing with C
5636 An implementation should support the following interface correspondences
5643 An Ada procedure corresponds to a void-returning C function.
5649 An Ada function corresponds to a non-void C function.
5655 An Ada @code{in} scalar parameter is passed as a scalar argument to a C
5662 An Ada @code{in} parameter of an access-to-object type with designated
5663 type @var{T} is passed as a @code{@var{t}*} argument to a C function,
5664 where @var{t} is the C type corresponding to the Ada type @var{T}.
5670 An Ada access @var{T} parameter, or an Ada @code{out} or @code{in out}
5671 parameter of an elementary type @var{T}, is passed as a @code{@var{t}*}
5672 argument to a C function, where @var{t} is the C type corresponding to
5673 the Ada type @var{T}. In the case of an elementary @code{out} or
5674 @code{in out} parameter, a pointer to a temporary copy is used to
5675 preserve by-copy semantics.
5681 An Ada parameter of a record type @var{T}, of any mode, is passed as a
5682 @code{@var{t}*} argument to a C function, where @var{t} is the C
5683 structure corresponding to the Ada type @var{T}.
5685 Followed. This convention may be overridden by the use of the C_Pass_By_Copy
5686 pragma, or Convention, or by explicitly specifying the mechanism for a given
5687 call using an extended import or export pragma.
5691 An Ada parameter of an array type with component type @var{T}, of any
5692 mode, is passed as a @code{@var{t}*} argument to a C function, where
5693 @var{t} is the C type corresponding to the Ada type @var{T}.
5699 An Ada parameter of an access-to-subprogram type is passed as a pointer
5700 to a C function whose prototype corresponds to the designated
5701 subprogram's specification.
5705 @cindex COBOL, interfacing with
5706 @unnumberedsec B.4(95-98): Interfacing with COBOL
5709 An Ada implementation should support the following interface
5710 correspondences between Ada and COBOL@.
5716 An Ada access @var{T} parameter is passed as a @samp{BY REFERENCE} data item of
5717 the COBOL type corresponding to @var{T}.
5723 An Ada in scalar parameter is passed as a @samp{BY CONTENT} data item of
5724 the corresponding COBOL type.
5730 Any other Ada parameter is passed as a @samp{BY REFERENCE} data item of the
5731 COBOL type corresponding to the Ada parameter type; for scalars, a local
5732 copy is used if necessary to ensure by-copy semantics.
5736 @cindex Fortran, interfacing with
5737 @unnumberedsec B.5(22-26): Interfacing with Fortran
5740 An Ada implementation should support the following interface
5741 correspondences between Ada and Fortran:
5747 An Ada procedure corresponds to a Fortran subroutine.
5753 An Ada function corresponds to a Fortran function.
5759 An Ada parameter of an elementary, array, or record type @var{T} is
5760 passed as a @var{T} argument to a Fortran procedure, where @var{T} is
5761 the Fortran type corresponding to the Ada type @var{T}, and where the
5762 INTENT attribute of the corresponding dummy argument matches the Ada
5763 formal parameter mode; the Fortran implementation's parameter passing
5764 conventions are used. For elementary types, a local copy is used if
5765 necessary to ensure by-copy semantics.
5771 An Ada parameter of an access-to-subprogram type is passed as a
5772 reference to a Fortran procedure whose interface corresponds to the
5773 designated subprogram's specification.
5777 @cindex Machine operations
5778 @unnumberedsec C.1(3-5): Access to Machine Operations
5781 The machine code or intrinsic support should allow access to all
5782 operations normally available to assembly language programmers for the
5783 target environment, including privileged instructions, if any.
5789 The interfacing pragmas (see Annex B) should support interface to
5790 assembler; the default assembler should be associated with the
5791 convention identifier @code{Assembler}.
5797 If an entity is exported to assembly language, then the implementation
5798 should allocate it at an addressable location, and should ensure that it
5799 is retained by the linking process, even if not otherwise referenced
5800 from the Ada code. The implementation should assume that any call to a
5801 machine code or assembler subprogram is allowed to read or update every
5802 object that is specified as exported.
5806 @unnumberedsec C.1(10-16): Access to Machine Operations
5809 The implementation should ensure that little or no overhead is
5810 associated with calling intrinsic and machine-code subprograms.
5812 Followed for both intrinsics and machine-code subprograms.
5816 It is recommended that intrinsic subprograms be provided for convenient
5817 access to any machine operations that provide special capabilities or
5818 efficiency and that are not otherwise available through the language
5821 Followed. A full set of machine operation intrinsic subprograms is provided.
5825 Atomic read-modify-write operations---e.g.@:, test and set, compare and
5826 swap, decrement and test, enqueue/dequeue.
5828 Followed on any target supporting such operations.
5832 Standard numeric functions---e.g.@:, sin, log.
5834 Followed on any target supporting such operations.
5838 String manipulation operations---e.g.@:, translate and test.
5840 Followed on any target supporting such operations.
5844 Vector operations---e.g.@:, compare vector against thresholds.
5846 Followed on any target supporting such operations.
5850 Direct operations on I/O ports.
5852 Followed on any target supporting such operations.
5854 @cindex Interrupt support
5855 @unnumberedsec C.3(28): Interrupt Support
5858 If the @code{Ceiling_Locking} policy is not in effect, the
5859 implementation should provide means for the application to specify which
5860 interrupts are to be blocked during protected actions, if the underlying
5861 system allows for a finer-grain control of interrupt blocking.
5863 Followed. The underlying system does not allow for finer-grain control
5864 of interrupt blocking.
5866 @cindex Protected procedure handlers
5867 @unnumberedsec C.3.1(20-21): Protected Procedure Handlers
5870 Whenever possible, the implementation should allow interrupt handlers to
5871 be called directly by the hardware.
5875 This is never possible under IRIX, so this is followed by default.
5877 Followed on any target where the underlying operating system permits
5882 Whenever practical, violations of any
5883 implementation-defined restrictions should be detected before run time.
5885 Followed. Compile time warnings are given when possible.
5887 @cindex Package @code{Interrupts}
5889 @unnumberedsec C.3.2(25): Package @code{Interrupts}
5893 If implementation-defined forms of interrupt handler procedures are
5894 supported, such as protected procedures with parameters, then for each
5895 such form of a handler, a type analogous to @code{Parameterless_Handler}
5896 should be specified in a child package of @code{Interrupts}, with the
5897 same operations as in the predefined package Interrupts.
5901 @cindex Pre-elaboration requirements
5902 @unnumberedsec C.4(14): Pre-elaboration Requirements
5905 It is recommended that pre-elaborated packages be implemented in such a
5906 way that there should be little or no code executed at run time for the
5907 elaboration of entities not already covered by the Implementation
5910 Followed. Executable code is generated in some cases, e.g.@: loops
5911 to initialize large arrays.
5913 @unnumberedsec C.5(8): Pragma @code{Discard_Names}
5917 If the pragma applies to an entity, then the implementation should
5918 reduce the amount of storage used for storing names associated with that
5923 @cindex Package @code{Task_Attributes}
5924 @findex Task_Attributes
5925 @unnumberedsec C.7.2(30): The Package Task_Attributes
5928 Some implementations are targeted to domains in which memory use at run
5929 time must be completely deterministic. For such implementations, it is
5930 recommended that the storage for task attributes will be pre-allocated
5931 statically and not from the heap. This can be accomplished by either
5932 placing restrictions on the number and the size of the task's
5933 attributes, or by using the pre-allocated storage for the first @var{N}
5934 attribute objects, and the heap for the others. In the latter case,
5935 @var{N} should be documented.
5937 Not followed. This implementation is not targeted to such a domain.
5939 @cindex Locking Policies
5940 @unnumberedsec D.3(17): Locking Policies
5944 The implementation should use names that end with @samp{_Locking} for
5945 locking policies defined by the implementation.
5947 Followed. A single implementation-defined locking policy is defined,
5948 whose name (@code{Inheritance_Locking}) follows this suggestion.
5950 @cindex Entry queuing policies
5951 @unnumberedsec D.4(16): Entry Queuing Policies
5954 Names that end with @samp{_Queuing} should be used
5955 for all implementation-defined queuing policies.
5957 Followed. No such implementation-defined queuing policies exist.
5959 @cindex Preemptive abort
5960 @unnumberedsec D.6(9-10): Preemptive Abort
5963 Even though the @code{abort_statement} is included in the list of
5964 potentially blocking operations (see 9.5.1), it is recommended that this
5965 statement be implemented in a way that never requires the task executing
5966 the @code{abort_statement} to block.
5972 On a multi-processor, the delay associated with aborting a task on
5973 another processor should be bounded; the implementation should use
5974 periodic polling, if necessary, to achieve this.
5978 @cindex Tasking restrictions
5979 @unnumberedsec D.7(21): Tasking Restrictions
5982 When feasible, the implementation should take advantage of the specified
5983 restrictions to produce a more efficient implementation.
5985 GNAT currently takes advantage of these restrictions by providing an optimized
5986 run time when the Ravenscar profile and the GNAT restricted run time set
5987 of restrictions are specified. See pragma @code{Ravenscar} and pragma
5988 @code{Restricted_Run_Time} for more details.
5990 @cindex Time, monotonic
5991 @unnumberedsec D.8(47-49): Monotonic Time
5994 When appropriate, implementations should provide configuration
5995 mechanisms to change the value of @code{Tick}.
5997 Such configuration mechanisms are not appropriate to this implementation
5998 and are thus not supported.
6002 It is recommended that @code{Calendar.Clock} and @code{Real_Time.Clock}
6003 be implemented as transformations of the same time base.
6009 It is recommended that the @dfn{best} time base which exists in
6010 the underlying system be available to the application through
6011 @code{Clock}. @dfn{Best} may mean highest accuracy or largest range.
6015 @cindex Partition communication subsystem
6017 @unnumberedsec E.5(28-29): Partition Communication Subsystem
6020 Whenever possible, the PCS on the called partition should allow for
6021 multiple tasks to call the RPC-receiver with different messages and
6022 should allow them to block until the corresponding subprogram body
6025 Followed by GLADE, a separately supplied PCS that can be used with
6030 The @code{Write} operation on a stream of type @code{Params_Stream_Type}
6031 should raise @code{Storage_Error} if it runs out of space trying to
6032 write the @code{Item} into the stream.
6034 Followed by GLADE, a separately supplied PCS that can be used with
6037 @cindex COBOL support
6038 @unnumberedsec F(7): COBOL Support
6041 If COBOL (respectively, C) is widely supported in the target
6042 environment, implementations supporting the Information Systems Annex
6043 should provide the child package @code{Interfaces.COBOL} (respectively,
6044 @code{Interfaces.C}) specified in Annex B and should support a
6045 @code{convention_identifier} of COBOL (respectively, C) in the interfacing
6046 pragmas (see Annex B), thus allowing Ada programs to interface with
6047 programs written in that language.
6051 @cindex Decimal radix support
6052 @unnumberedsec F.1(2): Decimal Radix Support
6055 Packed decimal should be used as the internal representation for objects
6056 of subtype @var{S} when @var{S}'Machine_Radix = 10.
6058 Not followed. GNAT ignores @var{S}'Machine_Radix and always uses binary
6062 @unnumberedsec G: Numerics
6065 If Fortran (respectively, C) is widely supported in the target
6066 environment, implementations supporting the Numerics Annex
6067 should provide the child package @code{Interfaces.Fortran} (respectively,
6068 @code{Interfaces.C}) specified in Annex B and should support a
6069 @code{convention_identifier} of Fortran (respectively, C) in the interfacing
6070 pragmas (see Annex B), thus allowing Ada programs to interface with
6071 programs written in that language.
6075 @cindex Complex types
6076 @unnumberedsec G.1.1(56-58): Complex Types
6079 Because the usual mathematical meaning of multiplication of a complex
6080 operand and a real operand is that of the scaling of both components of
6081 the former by the latter, an implementation should not perform this
6082 operation by first promoting the real operand to complex type and then
6083 performing a full complex multiplication. In systems that, in the
6084 future, support an Ada binding to IEC 559:1989, the latter technique
6085 will not generate the required result when one of the components of the
6086 complex operand is infinite. (Explicit multiplication of the infinite
6087 component by the zero component obtained during promotion yields a NaN
6088 that propagates into the final result.) Analogous advice applies in the
6089 case of multiplication of a complex operand and a pure-imaginary
6090 operand, and in the case of division of a complex operand by a real or
6091 pure-imaginary operand.
6097 Similarly, because the usual mathematical meaning of addition of a
6098 complex operand and a real operand is that the imaginary operand remains
6099 unchanged, an implementation should not perform this operation by first
6100 promoting the real operand to complex type and then performing a full
6101 complex addition. In implementations in which the @code{Signed_Zeros}
6102 attribute of the component type is @code{True} (and which therefore
6103 conform to IEC 559:1989 in regard to the handling of the sign of zero in
6104 predefined arithmetic operations), the latter technique will not
6105 generate the required result when the imaginary component of the complex
6106 operand is a negatively signed zero. (Explicit addition of the negative
6107 zero to the zero obtained during promotion yields a positive zero.)
6108 Analogous advice applies in the case of addition of a complex operand
6109 and a pure-imaginary operand, and in the case of subtraction of a
6110 complex operand and a real or pure-imaginary operand.
6116 Implementations in which @code{Real'Signed_Zeros} is @code{True} should
6117 attempt to provide a rational treatment of the signs of zero results and
6118 result components. As one example, the result of the @code{Argument}
6119 function should have the sign of the imaginary component of the
6120 parameter @code{X} when the point represented by that parameter lies on
6121 the positive real axis; as another, the sign of the imaginary component
6122 of the @code{Compose_From_Polar} function should be the same as
6123 (respectively, the opposite of) that of the @code{Argument} parameter when that
6124 parameter has a value of zero and the @code{Modulus} parameter has a
6125 nonnegative (respectively, negative) value.
6129 @cindex Complex elementary functions
6130 @unnumberedsec G.1.2(49): Complex Elementary Functions
6133 Implementations in which @code{Complex_Types.Real'Signed_Zeros} is
6134 @code{True} should attempt to provide a rational treatment of the signs
6135 of zero results and result components. For example, many of the complex
6136 elementary functions have components that are odd functions of one of
6137 the parameter components; in these cases, the result component should
6138 have the sign of the parameter component at the origin. Other complex
6139 elementary functions have zero components whose sign is opposite that of
6140 a parameter component at the origin, or is always positive or always
6145 @cindex Accuracy requirements
6146 @unnumberedsec G.2.4(19): Accuracy Requirements
6149 The versions of the forward trigonometric functions without a
6150 @code{Cycle} parameter should not be implemented by calling the
6151 corresponding version with a @code{Cycle} parameter of
6152 @code{2.0*Numerics.Pi}, since this will not provide the required
6153 accuracy in some portions of the domain. For the same reason, the
6154 version of @code{Log} without a @code{Base} parameter should not be
6155 implemented by calling the corresponding version with a @code{Base}
6156 parameter of @code{Numerics.e}.
6160 @cindex Complex arithmetic accuracy
6161 @cindex Accuracy, complex arithmetic
6162 @unnumberedsec G.2.6(15): Complex Arithmetic Accuracy
6166 The version of the @code{Compose_From_Polar} function without a
6167 @code{Cycle} parameter should not be implemented by calling the
6168 corresponding version with a @code{Cycle} parameter of
6169 @code{2.0*Numerics.Pi}, since this will not provide the required
6170 accuracy in some portions of the domain.
6174 @c -----------------------------------------
6175 @node Implementation Defined Characteristics
6176 @chapter Implementation Defined Characteristics
6179 In addition to the implementation dependent pragmas and attributes, and
6180 the implementation advice, there are a number of other features of Ada
6181 95 that are potentially implementation dependent. These are mentioned
6182 throughout the Ada 95 Reference Manual, and are summarized in annex M@.
6184 A requirement for conforming Ada compilers is that they provide
6185 documentation describing how the implementation deals with each of these
6186 issues. In this chapter, you will find each point in annex M listed
6187 followed by a description in italic font of how GNAT
6191 implementation on IRIX 5.3 operating system or greater
6193 handles the implementation dependence.
6195 You can use this chapter as a guide to minimizing implementation
6196 dependent features in your programs if portability to other compilers
6197 and other operating systems is an important consideration. The numbers
6198 in each section below correspond to the paragraph number in the Ada 95
6204 @strong{2}. Whether or not each recommendation given in Implementation
6205 Advice is followed. See 1.1.2(37).
6208 @xref{Implementation Advice}.
6213 @strong{3}. Capacity limitations of the implementation. See 1.1.3(3).
6216 The complexity of programs that can be processed is limited only by the
6217 total amount of available virtual memory, and disk space for the
6218 generated object files.
6223 @strong{4}. Variations from the standard that are impractical to avoid
6224 given the implementation's execution environment. See 1.1.3(6).
6227 There are no variations from the standard.
6232 @strong{5}. Which @code{code_statement}s cause external
6233 interactions. See 1.1.3(10).
6236 Any @code{code_statement} can potentially cause external interactions.
6241 @strong{6}. The coded representation for the text of an Ada
6242 program. See 2.1(4).
6245 See separate section on source representation.
6250 @strong{7}. The control functions allowed in comments. See 2.1(14).
6253 See separate section on source representation.
6258 @strong{8}. The representation for an end of line. See 2.2(2).
6261 See separate section on source representation.
6266 @strong{9}. Maximum supported line length and lexical element
6267 length. See 2.2(15).
6270 The maximum line length is 255 characters an the maximum length of a
6271 lexical element is also 255 characters.
6276 @strong{10}. Implementation defined pragmas. See 2.8(14).
6280 @xref{Implementation Defined Pragmas}.
6285 @strong{11}. Effect of pragma @code{Optimize}. See 2.8(27).
6288 Pragma @code{Optimize}, if given with a @code{Time} or @code{Space}
6289 parameter, checks that the optimization flag is set, and aborts if it is
6295 @strong{12}. The sequence of characters of the value returned by
6296 @code{@var{S}'Image} when some of the graphic characters of
6297 @code{@var{S}'Wide_Image} are not defined in @code{Character}. See
6301 The sequence of characters is as defined by the wide character encoding
6302 method used for the source. See section on source representation for
6308 @strong{13}. The predefined integer types declared in
6309 @code{Standard}. See 3.5.4(25).
6313 @item Short_Short_Integer
6316 (Short) 16 bit signed
6320 64 bit signed (Alpha OpenVMS only)
6321 32 bit signed (all other targets)
6322 @item Long_Long_Integer
6329 @strong{14}. Any nonstandard integer types and the operators defined
6330 for them. See 3.5.4(26).
6333 There are no nonstandard integer types.
6338 @strong{15}. Any nonstandard real types and the operators defined for
6342 There are no nonstandard real types.
6347 @strong{16}. What combinations of requested decimal precision and range
6348 are supported for floating point types. See 3.5.7(7).
6351 The precision and range is as defined by the IEEE standard.
6356 @strong{17}. The predefined floating point types declared in
6357 @code{Standard}. See 3.5.7(16).
6364 (Short) 32 bit IEEE short
6367 @item Long_Long_Float
6368 64 bit IEEE long (80 bit IEEE long on x86 processors)
6374 @strong{18}. The small of an ordinary fixed point type. See 3.5.9(8).
6377 @code{Fine_Delta} is 2**(@minus{}63)
6382 @strong{19}. What combinations of small, range, and digits are
6383 supported for fixed point types. See 3.5.9(10).
6386 Any combinations are permitted that do not result in a small less than
6387 @code{Fine_Delta} and do not result in a mantissa larger than 63 bits.
6388 If the mantissa is larger than 53 bits on machines where Long_Long_Float
6389 is 64 bits (true of all architectures except ia32), then the output from
6390 Text_IO is accurate to only 53 bits, rather than the full mantissa. This
6391 is because floating-point conversions are used to convert fixed point.
6396 @strong{20}. The result of @code{Tags.Expanded_Name} for types declared
6397 within an unnamed @code{block_statement}. See 3.9(10).
6400 Block numbers of the form @code{B@var{nnn}}, where @var{nnn} is a
6401 decimal integer are allocated.
6406 @strong{21}. Implementation-defined attributes. See 4.1.4(12).
6409 @xref{Implementation Defined Attributes}.
6414 @strong{22}. Any implementation-defined time types. See 9.6(6).
6417 There are no implementation-defined time types.
6422 @strong{23}. The time base associated with relative delays.
6425 See 9.6(20). The time base used is that provided by the C library
6426 function @code{gettimeofday}.
6431 @strong{24}. The time base of the type @code{Calendar.Time}. See
6435 The time base used is that provided by the C library function
6436 @code{gettimeofday}.
6441 @strong{25}. The time zone used for package @code{Calendar}
6442 operations. See 9.6(24).
6445 The time zone used by package @code{Calendar} is the current system time zone
6446 setting for local time, as accessed by the C library function
6452 @strong{26}. Any limit on @code{delay_until_statements} of
6453 @code{select_statements}. See 9.6(29).
6456 There are no such limits.
6461 @strong{27}. Whether or not two non overlapping parts of a composite
6462 object are independently addressable, in the case where packing, record
6463 layout, or @code{Component_Size} is specified for the object. See
6467 Separate components are independently addressable if they do not share
6468 overlapping storage units.
6473 @strong{28}. The representation for a compilation. See 10.1(2).
6476 A compilation is represented by a sequence of files presented to the
6477 compiler in a single invocation of the @code{gcc} command.
6482 @strong{29}. Any restrictions on compilations that contain multiple
6483 compilation_units. See 10.1(4).
6486 No single file can contain more than one compilation unit, but any
6487 sequence of files can be presented to the compiler as a single
6493 @strong{30}. The mechanisms for creating an environment and for adding
6494 and replacing compilation units. See 10.1.4(3).
6497 See separate section on compilation model.
6502 @strong{31}. The manner of explicitly assigning library units to a
6503 partition. See 10.2(2).
6506 If a unit contains an Ada main program, then the Ada units for the partition
6507 are determined by recursive application of the rules in the Ada Reference
6508 Manual section 10.2(2-6). In other words, the Ada units will be those that
6509 are needed by the main program, and then this definition of need is applied
6510 recursively to those units, and the partition contains the transitive
6511 closure determined by this relationship. In short, all the necessary units
6512 are included, with no need to explicitly specify the list. If additional
6513 units are required, e.g.@: by foreign language units, then all units must be
6514 mentioned in the context clause of one of the needed Ada units.
6516 If the partition contains no main program, or if the main program is in
6517 a language other than Ada, then GNAT
6518 provides the binder options @code{-z} and @code{-n} respectively, and in
6519 this case a list of units can be explicitly supplied to the binder for
6520 inclusion in the partition (all units needed by these units will also
6521 be included automatically). For full details on the use of these
6522 options, refer to the @cite{GNAT User's Guide} sections on Binding
6528 @strong{32}. The implementation-defined means, if any, of specifying
6529 which compilation units are needed by a given compilation unit. See
6533 The units needed by a given compilation unit are as defined in
6534 the Ada Reference Manual section 10.2(2-6). There are no
6535 implementation-defined pragmas or other implementation-defined
6536 means for specifying needed units.
6541 @strong{33}. The manner of designating the main subprogram of a
6542 partition. See 10.2(7).
6545 The main program is designated by providing the name of the
6546 corresponding @file{ALI} file as the input parameter to the binder.
6551 @strong{34}. The order of elaboration of @code{library_items}. See
6555 The first constraint on ordering is that it meets the requirements of
6556 chapter 10 of the Ada 95 Reference Manual. This still leaves some
6557 implementation dependent choices, which are resolved by first
6558 elaborating bodies as early as possible (i.e.@: in preference to specs
6559 where there is a choice), and second by evaluating the immediate with
6560 clauses of a unit to determine the probably best choice, and
6561 third by elaborating in alphabetical order of unit names
6562 where a choice still remains.
6567 @strong{35}. Parameter passing and function return for the main
6568 subprogram. See 10.2(21).
6571 The main program has no parameters. It may be a procedure, or a function
6572 returning an integer type. In the latter case, the returned integer
6573 value is the return code of the program (overriding any value that
6574 may have been set by a call to @code{Ada.Command_Line.Set_Exit_Status}).
6579 @strong{36}. The mechanisms for building and running partitions. See
6583 GNAT itself supports programs with only a single partition. The GNATDIST
6584 tool provided with the GLADE package (which also includes an implementation
6585 of the PCS) provides a completely flexible method for building and running
6586 programs consisting of multiple partitions. See the separate GLADE manual
6592 @strong{37}. The details of program execution, including program
6593 termination. See 10.2(25).
6596 See separate section on compilation model.
6601 @strong{38}. The semantics of any non-active partitions supported by the
6602 implementation. See 10.2(28).
6605 Passive partitions are supported on targets where shared memory is
6606 provided by the operating system. See the GLADE reference manual for
6612 @strong{39}. The information returned by @code{Exception_Message}. See
6616 Exception message returns the null string unless a specific message has
6617 been passed by the program.
6622 @strong{40}. The result of @code{Exceptions.Exception_Name} for types
6623 declared within an unnamed @code{block_statement}. See 11.4.1(12).
6626 Blocks have implementation defined names of the form @code{B@var{nnn}}
6627 where @var{nnn} is an integer.
6632 @strong{41}. The information returned by
6633 @code{Exception_Information}. See 11.4.1(13).
6636 @code{Exception_Information} returns a string in the following format:
6639 @emph{Exception_Name:} nnnnn
6640 @emph{Message:} mmmmm
6642 @emph{Call stack traceback locations:}
6643 0xhhhh 0xhhhh 0xhhhh ... 0xhhh
6651 @code{nnnn} is the fully qualified name of the exception in all upper
6652 case letters. This line is always present.
6655 @code{mmmm} is the message (this line present only if message is non-null)
6658 @code{ppp} is the Process Id value as a decimal integer (this line is
6659 present only if the Process Id is non-zero). Currently we are
6660 not making use of this field.
6663 The Call stack traceback locations line and the following values
6664 are present only if at least one traceback location was recorded.
6665 The values are given in C style format, with lower case letters
6666 for a-f, and only as many digits present as are necessary.
6670 The line terminator sequence at the end of each line, including
6671 the last line is a single @code{LF} character (@code{16#0A#}).
6676 @strong{42}. Implementation-defined check names. See 11.5(27).
6679 No implementation-defined check names are supported.
6684 @strong{43}. The interpretation of each aspect of representation. See
6688 See separate section on data representations.
6693 @strong{44}. Any restrictions placed upon representation items. See
6697 See separate section on data representations.
6702 @strong{45}. The meaning of @code{Size} for indefinite subtypes. See
6706 Size for an indefinite subtype is the maximum possible size, except that
6707 for the case of a subprogram parameter, the size of the parameter object
6713 @strong{46}. The default external representation for a type tag. See
6717 The default external representation for a type tag is the fully expanded
6718 name of the type in upper case letters.
6723 @strong{47}. What determines whether a compilation unit is the same in
6724 two different partitions. See 13.3(76).
6727 A compilation unit is the same in two different partitions if and only
6728 if it derives from the same source file.
6733 @strong{48}. Implementation-defined components. See 13.5.1(15).
6736 The only implementation defined component is the tag for a tagged type,
6737 which contains a pointer to the dispatching table.
6742 @strong{49}. If @code{Word_Size} = @code{Storage_Unit}, the default bit
6743 ordering. See 13.5.3(5).
6746 @code{Word_Size} (32) is not the same as @code{Storage_Unit} (8) for this
6747 implementation, so no non-default bit ordering is supported. The default
6748 bit ordering corresponds to the natural endianness of the target architecture.
6753 @strong{50}. The contents of the visible part of package @code{System}
6754 and its language-defined children. See 13.7(2).
6757 See the definition of these packages in files @file{system.ads} and
6758 @file{s-stoele.ads}.
6763 @strong{51}. The contents of the visible part of package
6764 @code{System.Machine_Code}, and the meaning of
6765 @code{code_statements}. See 13.8(7).
6768 See the definition and documentation in file @file{s-maccod.ads}.
6773 @strong{52}. The effect of unchecked conversion. See 13.9(11).
6776 Unchecked conversion between types of the same size
6777 and results in an uninterpreted transmission of the bits from one type
6778 to the other. If the types are of unequal sizes, then in the case of
6779 discrete types, a shorter source is first zero or sign extended as
6780 necessary, and a shorter target is simply truncated on the left.
6781 For all non-discrete types, the source is first copied if necessary
6782 to ensure that the alignment requirements of the target are met, then
6783 a pointer is constructed to the source value, and the result is obtained
6784 by dereferencing this pointer after converting it to be a pointer to the
6790 @strong{53}. The manner of choosing a storage pool for an access type
6791 when @code{Storage_Pool} is not specified for the type. See 13.11(17).
6794 There are 3 different standard pools used by the compiler when
6795 @code{Storage_Pool} is not specified depending whether the type is local
6796 to a subprogram or defined at the library level and whether
6797 @code{Storage_Size}is specified or not. See documentation in the runtime
6798 library units @code{System.Pool_Global}, @code{System.Pool_Size} and
6799 @code{System.Pool_Local} in files @file{s-poosiz.ads},
6800 @file{s-pooglo.ads} and @file{s-pooloc.ads} for full details on the
6806 @strong{54}. Whether or not the implementation provides user-accessible
6807 names for the standard pool type(s). See 13.11(17).
6811 See documentation in the sources of the run time mentioned in paragraph
6812 @strong{53} . All these pools are accessible by means of @code{with}'ing
6818 @strong{55}. The meaning of @code{Storage_Size}. See 13.11(18).
6821 @code{Storage_Size} is measured in storage units, and refers to the
6822 total space available for an access type collection, or to the primary
6823 stack space for a task.
6828 @strong{56}. Implementation-defined aspects of storage pools. See
6832 See documentation in the sources of the run time mentioned in paragraph
6833 @strong{53} for details on GNAT-defined aspects of storage pools.
6838 @strong{57}. The set of restrictions allowed in a pragma
6839 @code{Restrictions}. See 13.12(7).
6842 All RM defined Restriction identifiers are implemented. The following
6843 additional restriction identifiers are provided. There are two separate
6844 lists of implementation dependent restriction identifiers. The first
6845 set requires consistency throughout a partition (in other words, if the
6846 restriction identifier is used for any compilation unit in the partition,
6847 then all compilation units in the partition must obey the restriction.
6851 @item Simple_Barriers
6852 @findex Simple_Barriers
6853 This restriction ensures at compile time that barriers in entry declarations
6854 for protected types are restricted to either static boolean expressions or
6855 references to simple boolean variables defined in the private part of the
6856 protected type. No other form of entry barriers is permitted. This is one
6857 of the restrictions of the Ravenscar profile for limited tasking (see also
6858 pragma @code{Ravenscar}).
6860 @item Max_Entry_Queue_Depth => Expr
6861 @findex Max_Entry_Queue_Depth
6862 This restriction is a declaration that any protected entry compiled in
6863 the scope of the restriction has at most the specified number of
6864 tasks waiting on the entry
6865 at any one time, and so no queue is required. This restriction is not
6866 checked at compile time. A program execution is erroneous if an attempt
6867 is made to queue more than the specified number of tasks on such an entry.
6871 This restriction ensures at compile time that there is no implicit or
6872 explicit dependence on the package @code{Ada.Calendar}.
6874 @item No_Direct_Boolean_Operators
6875 @findex No_Direct_Boolean_Operators
6876 This restriction ensures that no logical (and/or/xor) or comparison
6877 operators are used on operands of type Boolean (or any type derived
6878 from Boolean). This is intended for use in safety critical programs
6879 where the certification protocol requires the use of short-circuit
6880 (and then, or else) forms for all composite boolean operations.
6882 @item No_Dynamic_Interrupts
6883 @findex No_Dynamic_Interrupts
6884 This restriction ensures at compile time that there is no attempt to
6885 dynamically associate interrupts. Only static association is allowed.
6887 @item No_Enumeration_Maps
6888 @findex No_Enumeration_Maps
6889 This restriction ensures at compile time that no operations requiring
6890 enumeration maps are used (that is Image and Value attributes applied
6891 to enumeration types).
6893 @item No_Entry_Calls_In_Elaboration_Code
6894 @findex No_Entry_Calls_In_Elaboration_Code
6895 This restriction ensures at compile time that no task or protected entry
6896 calls are made during elaboration code. As a result of the use of this
6897 restriction, the compiler can assume that no code past an accept statement
6898 in a task can be executed at elaboration time.
6900 @item No_Exception_Handlers
6901 @findex No_Exception_Handlers
6902 This restriction ensures at compile time that there are no explicit
6903 exception handlers. It also indicates that no exception propagation will
6904 be provided. In this mode, exceptions may be raised but will result in
6905 an immediate call to the last chance handler, a routine that the user
6906 must define with the following profile:
6908 procedure Last_Chance_Handler
6909 (Source_Location : System.Address; Line : Integer);
6910 pragma Export (C, Last_Chance_Handler,
6911 "__gnat_last_chance_handler");
6913 The parameter is a C null-terminated string representing a message to be
6914 associated with the exception (typically the source location of the raise
6915 statement generated by the compiler). The Line parameter when non-zero
6916 represents the line number in the source program where the raise occurs.
6918 @item No_Exception_Streams
6919 @findex No_Exception_Streams
6920 This restriction ensures at compile time that no stream operations for
6921 types Exception_Id or Exception_Occurrence are used. This also makes it
6922 impossible to pass exceptions to or from a partition with this restriction
6923 in a distributed environment. If this exception is active, then the generated
6924 code is simplified by omitting the otherwise-required global registration
6925 of exceptions when they are declared.
6927 @item No_Implicit_Conditionals
6928 @findex No_Implicit_Conditionals
6929 This restriction ensures that the generated code does not contain any
6930 implicit conditionals, either by modifying the generated code where possible,
6931 or by rejecting any construct that would otherwise generate an implicit
6934 @item No_Implicit_Dynamic_Code
6935 @findex No_Implicit_Dynamic_Code
6936 This restriction prevents the compiler from building ``trampolines''.
6937 This is a structure that is built on the stack and contains dynamic
6938 code to be executed at run time. A trampoline is needed to indirectly
6939 address a nested subprogram (that is a subprogram that is not at the
6940 library level). The restriction prevents the use of any of the
6941 attributes @code{Address}, @code{Access} or @code{Unrestricted_Access}
6942 being applied to a subprogram that is not at the library level.
6944 @item No_Implicit_Loops
6945 @findex No_Implicit_Loops
6946 This restriction ensures that the generated code does not contain any
6947 implicit @code{for} loops, either by modifying
6948 the generated code where possible,
6949 or by rejecting any construct that would otherwise generate an implicit
6952 @item No_Initialize_Scalars
6953 @findex No_Initialize_Scalars
6954 This restriction ensures that no unit in the partition is compiled with
6955 pragma Initialize_Scalars. This allows the generation of more efficient
6956 code, and in particular eliminates dummy null initialization routines that
6957 are otherwise generated for some record and array types.
6959 @item No_Local_Protected_Objects
6960 @findex No_Local_Protected_Objects
6961 This restriction ensures at compile time that protected objects are
6962 only declared at the library level.
6964 @item No_Protected_Type_Allocators
6965 @findex No_Protected_Type_Allocators
6966 This restriction ensures at compile time that there are no allocator
6967 expressions that attempt to allocate protected objects.
6969 @item No_Secondary_Stack
6970 @findex No_Secondary_Stack
6971 This restriction ensures at compile time that the generated code does not
6972 contain any reference to the secondary stack. The secondary stack is used
6973 to implement functions returning unconstrained objects (arrays or records)
6976 @item No_Select_Statements
6977 @findex No_Select_Statements
6978 This restriction ensures at compile time no select statements of any kind
6979 are permitted, that is the keyword @code{select} may not appear.
6980 This is one of the restrictions of the Ravenscar
6981 profile for limited tasking (see also pragma @code{Ravenscar}).
6983 @item No_Standard_Storage_Pools
6984 @findex No_Standard_Storage_Pools
6985 This restriction ensures at compile time that no access types
6986 use the standard default storage pool. Any access type declared must
6987 have an explicit Storage_Pool attribute defined specifying a
6988 user-defined storage pool.
6992 This restriction ensures at compile time that there are no implicit or
6993 explicit dependencies on the package @code{Ada.Streams}.
6995 @item No_Task_Attributes_Package
6996 @findex No_Task_Attributes_Package
6997 This restriction ensures at compile time that there are no implicit or
6998 explicit dependencies on the package @code{Ada.Task_Attributes}.
7000 @item No_Task_Termination
7001 @findex No_Task_Termination
7002 This restriction ensures at compile time that no terminate alternatives
7003 appear in any task body.
7007 This restriction prevents the declaration of tasks or task types throughout
7008 the partition. It is similar in effect to the use of @code{Max_Tasks => 0}
7009 except that violations are caught at compile time and cause an error message
7010 to be output either by the compiler or binder.
7012 @item No_Wide_Characters
7013 @findex No_Wide_Characters
7014 This restriction ensures at compile time that no uses of the types
7015 @code{Wide_Character} or @code{Wide_String}
7016 appear, and that no wide character literals
7017 appear in the program (that is literals representing characters not in
7018 type @code{Character}.
7020 @item Static_Priorities
7021 @findex Static_Priorities
7022 This restriction ensures at compile time that all priority expressions
7023 are static, and that there are no dependencies on the package
7024 @code{Ada.Dynamic_Priorities}.
7026 @item Static_Storage_Size
7027 @findex Static_Storage_Size
7028 This restriction ensures at compile time that any expression appearing
7029 in a Storage_Size pragma or attribute definition clause is static.
7034 The second set of implementation dependent restriction identifiers
7035 does not require partition-wide consistency.
7036 The restriction may be enforced for a single
7037 compilation unit without any effect on any of the
7038 other compilation units in the partition.
7042 @item No_Elaboration_Code
7043 @findex No_Elaboration_Code
7044 This restriction ensures at compile time that no elaboration code is
7045 generated. Note that this is not the same condition as is enforced
7046 by pragma @code{Preelaborate}. There are cases in which pragma
7047 @code{Preelaborate} still permits code to be generated (e.g.@: code
7048 to initialize a large array to all zeroes), and there are cases of units
7049 which do not meet the requirements for pragma @code{Preelaborate},
7050 but for which no elaboration code is generated. Generally, it is
7051 the case that preelaborable units will meet the restrictions, with
7052 the exception of large aggregates initialized with an others_clause,
7053 and exception declarations (which generate calls to a run-time
7054 registry procedure). Note that this restriction is enforced on
7055 a unit by unit basis, it need not be obeyed consistently
7056 throughout a partition.
7058 @item No_Entry_Queue
7059 @findex No_Entry_Queue
7060 This restriction is a declaration that any protected entry compiled in
7061 the scope of the restriction has at most one task waiting on the entry
7062 at any one time, and so no queue is required. This restriction is not
7063 checked at compile time. A program execution is erroneous if an attempt
7064 is made to queue a second task on such an entry.
7066 @item No_Implementation_Attributes
7067 @findex No_Implementation_Attributes
7068 This restriction checks at compile time that no GNAT-defined attributes
7069 are present. With this restriction, the only attributes that can be used
7070 are those defined in the Ada 95 Reference Manual.
7072 @item No_Implementation_Pragmas
7073 @findex No_Implementation_Pragmas
7074 This restriction checks at compile time that no GNAT-defined pragmas
7075 are present. With this restriction, the only pragmas that can be used
7076 are those defined in the Ada 95 Reference Manual.
7078 @item No_Implementation_Restrictions
7079 @findex No_Implementation_Restrictions
7080 This restriction checks at compile time that no GNAT-defined restriction
7081 identifiers (other than @code{No_Implementation_Restrictions} itself)
7082 are present. With this restriction, the only other restriction identifiers
7083 that can be used are those defined in the Ada 95 Reference Manual.
7090 @strong{58}. The consequences of violating limitations on
7091 @code{Restrictions} pragmas. See 13.12(9).
7094 Restrictions that can be checked at compile time result in illegalities
7095 if violated. Currently there are no other consequences of violating
7101 @strong{59}. The representation used by the @code{Read} and
7102 @code{Write} attributes of elementary types in terms of stream
7103 elements. See 13.13.2(9).
7106 The representation is the in-memory representation of the base type of
7107 the type, using the number of bits corresponding to the
7108 @code{@var{type}'Size} value, and the natural ordering of the machine.
7113 @strong{60}. The names and characteristics of the numeric subtypes
7114 declared in the visible part of package @code{Standard}. See A.1(3).
7117 See items describing the integer and floating-point types supported.
7122 @strong{61}. The accuracy actually achieved by the elementary
7123 functions. See A.5.1(1).
7126 The elementary functions correspond to the functions available in the C
7127 library. Only fast math mode is implemented.
7132 @strong{62}. The sign of a zero result from some of the operators or
7133 functions in @code{Numerics.Generic_Elementary_Functions}, when
7134 @code{Float_Type'Signed_Zeros} is @code{True}. See A.5.1(46).
7137 The sign of zeroes follows the requirements of the IEEE 754 standard on
7143 @strong{63}. The value of
7144 @code{Numerics.Float_Random.Max_Image_Width}. See A.5.2(27).
7147 Maximum image width is 649, see library file @file{a-numran.ads}.
7152 @strong{64}. The value of
7153 @code{Numerics.Discrete_Random.Max_Image_Width}. See A.5.2(27).
7156 Maximum image width is 80, see library file @file{a-nudira.ads}.
7161 @strong{65}. The algorithms for random number generation. See
7165 The algorithm is documented in the source files @file{a-numran.ads} and
7166 @file{a-numran.adb}.
7171 @strong{66}. The string representation of a random number generator's
7172 state. See A.5.2(38).
7175 See the documentation contained in the file @file{a-numran.adb}.
7180 @strong{67}. The minimum time interval between calls to the
7181 time-dependent Reset procedure that are guaranteed to initiate different
7182 random number sequences. See A.5.2(45).
7185 The minimum period between reset calls to guarantee distinct series of
7186 random numbers is one microsecond.
7191 @strong{68}. The values of the @code{Model_Mantissa},
7192 @code{Model_Emin}, @code{Model_Epsilon}, @code{Model},
7193 @code{Safe_First}, and @code{Safe_Last} attributes, if the Numerics
7194 Annex is not supported. See A.5.3(72).
7197 See the source file @file{ttypef.ads} for the values of all numeric
7203 @strong{69}. Any implementation-defined characteristics of the
7204 input-output packages. See A.7(14).
7207 There are no special implementation defined characteristics for these
7213 @strong{70}. The value of @code{Buffer_Size} in @code{Storage_IO}. See
7217 All type representations are contiguous, and the @code{Buffer_Size} is
7218 the value of @code{@var{type}'Size} rounded up to the next storage unit
7224 @strong{71}. External files for standard input, standard output, and
7225 standard error See A.10(5).
7228 These files are mapped onto the files provided by the C streams
7229 libraries. See source file @file{i-cstrea.ads} for further details.
7234 @strong{72}. The accuracy of the value produced by @code{Put}. See
7238 If more digits are requested in the output than are represented by the
7239 precision of the value, zeroes are output in the corresponding least
7240 significant digit positions.
7245 @strong{73}. The meaning of @code{Argument_Count}, @code{Argument}, and
7246 @code{Command_Name}. See A.15(1).
7249 These are mapped onto the @code{argv} and @code{argc} parameters of the
7250 main program in the natural manner.
7255 @strong{74}. Implementation-defined convention names. See B.1(11).
7258 The following convention names are supported
7266 Synonym for Assembler
7268 Synonym for Assembler
7271 @item C_Pass_By_Copy
7272 Allowed only for record types, like C, but also notes that record
7273 is to be passed by copy rather than reference.
7279 Treated the same as C
7281 Treated the same as C
7285 For support of pragma @code{Import} with convention Intrinsic, see
7286 separate section on Intrinsic Subprograms.
7288 Stdcall (used for Windows implementations only). This convention correspond
7289 to the WINAPI (previously called Pascal convention) C/C++ convention under
7290 Windows. A function with this convention cleans the stack before exit.
7296 Stubbed is a special convention used to indicate that the body of the
7297 subprogram will be entirely ignored. Any call to the subprogram
7298 is converted into a raise of the @code{Program_Error} exception. If a
7299 pragma @code{Import} specifies convention @code{stubbed} then no body need
7300 be present at all. This convention is useful during development for the
7301 inclusion of subprograms whose body has not yet been written.
7305 In addition, all otherwise unrecognized convention names are also
7306 treated as being synonymous with convention C@. In all implementations
7307 except for VMS, use of such other names results in a warning. In VMS
7308 implementations, these names are accepted silently.
7313 @strong{75}. The meaning of link names. See B.1(36).
7316 Link names are the actual names used by the linker.
7321 @strong{76}. The manner of choosing link names when neither the link
7322 name nor the address of an imported or exported entity is specified. See
7326 The default linker name is that which would be assigned by the relevant
7327 external language, interpreting the Ada name as being in all lower case
7333 @strong{77}. The effect of pragma @code{Linker_Options}. See B.1(37).
7336 The string passed to @code{Linker_Options} is presented uninterpreted as
7337 an argument to the link command, unless it contains Ascii.NUL characters.
7338 NUL characters if they appear act as argument separators, so for example
7340 @smallexample @c ada
7341 pragma Linker_Options ("-labc" & ASCII.Nul & "-ldef");
7345 causes two separate arguments @code{-labc} and @code{-ldef} to be passed to the
7346 linker. The order of linker options is preserved for a given unit. The final
7347 list of options passed to the linker is in reverse order of the elaboration
7348 order. For example, linker options fo a body always appear before the options
7349 from the corresponding package spec.
7354 @strong{78}. The contents of the visible part of package
7355 @code{Interfaces} and its language-defined descendants. See B.2(1).
7358 See files with prefix @file{i-} in the distributed library.
7363 @strong{79}. Implementation-defined children of package
7364 @code{Interfaces}. The contents of the visible part of package
7365 @code{Interfaces}. See B.2(11).
7368 See files with prefix @file{i-} in the distributed library.
7373 @strong{80}. The types @code{Floating}, @code{Long_Floating},
7374 @code{Binary}, @code{Long_Binary}, @code{Decimal_ Element}, and
7375 @code{COBOL_Character}; and the initialization of the variables
7376 @code{Ada_To_COBOL} and @code{COBOL_To_Ada}, in
7377 @code{Interfaces.COBOL}. See B.4(50).
7384 (Floating) Long_Float
7389 @item Decimal_Element
7391 @item COBOL_Character
7396 For initialization, see the file @file{i-cobol.ads} in the distributed library.
7401 @strong{81}. Support for access to machine instructions. See C.1(1).
7404 See documentation in file @file{s-maccod.ads} in the distributed library.
7409 @strong{82}. Implementation-defined aspects of access to machine
7410 operations. See C.1(9).
7413 See documentation in file @file{s-maccod.ads} in the distributed library.
7418 @strong{83}. Implementation-defined aspects of interrupts. See C.3(2).
7421 Interrupts are mapped to signals or conditions as appropriate. See
7423 @code{Ada.Interrupt_Names} in source file @file{a-intnam.ads} for details
7424 on the interrupts supported on a particular target.
7429 @strong{84}. Implementation-defined aspects of pre-elaboration. See
7433 GNAT does not permit a partition to be restarted without reloading,
7434 except under control of the debugger.
7439 @strong{85}. The semantics of pragma @code{Discard_Names}. See C.5(7).
7442 Pragma @code{Discard_Names} causes names of enumeration literals to
7443 be suppressed. In the presence of this pragma, the Image attribute
7444 provides the image of the Pos of the literal, and Value accepts
7450 @strong{86}. The result of the @code{Task_Identification.Image}
7451 attribute. See C.7.1(7).
7454 The result of this attribute is an 8-digit hexadecimal string
7455 representing the virtual address of the task control block.
7460 @strong{87}. The value of @code{Current_Task} when in a protected entry
7461 or interrupt handler. See C.7.1(17).
7464 Protected entries or interrupt handlers can be executed by any
7465 convenient thread, so the value of @code{Current_Task} is undefined.
7470 @strong{88}. The effect of calling @code{Current_Task} from an entry
7471 body or interrupt handler. See C.7.1(19).
7474 The effect of calling @code{Current_Task} from an entry body or
7475 interrupt handler is to return the identification of the task currently
7481 @strong{89}. Implementation-defined aspects of
7482 @code{Task_Attributes}. See C.7.2(19).
7485 There are no implementation-defined aspects of @code{Task_Attributes}.
7490 @strong{90}. Values of all @code{Metrics}. See D(2).
7493 The metrics information for GNAT depends on the performance of the
7494 underlying operating system. The sources of the run-time for tasking
7495 implementation, together with the output from @code{-gnatG} can be
7496 used to determine the exact sequence of operating systems calls made
7497 to implement various tasking constructs. Together with appropriate
7498 information on the performance of the underlying operating system,
7499 on the exact target in use, this information can be used to determine
7500 the required metrics.
7505 @strong{91}. The declarations of @code{Any_Priority} and
7506 @code{Priority}. See D.1(11).
7509 See declarations in file @file{system.ads}.
7514 @strong{92}. Implementation-defined execution resources. See D.1(15).
7517 There are no implementation-defined execution resources.
7522 @strong{93}. Whether, on a multiprocessor, a task that is waiting for
7523 access to a protected object keeps its processor busy. See D.2.1(3).
7526 On a multi-processor, a task that is waiting for access to a protected
7527 object does not keep its processor busy.
7532 @strong{94}. The affect of implementation defined execution resources
7533 on task dispatching. See D.2.1(9).
7538 Tasks map to IRIX threads, and the dispatching policy is as defined by
7539 the IRIX implementation of threads.
7541 Tasks map to threads in the threads package used by GNAT@. Where possible
7542 and appropriate, these threads correspond to native threads of the
7543 underlying operating system.
7548 @strong{95}. Implementation-defined @code{policy_identifiers} allowed
7549 in a pragma @code{Task_Dispatching_Policy}. See D.2.2(3).
7552 There are no implementation-defined policy-identifiers allowed in this
7558 @strong{96}. Implementation-defined aspects of priority inversion. See
7562 Execution of a task cannot be preempted by the implementation processing
7563 of delay expirations for lower priority tasks.
7568 @strong{97}. Implementation defined task dispatching. See D.2.2(18).
7573 Tasks map to IRIX threads, and the dispatching policy is as defied by
7574 the IRIX implementation of threads.
7576 The policy is the same as that of the underlying threads implementation.
7581 @strong{98}. Implementation-defined @code{policy_identifiers} allowed
7582 in a pragma @code{Locking_Policy}. See D.3(4).
7585 The only implementation defined policy permitted in GNAT is
7586 @code{Inheritance_Locking}. On targets that support this policy, locking
7587 is implemented by inheritance, i.e.@: the task owning the lock operates
7588 at a priority equal to the highest priority of any task currently
7589 requesting the lock.
7594 @strong{99}. Default ceiling priorities. See D.3(10).
7597 The ceiling priority of protected objects of the type
7598 @code{System.Interrupt_Priority'Last} as described in the Ada 95
7599 Reference Manual D.3(10),
7604 @strong{100}. The ceiling of any protected object used internally by
7605 the implementation. See D.3(16).
7608 The ceiling priority of internal protected objects is
7609 @code{System.Priority'Last}.
7614 @strong{101}. Implementation-defined queuing policies. See D.4(1).
7617 There are no implementation-defined queueing policies.
7622 @strong{102}. On a multiprocessor, any conditions that cause the
7623 completion of an aborted construct to be delayed later than what is
7624 specified for a single processor. See D.6(3).
7627 The semantics for abort on a multi-processor is the same as on a single
7628 processor, there are no further delays.
7633 @strong{103}. Any operations that implicitly require heap storage
7634 allocation. See D.7(8).
7637 The only operation that implicitly requires heap storage allocation is
7643 @strong{104}. Implementation-defined aspects of pragma
7644 @code{Restrictions}. See D.7(20).
7647 There are no such implementation-defined aspects.
7652 @strong{105}. Implementation-defined aspects of package
7653 @code{Real_Time}. See D.8(17).
7656 There are no implementation defined aspects of package @code{Real_Time}.
7661 @strong{106}. Implementation-defined aspects of
7662 @code{delay_statements}. See D.9(8).
7665 Any difference greater than one microsecond will cause the task to be
7666 delayed (see D.9(7)).
7671 @strong{107}. The upper bound on the duration of interrupt blocking
7672 caused by the implementation. See D.12(5).
7675 The upper bound is determined by the underlying operating system. In
7676 no cases is it more than 10 milliseconds.
7681 @strong{108}. The means for creating and executing distributed
7685 The GLADE package provides a utility GNATDIST for creating and executing
7686 distributed programs. See the GLADE reference manual for further details.
7691 @strong{109}. Any events that can result in a partition becoming
7692 inaccessible. See E.1(7).
7695 See the GLADE reference manual for full details on such events.
7700 @strong{110}. The scheduling policies, treatment of priorities, and
7701 management of shared resources between partitions in certain cases. See
7705 See the GLADE reference manual for full details on these aspects of
7706 multi-partition execution.
7711 @strong{111}. Events that cause the version of a compilation unit to
7715 Editing the source file of a compilation unit, or the source files of
7716 any units on which it is dependent in a significant way cause the version
7717 to change. No other actions cause the version number to change. All changes
7718 are significant except those which affect only layout, capitalization or
7724 @strong{112}. Whether the execution of the remote subprogram is
7725 immediately aborted as a result of cancellation. See E.4(13).
7728 See the GLADE reference manual for details on the effect of abort in
7729 a distributed application.
7734 @strong{113}. Implementation-defined aspects of the PCS@. See E.5(25).
7737 See the GLADE reference manual for a full description of all implementation
7738 defined aspects of the PCS@.
7743 @strong{114}. Implementation-defined interfaces in the PCS@. See
7747 See the GLADE reference manual for a full description of all
7748 implementation defined interfaces.
7753 @strong{115}. The values of named numbers in the package
7754 @code{Decimal}. See F.2(7).
7766 @item Max_Decimal_Digits
7773 @strong{116}. The value of @code{Max_Picture_Length} in the package
7774 @code{Text_IO.Editing}. See F.3.3(16).
7782 @strong{117}. The value of @code{Max_Picture_Length} in the package
7783 @code{Wide_Text_IO.Editing}. See F.3.4(5).
7791 @strong{118}. The accuracy actually achieved by the complex elementary
7792 functions and by other complex arithmetic operations. See G.1(1).
7795 Standard library functions are used for the complex arithmetic
7796 operations. Only fast math mode is currently supported.
7801 @strong{119}. The sign of a zero result (or a component thereof) from
7802 any operator or function in @code{Numerics.Generic_Complex_Types}, when
7803 @code{Real'Signed_Zeros} is True. See G.1.1(53).
7806 The signs of zero values are as recommended by the relevant
7807 implementation advice.
7812 @strong{120}. The sign of a zero result (or a component thereof) from
7813 any operator or function in
7814 @code{Numerics.Generic_Complex_Elementary_Functions}, when
7815 @code{Real'Signed_Zeros} is @code{True}. See G.1.2(45).
7818 The signs of zero values are as recommended by the relevant
7819 implementation advice.
7824 @strong{121}. Whether the strict mode or the relaxed mode is the
7825 default. See G.2(2).
7828 The strict mode is the default. There is no separate relaxed mode. GNAT
7829 provides a highly efficient implementation of strict mode.
7834 @strong{122}. The result interval in certain cases of fixed-to-float
7835 conversion. See G.2.1(10).
7838 For cases where the result interval is implementation dependent, the
7839 accuracy is that provided by performing all operations in 64-bit IEEE
7840 floating-point format.
7845 @strong{123}. The result of a floating point arithmetic operation in
7846 overflow situations, when the @code{Machine_Overflows} attribute of the
7847 result type is @code{False}. See G.2.1(13).
7850 Infinite and Nan values are produced as dictated by the IEEE
7851 floating-point standard.
7856 @strong{124}. The result interval for division (or exponentiation by a
7857 negative exponent), when the floating point hardware implements division
7858 as multiplication by a reciprocal. See G.2.1(16).
7861 Not relevant, division is IEEE exact.
7866 @strong{125}. The definition of close result set, which determines the
7867 accuracy of certain fixed point multiplications and divisions. See
7871 Operations in the close result set are performed using IEEE long format
7872 floating-point arithmetic. The input operands are converted to
7873 floating-point, the operation is done in floating-point, and the result
7874 is converted to the target type.
7879 @strong{126}. Conditions on a @code{universal_real} operand of a fixed
7880 point multiplication or division for which the result shall be in the
7881 perfect result set. See G.2.3(22).
7884 The result is only defined to be in the perfect result set if the result
7885 can be computed by a single scaling operation involving a scale factor
7886 representable in 64-bits.
7891 @strong{127}. The result of a fixed point arithmetic operation in
7892 overflow situations, when the @code{Machine_Overflows} attribute of the
7893 result type is @code{False}. See G.2.3(27).
7896 Not relevant, @code{Machine_Overflows} is @code{True} for fixed-point
7902 @strong{128}. The result of an elementary function reference in
7903 overflow situations, when the @code{Machine_Overflows} attribute of the
7904 result type is @code{False}. See G.2.4(4).
7907 IEEE infinite and Nan values are produced as appropriate.
7912 @strong{129}. The value of the angle threshold, within which certain
7913 elementary functions, complex arithmetic operations, and complex
7914 elementary functions yield results conforming to a maximum relative
7915 error bound. See G.2.4(10).
7918 Information on this subject is not yet available.
7923 @strong{130}. The accuracy of certain elementary functions for
7924 parameters beyond the angle threshold. See G.2.4(10).
7927 Information on this subject is not yet available.
7932 @strong{131}. The result of a complex arithmetic operation or complex
7933 elementary function reference in overflow situations, when the
7934 @code{Machine_Overflows} attribute of the corresponding real type is
7935 @code{False}. See G.2.6(5).
7938 IEEE infinite and Nan values are produced as appropriate.
7943 @strong{132}. The accuracy of certain complex arithmetic operations and
7944 certain complex elementary functions for parameters (or components
7945 thereof) beyond the angle threshold. See G.2.6(8).
7948 Information on those subjects is not yet available.
7953 @strong{133}. Information regarding bounded errors and erroneous
7954 execution. See H.2(1).
7957 Information on this subject is not yet available.
7962 @strong{134}. Implementation-defined aspects of pragma
7963 @code{Inspection_Point}. See H.3.2(8).
7966 Pragma @code{Inspection_Point} ensures that the variable is live and can
7967 be examined by the debugger at the inspection point.
7972 @strong{135}. Implementation-defined aspects of pragma
7973 @code{Restrictions}. See H.4(25).
7976 There are no implementation-defined aspects of pragma @code{Restrictions}. The
7977 use of pragma @code{Restrictions [No_Exceptions]} has no effect on the
7978 generated code. Checks must suppressed by use of pragma @code{Suppress}.
7983 @strong{136}. Any restrictions on pragma @code{Restrictions}. See
7987 There are no restrictions on pragma @code{Restrictions}.
7989 @node Intrinsic Subprograms
7990 @chapter Intrinsic Subprograms
7991 @cindex Intrinsic Subprograms
7994 * Intrinsic Operators::
7995 * Enclosing_Entity::
7996 * Exception_Information::
7997 * Exception_Message::
8005 * Shift_Right_Arithmetic::
8010 GNAT allows a user application program to write the declaration:
8012 @smallexample @c ada
8013 pragma Import (Intrinsic, name);
8017 providing that the name corresponds to one of the implemented intrinsic
8018 subprograms in GNAT, and that the parameter profile of the referenced
8019 subprogram meets the requirements. This chapter describes the set of
8020 implemented intrinsic subprograms, and the requirements on parameter profiles.
8021 Note that no body is supplied; as with other uses of pragma Import, the
8022 body is supplied elsewhere (in this case by the compiler itself). Note
8023 that any use of this feature is potentially non-portable, since the
8024 Ada standard does not require Ada compilers to implement this feature.
8026 @node Intrinsic Operators
8027 @section Intrinsic Operators
8028 @cindex Intrinsic operator
8031 All the predefined numeric operators in package Standard
8032 in @code{pragma Import (Intrinsic,..)}
8033 declarations. In the binary operator case, the operands must have the same
8034 size. The operand or operands must also be appropriate for
8035 the operator. For example, for addition, the operands must
8036 both be floating-point or both be fixed-point, and the
8037 right operand for @code{"**"} must have a root type of
8038 @code{Standard.Integer'Base}.
8039 You can use an intrinsic operator declaration as in the following example:
8041 @smallexample @c ada
8042 type Int1 is new Integer;
8043 type Int2 is new Integer;
8045 function "+" (X1 : Int1; X2 : Int2) return Int1;
8046 function "+" (X1 : Int1; X2 : Int2) return Int2;
8047 pragma Import (Intrinsic, "+");
8051 This declaration would permit ``mixed mode'' arithmetic on items
8052 of the differing types @code{Int1} and @code{Int2}.
8053 It is also possible to specify such operators for private types, if the
8054 full views are appropriate arithmetic types.
8056 @node Enclosing_Entity
8057 @section Enclosing_Entity
8058 @cindex Enclosing_Entity
8060 This intrinsic subprogram is used in the implementation of the
8061 library routine @code{GNAT.Source_Info}. The only useful use of the
8062 intrinsic import in this case is the one in this unit, so an
8063 application program should simply call the function
8064 @code{GNAT.Source_Info.Enclosing_Entity} to obtain the name of
8065 the current subprogram, package, task, entry, or protected subprogram.
8067 @node Exception_Information
8068 @section Exception_Information
8069 @cindex Exception_Information'
8071 This intrinsic subprogram is used in the implementation of the
8072 library routine @code{GNAT.Current_Exception}. The only useful
8073 use of the intrinsic import in this case is the one in this unit,
8074 so an application program should simply call the function
8075 @code{GNAT.Current_Exception.Exception_Information} to obtain
8076 the exception information associated with the current exception.
8078 @node Exception_Message
8079 @section Exception_Message
8080 @cindex Exception_Message
8082 This intrinsic subprogram is used in the implementation of the
8083 library routine @code{GNAT.Current_Exception}. The only useful
8084 use of the intrinsic import in this case is the one in this unit,
8085 so an application program should simply call the function
8086 @code{GNAT.Current_Exception.Exception_Message} to obtain
8087 the message associated with the current exception.
8089 @node Exception_Name
8090 @section Exception_Name
8091 @cindex Exception_Name
8093 This intrinsic subprogram is used in the implementation of the
8094 library routine @code{GNAT.Current_Exception}. The only useful
8095 use of the intrinsic import in this case is the one in this unit,
8096 so an application program should simply call the function
8097 @code{GNAT.Current_Exception.Exception_Name} to obtain
8098 the name of the current exception.
8104 This intrinsic subprogram is used in the implementation of the
8105 library routine @code{GNAT.Source_Info}. The only useful use of the
8106 intrinsic import in this case is the one in this unit, so an
8107 application program should simply call the function
8108 @code{GNAT.Source_Info.File} to obtain the name of the current
8115 This intrinsic subprogram is used in the implementation of the
8116 library routine @code{GNAT.Source_Info}. The only useful use of the
8117 intrinsic import in this case is the one in this unit, so an
8118 application program should simply call the function
8119 @code{GNAT.Source_Info.Line} to obtain the number of the current
8123 @section Rotate_Left
8126 In standard Ada 95, the @code{Rotate_Left} function is available only
8127 for the predefined modular types in package @code{Interfaces}. However, in
8128 GNAT it is possible to define a Rotate_Left function for a user
8129 defined modular type or any signed integer type as in this example:
8131 @smallexample @c ada
8133 (Value : My_Modular_Type;
8135 return My_Modular_Type;
8139 The requirements are that the profile be exactly as in the example
8140 above. The only modifications allowed are in the formal parameter
8141 names, and in the type of @code{Value} and the return type, which
8142 must be the same, and must be either a signed integer type, or
8143 a modular integer type with a binary modulus, and the size must
8144 be 8. 16, 32 or 64 bits.
8147 @section Rotate_Right
8148 @cindex Rotate_Right
8150 A @code{Rotate_Right} function can be defined for any user defined
8151 binary modular integer type, or signed integer type, as described
8152 above for @code{Rotate_Left}.
8158 A @code{Shift_Left} function can be defined for any user defined
8159 binary modular integer type, or signed integer type, as described
8160 above for @code{Rotate_Left}.
8163 @section Shift_Right
8166 A @code{Shift_Right} function can be defined for any user defined
8167 binary modular integer type, or signed integer type, as described
8168 above for @code{Rotate_Left}.
8170 @node Shift_Right_Arithmetic
8171 @section Shift_Right_Arithmetic
8172 @cindex Shift_Right_Arithmetic
8174 A @code{Shift_Right_Arithmetic} function can be defined for any user
8175 defined binary modular integer type, or signed integer type, as described
8176 above for @code{Rotate_Left}.
8178 @node Source_Location
8179 @section Source_Location
8180 @cindex Source_Location
8182 This intrinsic subprogram is used in the implementation of the
8183 library routine @code{GNAT.Source_Info}. The only useful use of the
8184 intrinsic import in this case is the one in this unit, so an
8185 application program should simply call the function
8186 @code{GNAT.Source_Info.Source_Location} to obtain the current
8187 source file location.
8189 @node Representation Clauses and Pragmas
8190 @chapter Representation Clauses and Pragmas
8191 @cindex Representation Clauses
8194 * Alignment Clauses::
8196 * Storage_Size Clauses::
8197 * Size of Variant Record Objects::
8198 * Biased Representation ::
8199 * Value_Size and Object_Size Clauses::
8200 * Component_Size Clauses::
8201 * Bit_Order Clauses::
8202 * Effect of Bit_Order on Byte Ordering::
8203 * Pragma Pack for Arrays::
8204 * Pragma Pack for Records::
8205 * Record Representation Clauses::
8206 * Enumeration Clauses::
8208 * Effect of Convention on Representation::
8209 * Determining the Representations chosen by GNAT::
8213 @cindex Representation Clause
8214 @cindex Representation Pragma
8215 @cindex Pragma, representation
8216 This section describes the representation clauses accepted by GNAT, and
8217 their effect on the representation of corresponding data objects.
8219 GNAT fully implements Annex C (Systems Programming). This means that all
8220 the implementation advice sections in chapter 13 are fully implemented.
8221 However, these sections only require a minimal level of support for
8222 representation clauses. GNAT provides much more extensive capabilities,
8223 and this section describes the additional capabilities provided.
8225 @node Alignment Clauses
8226 @section Alignment Clauses
8227 @cindex Alignment Clause
8230 GNAT requires that all alignment clauses specify a power of 2, and all
8231 default alignments are always a power of 2. The default alignment
8232 values are as follows:
8235 @item @emph{Primitive Types}.
8236 For primitive types, the alignment is the minimum of the actual size of
8237 objects of the type divided by @code{Storage_Unit},
8238 and the maximum alignment supported by the target.
8239 (This maximum alignment is given by the GNAT-specific attribute
8240 @code{Standard'Maximum_Alignment}; see @ref{Maximum_Alignment}.)
8241 @cindex @code{Maximum_Alignment} attribute
8242 For example, for type @code{Long_Float}, the object size is 8 bytes, and the
8243 default alignment will be 8 on any target that supports alignments
8244 this large, but on some targets, the maximum alignment may be smaller
8245 than 8, in which case objects of type @code{Long_Float} will be maximally
8248 @item @emph{Arrays}.
8249 For arrays, the alignment is equal to the alignment of the component type
8250 for the normal case where no packing or component size is given. If the
8251 array is packed, and the packing is effective (see separate section on
8252 packed arrays), then the alignment will be one for long packed arrays,
8253 or arrays whose length is not known at compile time. For short packed
8254 arrays, which are handled internally as modular types, the alignment
8255 will be as described for primitive types, e.g.@: a packed array of length
8256 31 bits will have an object size of four bytes, and an alignment of 4.
8258 @item @emph{Records}.
8259 For the normal non-packed case, the alignment of a record is equal to
8260 the maximum alignment of any of its components. For tagged records, this
8261 includes the implicit access type used for the tag. If a pragma @code{Pack} is
8262 used and all fields are packable (see separate section on pragma @code{Pack}),
8263 then the resulting alignment is 1.
8265 A special case is when:
8268 the size of the record is given explicitly, or a
8269 full record representation clause is given, and
8271 the size of the record is 2, 4, or 8 bytes.
8274 In this case, an alignment is chosen to match the
8275 size of the record. For example, if we have:
8277 @smallexample @c ada
8278 type Small is record
8281 for Small'Size use 16;
8285 then the default alignment of the record type @code{Small} is 2, not 1. This
8286 leads to more efficient code when the record is treated as a unit, and also
8287 allows the type to specified as @code{Atomic} on architectures requiring
8293 An alignment clause may
8294 always specify a larger alignment than the default value, up to some
8295 maximum value dependent on the target (obtainable by using the
8296 attribute reference @code{Standard'Maximum_Alignment}).
8298 it is permissible to specify a smaller alignment than the default value
8299 is for a record with a record representation clause.
8300 In this case, packable fields for which a component clause is
8301 given still result in a default alignment corresponding to the original
8302 type, but this may be overridden, since these components in fact only
8303 require an alignment of one byte. For example, given
8305 @smallexample @c ada
8311 A at 0 range 0 .. 31;
8314 for V'alignment use 1;
8318 @cindex Alignment, default
8319 The default alignment for the type @code{V} is 4, as a result of the
8320 Integer field in the record, but since this field is placed with a
8321 component clause, it is permissible, as shown, to override the default
8322 alignment of the record with a smaller value.
8325 @section Size Clauses
8329 The default size for a type @code{T} is obtainable through the
8330 language-defined attribute @code{T'Size} and also through the
8331 equivalent GNAT-defined attribute @code{T'Value_Size}.
8332 For objects of type @code{T}, GNAT will generally increase the type size
8333 so that the object size (obtainable through the GNAT-defined attribute
8334 @code{T'Object_Size})
8335 is a multiple of @code{T'Alignment * Storage_Unit}.
8338 @smallexample @c ada
8339 type Smallint is range 1 .. 6;
8348 In this example, @code{Smallint'Size} = @code{Smallint'Value_Size} = 3,
8349 as specified by the RM rules,
8350 but objects of this type will have a size of 8
8351 (@code{Smallint'Object_Size} = 8),
8352 since objects by default occupy an integral number
8353 of storage units. On some targets, notably older
8354 versions of the Digital Alpha, the size of stand
8355 alone objects of this type may be 32, reflecting
8356 the inability of the hardware to do byte load/stores.
8358 Similarly, the size of type @code{Rec} is 40 bits
8359 (@code{Rec'Size} = @code{Rec'Value_Size} = 40), but
8360 the alignment is 4, so objects of this type will have
8361 their size increased to 64 bits so that it is a multiple
8362 of the alignment (in bits). The reason for this decision, which is
8363 in accordance with the specific Implementation Advice in RM 13.3(43):
8366 A @code{Size} clause should be supported for an object if the specified
8367 @code{Size} is at least as large as its subtype's @code{Size}, and corresponds
8368 to a size in storage elements that is a multiple of the object's
8369 @code{Alignment} (if the @code{Alignment} is nonzero).
8373 An explicit size clause may be used to override the default size by
8374 increasing it. For example, if we have:
8376 @smallexample @c ada
8377 type My_Boolean is new Boolean;
8378 for My_Boolean'Size use 32;
8382 then values of this type will always be 32 bits long. In the case of
8383 discrete types, the size can be increased up to 64 bits, with the effect
8384 that the entire specified field is used to hold the value, sign- or
8385 zero-extended as appropriate. If more than 64 bits is specified, then
8386 padding space is allocated after the value, and a warning is issued that
8387 there are unused bits.
8389 Similarly the size of records and arrays may be increased, and the effect
8390 is to add padding bits after the value. This also causes a warning message
8393 The largest Size value permitted in GNAT is 2**31@minus{}1. Since this is a
8394 Size in bits, this corresponds to an object of size 256 megabytes (minus
8395 one). This limitation is true on all targets. The reason for this
8396 limitation is that it improves the quality of the code in many cases
8397 if it is known that a Size value can be accommodated in an object of
8400 @node Storage_Size Clauses
8401 @section Storage_Size Clauses
8402 @cindex Storage_Size Clause
8405 For tasks, the @code{Storage_Size} clause specifies the amount of space
8406 to be allocated for the task stack. This cannot be extended, and if the
8407 stack is exhausted, then @code{Storage_Error} will be raised (if stack
8408 checking is enabled). Use a @code{Storage_Size} attribute definition clause,
8409 or a @code{Storage_Size} pragma in the task definition to set the
8410 appropriate required size. A useful technique is to include in every
8411 task definition a pragma of the form:
8413 @smallexample @c ada
8414 pragma Storage_Size (Default_Stack_Size);
8418 Then @code{Default_Stack_Size} can be defined in a global package, and
8419 modified as required. Any tasks requiring stack sizes different from the
8420 default can have an appropriate alternative reference in the pragma.
8422 For access types, the @code{Storage_Size} clause specifies the maximum
8423 space available for allocation of objects of the type. If this space is
8424 exceeded then @code{Storage_Error} will be raised by an allocation attempt.
8425 In the case where the access type is declared local to a subprogram, the
8426 use of a @code{Storage_Size} clause triggers automatic use of a special
8427 predefined storage pool (@code{System.Pool_Size}) that ensures that all
8428 space for the pool is automatically reclaimed on exit from the scope in
8429 which the type is declared.
8431 A special case recognized by the compiler is the specification of a
8432 @code{Storage_Size} of zero for an access type. This means that no
8433 items can be allocated from the pool, and this is recognized at compile
8434 time, and all the overhead normally associated with maintaining a fixed
8435 size storage pool is eliminated. Consider the following example:
8437 @smallexample @c ada
8439 type R is array (Natural) of Character;
8440 type P is access all R;
8441 for P'Storage_Size use 0;
8442 -- Above access type intended only for interfacing purposes
8446 procedure g (m : P);
8447 pragma Import (C, g);
8458 As indicated in this example, these dummy storage pools are often useful in
8459 connection with interfacing where no object will ever be allocated. If you
8460 compile the above example, you get the warning:
8463 p.adb:16:09: warning: allocation from empty storage pool
8464 p.adb:16:09: warning: Storage_Error will be raised at run time
8468 Of course in practice, there will not be any explicit allocators in the
8469 case of such an access declaration.
8471 @node Size of Variant Record Objects
8472 @section Size of Variant Record Objects
8473 @cindex Size, variant record objects
8474 @cindex Variant record objects, size
8477 In the case of variant record objects, there is a question whether Size gives
8478 information about a particular variant, or the maximum size required
8479 for any variant. Consider the following program
8481 @smallexample @c ada
8482 with Text_IO; use Text_IO;
8484 type R1 (A : Boolean := False) is record
8486 when True => X : Character;
8495 Put_Line (Integer'Image (V1'Size));
8496 Put_Line (Integer'Image (V2'Size));
8501 Here we are dealing with a variant record, where the True variant
8502 requires 16 bits, and the False variant requires 8 bits.
8503 In the above example, both V1 and V2 contain the False variant,
8504 which is only 8 bits long. However, the result of running the
8513 The reason for the difference here is that the discriminant value of
8514 V1 is fixed, and will always be False. It is not possible to assign
8515 a True variant value to V1, therefore 8 bits is sufficient. On the
8516 other hand, in the case of V2, the initial discriminant value is
8517 False (from the default), but it is possible to assign a True
8518 variant value to V2, therefore 16 bits must be allocated for V2
8519 in the general case, even fewer bits may be needed at any particular
8520 point during the program execution.
8522 As can be seen from the output of this program, the @code{'Size}
8523 attribute applied to such an object in GNAT gives the actual allocated
8524 size of the variable, which is the largest size of any of the variants.
8525 The Ada Reference Manual is not completely clear on what choice should
8526 be made here, but the GNAT behavior seems most consistent with the
8527 language in the RM@.
8529 In some cases, it may be desirable to obtain the size of the current
8530 variant, rather than the size of the largest variant. This can be
8531 achieved in GNAT by making use of the fact that in the case of a
8532 subprogram parameter, GNAT does indeed return the size of the current
8533 variant (because a subprogram has no way of knowing how much space
8534 is actually allocated for the actual).
8536 Consider the following modified version of the above program:
8538 @smallexample @c ada
8539 with Text_IO; use Text_IO;
8541 type R1 (A : Boolean := False) is record
8543 when True => X : Character;
8550 function Size (V : R1) return Integer is
8556 Put_Line (Integer'Image (V2'Size));
8557 Put_Line (Integer'IMage (Size (V2)));
8559 Put_Line (Integer'Image (V2'Size));
8560 Put_Line (Integer'IMage (Size (V2)));
8565 The output from this program is
8575 Here we see that while the @code{'Size} attribute always returns
8576 the maximum size, regardless of the current variant value, the
8577 @code{Size} function does indeed return the size of the current
8580 @node Biased Representation
8581 @section Biased Representation
8582 @cindex Size for biased representation
8583 @cindex Biased representation
8586 In the case of scalars with a range starting at other than zero, it is
8587 possible in some cases to specify a size smaller than the default minimum
8588 value, and in such cases, GNAT uses an unsigned biased representation,
8589 in which zero is used to represent the lower bound, and successive values
8590 represent successive values of the type.
8592 For example, suppose we have the declaration:
8594 @smallexample @c ada
8595 type Small is range -7 .. -4;
8596 for Small'Size use 2;
8600 Although the default size of type @code{Small} is 4, the @code{Size}
8601 clause is accepted by GNAT and results in the following representation
8605 -7 is represented as 2#00#
8606 -6 is represented as 2#01#
8607 -5 is represented as 2#10#
8608 -4 is represented as 2#11#
8612 Biased representation is only used if the specified @code{Size} clause
8613 cannot be accepted in any other manner. These reduced sizes that force
8614 biased representation can be used for all discrete types except for
8615 enumeration types for which a representation clause is given.
8617 @node Value_Size and Object_Size Clauses
8618 @section Value_Size and Object_Size Clauses
8621 @cindex Size, of objects
8624 In Ada 95, @code{T'Size} for a type @code{T} is the minimum number of bits
8625 required to hold values of type @code{T}. Although this interpretation was
8626 allowed in Ada 83, it was not required, and this requirement in practice
8627 can cause some significant difficulties. For example, in most Ada 83
8628 compilers, @code{Natural'Size} was 32. However, in Ada 95,
8629 @code{Natural'Size} is
8630 typically 31. This means that code may change in behavior when moving
8631 from Ada 83 to Ada 95. For example, consider:
8633 @smallexample @c ada
8640 at 0 range 0 .. Natural'Size - 1;
8641 at 0 range Natural'Size .. 2 * Natural'Size - 1;
8646 In the above code, since the typical size of @code{Natural} objects
8647 is 32 bits and @code{Natural'Size} is 31, the above code can cause
8648 unexpected inefficient packing in Ada 95, and in general there are
8649 cases where the fact that the object size can exceed the
8650 size of the type causes surprises.
8652 To help get around this problem GNAT provides two implementation
8653 defined attributes, @code{Value_Size} and @code{Object_Size}. When
8654 applied to a type, these attributes yield the size of the type
8655 (corresponding to the RM defined size attribute), and the size of
8656 objects of the type respectively.
8658 The @code{Object_Size} is used for determining the default size of
8659 objects and components. This size value can be referred to using the
8660 @code{Object_Size} attribute. The phrase ``is used'' here means that it is
8661 the basis of the determination of the size. The backend is free to
8662 pad this up if necessary for efficiency, e.g.@: an 8-bit stand-alone
8663 character might be stored in 32 bits on a machine with no efficient
8664 byte access instructions such as the Alpha.
8666 The default rules for the value of @code{Object_Size} for
8667 discrete types are as follows:
8671 The @code{Object_Size} for base subtypes reflect the natural hardware
8672 size in bits (run the compiler with @option{-gnatS} to find those values
8673 for numeric types). Enumeration types and fixed-point base subtypes have
8674 8, 16, 32 or 64 bits for this size, depending on the range of values
8678 The @code{Object_Size} of a subtype is the same as the
8679 @code{Object_Size} of
8680 the type from which it is obtained.
8683 The @code{Object_Size} of a derived base type is copied from the parent
8684 base type, and the @code{Object_Size} of a derived first subtype is copied
8685 from the parent first subtype.
8689 The @code{Value_Size} attribute
8690 is the (minimum) number of bits required to store a value
8692 This value is used to determine how tightly to pack
8693 records or arrays with components of this type, and also affects
8694 the semantics of unchecked conversion (unchecked conversions where
8695 the @code{Value_Size} values differ generate a warning, and are potentially
8698 The default rules for the value of @code{Value_Size} are as follows:
8702 The @code{Value_Size} for a base subtype is the minimum number of bits
8703 required to store all values of the type (including the sign bit
8704 only if negative values are possible).
8707 If a subtype statically matches the first subtype of a given type, then it has
8708 by default the same @code{Value_Size} as the first subtype. This is a
8709 consequence of RM 13.1(14) (``if two subtypes statically match,
8710 then their subtype-specific aspects are the same''.)
8713 All other subtypes have a @code{Value_Size} corresponding to the minimum
8714 number of bits required to store all values of the subtype. For
8715 dynamic bounds, it is assumed that the value can range down or up
8716 to the corresponding bound of the ancestor
8720 The RM defined attribute @code{Size} corresponds to the
8721 @code{Value_Size} attribute.
8723 The @code{Size} attribute may be defined for a first-named subtype. This sets
8724 the @code{Value_Size} of
8725 the first-named subtype to the given value, and the
8726 @code{Object_Size} of this first-named subtype to the given value padded up
8727 to an appropriate boundary. It is a consequence of the default rules
8728 above that this @code{Object_Size} will apply to all further subtypes. On the
8729 other hand, @code{Value_Size} is affected only for the first subtype, any
8730 dynamic subtypes obtained from it directly, and any statically matching
8731 subtypes. The @code{Value_Size} of any other static subtypes is not affected.
8733 @code{Value_Size} and
8734 @code{Object_Size} may be explicitly set for any subtype using
8735 an attribute definition clause. Note that the use of these attributes
8736 can cause the RM 13.1(14) rule to be violated. If two access types
8737 reference aliased objects whose subtypes have differing @code{Object_Size}
8738 values as a result of explicit attribute definition clauses, then it
8739 is erroneous to convert from one access subtype to the other.
8741 At the implementation level, Esize stores the Object_Size and the
8742 RM_Size field stores the @code{Value_Size} (and hence the value of the
8743 @code{Size} attribute,
8744 which, as noted above, is equivalent to @code{Value_Size}).
8746 To get a feel for the difference, consider the following examples (note
8747 that in each case the base is @code{Short_Short_Integer} with a size of 8):
8750 Object_Size Value_Size
8752 type x1 is range 0 .. 5; 8 3
8754 type x2 is range 0 .. 5;
8755 for x2'size use 12; 16 12
8757 subtype x3 is x2 range 0 .. 3; 16 2
8759 subtype x4 is x2'base range 0 .. 10; 8 4
8761 subtype x5 is x2 range 0 .. dynamic; 16 3*
8763 subtype x6 is x2'base range 0 .. dynamic; 8 3*
8768 Note: the entries marked ``3*'' are not actually specified by the Ada 95 RM,
8769 but it seems in the spirit of the RM rules to allocate the minimum number
8770 of bits (here 3, given the range for @code{x2})
8771 known to be large enough to hold the given range of values.
8773 So far, so good, but GNAT has to obey the RM rules, so the question is
8774 under what conditions must the RM @code{Size} be used.
8775 The following is a list
8776 of the occasions on which the RM @code{Size} must be used:
8780 Component size for packed arrays or records
8783 Value of the attribute @code{Size} for a type
8786 Warning about sizes not matching for unchecked conversion
8790 For record types, the @code{Object_Size} is always a multiple of the
8791 alignment of the type (this is true for all types). In some cases the
8792 @code{Value_Size} can be smaller. Consider:
8802 On a typical 32-bit architecture, the X component will be four bytes, and
8803 require four-byte alignment, and the Y component will be one byte. In this
8804 case @code{R'Value_Size} will be 40 (bits) since this is the minimum size
8805 required to store a value of this type, and for example, it is permissible
8806 to have a component of type R in an outer record whose component size is
8807 specified to be 48 bits. However, @code{R'Object_Size} will be 64 (bits),
8808 since it must be rounded up so that this value is a multiple of the
8809 alignment (4 bytes = 32 bits).
8812 For all other types, the @code{Object_Size}
8813 and Value_Size are the same (and equivalent to the RM attribute @code{Size}).
8814 Only @code{Size} may be specified for such types.
8816 @node Component_Size Clauses
8817 @section Component_Size Clauses
8818 @cindex Component_Size Clause
8821 Normally, the value specified in a component clause must be consistent
8822 with the subtype of the array component with regard to size and alignment.
8823 In other words, the value specified must be at least equal to the size
8824 of this subtype, and must be a multiple of the alignment value.
8826 In addition, component size clauses are allowed which cause the array
8827 to be packed, by specifying a smaller value. The cases in which this
8828 is allowed are for component size values in the range 1 through 63. The value
8829 specified must not be smaller than the Size of the subtype. GNAT will
8830 accurately honor all packing requests in this range. For example, if
8833 @smallexample @c ada
8834 type r is array (1 .. 8) of Natural;
8835 for r'Component_Size use 31;
8839 then the resulting array has a length of 31 bytes (248 bits = 8 * 31).
8840 Of course access to the components of such an array is considerably
8841 less efficient than if the natural component size of 32 is used.
8843 @node Bit_Order Clauses
8844 @section Bit_Order Clauses
8845 @cindex Bit_Order Clause
8846 @cindex bit ordering
8847 @cindex ordering, of bits
8850 For record subtypes, GNAT permits the specification of the @code{Bit_Order}
8851 attribute. The specification may either correspond to the default bit
8852 order for the target, in which case the specification has no effect and
8853 places no additional restrictions, or it may be for the non-standard
8854 setting (that is the opposite of the default).
8856 In the case where the non-standard value is specified, the effect is
8857 to renumber bits within each byte, but the ordering of bytes is not
8858 affected. There are certain
8859 restrictions placed on component clauses as follows:
8863 @item Components fitting within a single storage unit.
8865 These are unrestricted, and the effect is merely to renumber bits. For
8866 example if we are on a little-endian machine with @code{Low_Order_First}
8867 being the default, then the following two declarations have exactly
8870 @smallexample @c ada
8873 B : Integer range 1 .. 120;
8877 A at 0 range 0 .. 0;
8878 B at 0 range 1 .. 7;
8883 B : Integer range 1 .. 120;
8886 for R2'Bit_Order use High_Order_First;
8889 A at 0 range 7 .. 7;
8890 B at 0 range 0 .. 6;
8895 The useful application here is to write the second declaration with the
8896 @code{Bit_Order} attribute definition clause, and know that it will be treated
8897 the same, regardless of whether the target is little-endian or big-endian.
8899 @item Components occupying an integral number of bytes.
8901 These are components that exactly fit in two or more bytes. Such component
8902 declarations are allowed, but have no effect, since it is important to realize
8903 that the @code{Bit_Order} specification does not affect the ordering of bytes.
8904 In particular, the following attempt at getting an endian-independent integer
8907 @smallexample @c ada
8912 for R2'Bit_Order use High_Order_First;
8915 A at 0 range 0 .. 31;
8920 This declaration will result in a little-endian integer on a
8921 little-endian machine, and a big-endian integer on a big-endian machine.
8922 If byte flipping is required for interoperability between big- and
8923 little-endian machines, this must be explicitly programmed. This capability
8924 is not provided by @code{Bit_Order}.
8926 @item Components that are positioned across byte boundaries
8928 but do not occupy an integral number of bytes. Given that bytes are not
8929 reordered, such fields would occupy a non-contiguous sequence of bits
8930 in memory, requiring non-trivial code to reassemble. They are for this
8931 reason not permitted, and any component clause specifying such a layout
8932 will be flagged as illegal by GNAT@.
8937 Since the misconception that Bit_Order automatically deals with all
8938 endian-related incompatibilities is a common one, the specification of
8939 a component field that is an integral number of bytes will always
8940 generate a warning. This warning may be suppressed using
8941 @code{pragma Suppress} if desired. The following section contains additional
8942 details regarding the issue of byte ordering.
8944 @node Effect of Bit_Order on Byte Ordering
8945 @section Effect of Bit_Order on Byte Ordering
8946 @cindex byte ordering
8947 @cindex ordering, of bytes
8950 In this section we will review the effect of the @code{Bit_Order} attribute
8951 definition clause on byte ordering. Briefly, it has no effect at all, but
8952 a detailed example will be helpful. Before giving this
8953 example, let us review the precise
8954 definition of the effect of defining @code{Bit_Order}. The effect of a
8955 non-standard bit order is described in section 15.5.3 of the Ada
8959 2 A bit ordering is a method of interpreting the meaning of
8960 the storage place attributes.
8964 To understand the precise definition of storage place attributes in
8965 this context, we visit section 13.5.1 of the manual:
8968 13 A record_representation_clause (without the mod_clause)
8969 specifies the layout. The storage place attributes (see 13.5.2)
8970 are taken from the values of the position, first_bit, and last_bit
8971 expressions after normalizing those values so that first_bit is
8972 less than Storage_Unit.
8976 The critical point here is that storage places are taken from
8977 the values after normalization, not before. So the @code{Bit_Order}
8978 interpretation applies to normalized values. The interpretation
8979 is described in the later part of the 15.5.3 paragraph:
8982 2 A bit ordering is a method of interpreting the meaning of
8983 the storage place attributes. High_Order_First (known in the
8984 vernacular as ``big endian'') means that the first bit of a
8985 storage element (bit 0) is the most significant bit (interpreting
8986 the sequence of bits that represent a component as an unsigned
8987 integer value). Low_Order_First (known in the vernacular as
8988 ``little endian'') means the opposite: the first bit is the
8993 Note that the numbering is with respect to the bits of a storage
8994 unit. In other words, the specification affects only the numbering
8995 of bits within a single storage unit.
8997 We can make the effect clearer by giving an example.
8999 Suppose that we have an external device which presents two bytes, the first
9000 byte presented, which is the first (low addressed byte) of the two byte
9001 record is called Master, and the second byte is called Slave.
9003 The left most (most significant bit is called Control for each byte, and
9004 the remaining 7 bits are called V1, V2, @dots{} V7, where V7 is the rightmost
9005 (least significant) bit.
9007 On a big-endian machine, we can write the following representation clause
9009 @smallexample @c ada
9011 Master_Control : Bit;
9019 Slave_Control : Bit;
9030 Master_Control at 0 range 0 .. 0;
9031 Master_V1 at 0 range 1 .. 1;
9032 Master_V2 at 0 range 2 .. 2;
9033 Master_V3 at 0 range 3 .. 3;
9034 Master_V4 at 0 range 4 .. 4;
9035 Master_V5 at 0 range 5 .. 5;
9036 Master_V6 at 0 range 6 .. 6;
9037 Master_V7 at 0 range 7 .. 7;
9038 Slave_Control at 1 range 0 .. 0;
9039 Slave_V1 at 1 range 1 .. 1;
9040 Slave_V2 at 1 range 2 .. 2;
9041 Slave_V3 at 1 range 3 .. 3;
9042 Slave_V4 at 1 range 4 .. 4;
9043 Slave_V5 at 1 range 5 .. 5;
9044 Slave_V6 at 1 range 6 .. 6;
9045 Slave_V7 at 1 range 7 .. 7;
9050 Now if we move this to a little endian machine, then the bit ordering within
9051 the byte is backwards, so we have to rewrite the record rep clause as:
9053 @smallexample @c ada
9055 Master_Control at 0 range 7 .. 7;
9056 Master_V1 at 0 range 6 .. 6;
9057 Master_V2 at 0 range 5 .. 5;
9058 Master_V3 at 0 range 4 .. 4;
9059 Master_V4 at 0 range 3 .. 3;
9060 Master_V5 at 0 range 2 .. 2;
9061 Master_V6 at 0 range 1 .. 1;
9062 Master_V7 at 0 range 0 .. 0;
9063 Slave_Control at 1 range 7 .. 7;
9064 Slave_V1 at 1 range 6 .. 6;
9065 Slave_V2 at 1 range 5 .. 5;
9066 Slave_V3 at 1 range 4 .. 4;
9067 Slave_V4 at 1 range 3 .. 3;
9068 Slave_V5 at 1 range 2 .. 2;
9069 Slave_V6 at 1 range 1 .. 1;
9070 Slave_V7 at 1 range 0 .. 0;
9075 It is a nuisance to have to rewrite the clause, especially if
9076 the code has to be maintained on both machines. However,
9077 this is a case that we can handle with the
9078 @code{Bit_Order} attribute if it is implemented.
9079 Note that the implementation is not required on byte addressed
9080 machines, but it is indeed implemented in GNAT.
9081 This means that we can simply use the
9082 first record clause, together with the declaration
9084 @smallexample @c ada
9085 for Data'Bit_Order use High_Order_First;
9089 and the effect is what is desired, namely the layout is exactly the same,
9090 independent of whether the code is compiled on a big-endian or little-endian
9093 The important point to understand is that byte ordering is not affected.
9094 A @code{Bit_Order} attribute definition never affects which byte a field
9095 ends up in, only where it ends up in that byte.
9096 To make this clear, let us rewrite the record rep clause of the previous
9099 @smallexample @c ada
9100 for Data'Bit_Order use High_Order_First;
9102 Master_Control at 0 range 0 .. 0;
9103 Master_V1 at 0 range 1 .. 1;
9104 Master_V2 at 0 range 2 .. 2;
9105 Master_V3 at 0 range 3 .. 3;
9106 Master_V4 at 0 range 4 .. 4;
9107 Master_V5 at 0 range 5 .. 5;
9108 Master_V6 at 0 range 6 .. 6;
9109 Master_V7 at 0 range 7 .. 7;
9110 Slave_Control at 0 range 8 .. 8;
9111 Slave_V1 at 0 range 9 .. 9;
9112 Slave_V2 at 0 range 10 .. 10;
9113 Slave_V3 at 0 range 11 .. 11;
9114 Slave_V4 at 0 range 12 .. 12;
9115 Slave_V5 at 0 range 13 .. 13;
9116 Slave_V6 at 0 range 14 .. 14;
9117 Slave_V7 at 0 range 15 .. 15;
9122 This is exactly equivalent to saying (a repeat of the first example):
9124 @smallexample @c ada
9125 for Data'Bit_Order use High_Order_First;
9127 Master_Control at 0 range 0 .. 0;
9128 Master_V1 at 0 range 1 .. 1;
9129 Master_V2 at 0 range 2 .. 2;
9130 Master_V3 at 0 range 3 .. 3;
9131 Master_V4 at 0 range 4 .. 4;
9132 Master_V5 at 0 range 5 .. 5;
9133 Master_V6 at 0 range 6 .. 6;
9134 Master_V7 at 0 range 7 .. 7;
9135 Slave_Control at 1 range 0 .. 0;
9136 Slave_V1 at 1 range 1 .. 1;
9137 Slave_V2 at 1 range 2 .. 2;
9138 Slave_V3 at 1 range 3 .. 3;
9139 Slave_V4 at 1 range 4 .. 4;
9140 Slave_V5 at 1 range 5 .. 5;
9141 Slave_V6 at 1 range 6 .. 6;
9142 Slave_V7 at 1 range 7 .. 7;
9147 Why are they equivalent? Well take a specific field, the @code{Slave_V2}
9148 field. The storage place attributes are obtained by normalizing the
9149 values given so that the @code{First_Bit} value is less than 8. After
9150 normalizing the values (0,10,10) we get (1,2,2) which is exactly what
9151 we specified in the other case.
9153 Now one might expect that the @code{Bit_Order} attribute might affect
9154 bit numbering within the entire record component (two bytes in this
9155 case, thus affecting which byte fields end up in), but that is not
9156 the way this feature is defined, it only affects numbering of bits,
9157 not which byte they end up in.
9159 Consequently it never makes sense to specify a starting bit number
9160 greater than 7 (for a byte addressable field) if an attribute
9161 definition for @code{Bit_Order} has been given, and indeed it
9162 may be actively confusing to specify such a value, so the compiler
9163 generates a warning for such usage.
9165 If you do need to control byte ordering then appropriate conditional
9166 values must be used. If in our example, the slave byte came first on
9167 some machines we might write:
9169 @smallexample @c ada
9170 Master_Byte_First constant Boolean := @dots{};
9172 Master_Byte : constant Natural :=
9173 1 - Boolean'Pos (Master_Byte_First);
9174 Slave_Byte : constant Natural :=
9175 Boolean'Pos (Master_Byte_First);
9177 for Data'Bit_Order use High_Order_First;
9179 Master_Control at Master_Byte range 0 .. 0;
9180 Master_V1 at Master_Byte range 1 .. 1;
9181 Master_V2 at Master_Byte range 2 .. 2;
9182 Master_V3 at Master_Byte range 3 .. 3;
9183 Master_V4 at Master_Byte range 4 .. 4;
9184 Master_V5 at Master_Byte range 5 .. 5;
9185 Master_V6 at Master_Byte range 6 .. 6;
9186 Master_V7 at Master_Byte range 7 .. 7;
9187 Slave_Control at Slave_Byte range 0 .. 0;
9188 Slave_V1 at Slave_Byte range 1 .. 1;
9189 Slave_V2 at Slave_Byte range 2 .. 2;
9190 Slave_V3 at Slave_Byte range 3 .. 3;
9191 Slave_V4 at Slave_Byte range 4 .. 4;
9192 Slave_V5 at Slave_Byte range 5 .. 5;
9193 Slave_V6 at Slave_Byte range 6 .. 6;
9194 Slave_V7 at Slave_Byte range 7 .. 7;
9199 Now to switch between machines, all that is necessary is
9200 to set the boolean constant @code{Master_Byte_First} in
9201 an appropriate manner.
9203 @node Pragma Pack for Arrays
9204 @section Pragma Pack for Arrays
9205 @cindex Pragma Pack (for arrays)
9208 Pragma @code{Pack} applied to an array has no effect unless the component type
9209 is packable. For a component type to be packable, it must be one of the
9216 Any type whose size is specified with a size clause
9218 Any packed array type with a static size
9222 For all these cases, if the component subtype size is in the range
9223 1 through 63, then the effect of the pragma @code{Pack} is exactly as though a
9224 component size were specified giving the component subtype size.
9225 For example if we have:
9227 @smallexample @c ada
9228 type r is range 0 .. 17;
9230 type ar is array (1 .. 8) of r;
9235 Then the component size of @code{ar} will be set to 5 (i.e.@: to @code{r'size},
9236 and the size of the array @code{ar} will be exactly 40 bits.
9238 Note that in some cases this rather fierce approach to packing can produce
9239 unexpected effects. For example, in Ada 95, type Natural typically has a
9240 size of 31, meaning that if you pack an array of Natural, you get 31-bit
9241 close packing, which saves a few bits, but results in far less efficient
9242 access. Since many other Ada compilers will ignore such a packing request,
9243 GNAT will generate a warning on some uses of pragma @code{Pack} that it guesses
9244 might not be what is intended. You can easily remove this warning by
9245 using an explicit @code{Component_Size} setting instead, which never generates
9246 a warning, since the intention of the programmer is clear in this case.
9248 GNAT treats packed arrays in one of two ways. If the size of the array is
9249 known at compile time and is less than 64 bits, then internally the array
9250 is represented as a single modular type, of exactly the appropriate number
9251 of bits. If the length is greater than 63 bits, or is not known at compile
9252 time, then the packed array is represented as an array of bytes, and the
9253 length is always a multiple of 8 bits.
9255 Note that to represent a packed array as a modular type, the alignment must
9256 be suitable for the modular type involved. For example, on typical machines
9257 a 32-bit packed array will be represented by a 32-bit modular integer with
9258 an alignment of four bytes. If you explicitly override the default alignment
9259 with an alignment clause that is too small, the modular representation
9260 cannot be used. For example, consider the following set of declarations:
9262 @smallexample @c ada
9263 type R is range 1 .. 3;
9264 type S is array (1 .. 31) of R;
9265 for S'Component_Size use 2;
9267 for S'Alignment use 1;
9271 If the alignment clause were not present, then a 62-bit modular
9272 representation would be chosen (typically with an alignment of 4 or 8
9273 bytes depending on the target). But the default alignment is overridden
9274 with the explicit alignment clause. This means that the modular
9275 representation cannot be used, and instead the array of bytes
9276 representation must be used, meaning that the length must be a multiple
9277 of 8. Thus the above set of declarations will result in a diagnostic
9278 rejecting the size clause and noting that the minimum size allowed is 64.
9280 @cindex Pragma Pack (for type Natural)
9281 @cindex Pragma Pack warning
9283 One special case that is worth noting occurs when the base type of the
9284 component size is 8/16/32 and the subtype is one bit less. Notably this
9285 occurs with subtype @code{Natural}. Consider:
9287 @smallexample @c ada
9288 type Arr is array (1 .. 32) of Natural;
9293 In all commonly used Ada 83 compilers, this pragma Pack would be ignored,
9294 since typically @code{Natural'Size} is 32 in Ada 83, and in any case most
9295 Ada 83 compilers did not attempt 31 bit packing.
9297 In Ada 95, @code{Natural'Size} is required to be 31. Furthermore, GNAT really
9298 does pack 31-bit subtype to 31 bits. This may result in a substantial
9299 unintended performance penalty when porting legacy Ada 83 code. To help
9300 prevent this, GNAT generates a warning in such cases. If you really want 31
9301 bit packing in a case like this, you can set the component size explicitly:
9303 @smallexample @c ada
9304 type Arr is array (1 .. 32) of Natural;
9305 for Arr'Component_Size use 31;
9309 Here 31-bit packing is achieved as required, and no warning is generated,
9310 since in this case the programmer intention is clear.
9312 @node Pragma Pack for Records
9313 @section Pragma Pack for Records
9314 @cindex Pragma Pack (for records)
9317 Pragma @code{Pack} applied to a record will pack the components to reduce
9318 wasted space from alignment gaps and by reducing the amount of space
9319 taken by components. We distinguish between @emph{packable} components and
9320 @emph{non-packable} components.
9321 Components of the following types are considered packable:
9324 All primitive types are packable.
9327 Small packed arrays, whose size does not exceed 64 bits, and where the
9328 size is statically known at compile time, are represented internally
9329 as modular integers, and so they are also packable.
9334 All packable components occupy the exact number of bits corresponding to
9335 their @code{Size} value, and are packed with no padding bits, i.e.@: they
9336 can start on an arbitrary bit boundary.
9338 All other types are non-packable, they occupy an integral number of
9340 are placed at a boundary corresponding to their alignment requirements.
9342 For example, consider the record
9344 @smallexample @c ada
9345 type Rb1 is array (1 .. 13) of Boolean;
9348 type Rb2 is array (1 .. 65) of Boolean;
9363 The representation for the record x2 is as follows:
9365 @smallexample @c ada
9366 for x2'Size use 224;
9368 l1 at 0 range 0 .. 0;
9369 l2 at 0 range 1 .. 64;
9370 l3 at 12 range 0 .. 31;
9371 l4 at 16 range 0 .. 0;
9372 l5 at 16 range 1 .. 13;
9373 l6 at 18 range 0 .. 71;
9378 Studying this example, we see that the packable fields @code{l1}
9380 of length equal to their sizes, and placed at specific bit boundaries (and
9381 not byte boundaries) to
9382 eliminate padding. But @code{l3} is of a non-packable float type, so
9383 it is on the next appropriate alignment boundary.
9385 The next two fields are fully packable, so @code{l4} and @code{l5} are
9386 minimally packed with no gaps. However, type @code{Rb2} is a packed
9387 array that is longer than 64 bits, so it is itself non-packable. Thus
9388 the @code{l6} field is aligned to the next byte boundary, and takes an
9389 integral number of bytes, i.e.@: 72 bits.
9391 @node Record Representation Clauses
9392 @section Record Representation Clauses
9393 @cindex Record Representation Clause
9396 Record representation clauses may be given for all record types, including
9397 types obtained by record extension. Component clauses are allowed for any
9398 static component. The restrictions on component clauses depend on the type
9401 @cindex Component Clause
9402 For all components of an elementary type, the only restriction on component
9403 clauses is that the size must be at least the 'Size value of the type
9404 (actually the Value_Size). There are no restrictions due to alignment,
9405 and such components may freely cross storage boundaries.
9407 Packed arrays with a size up to and including 64 bits are represented
9408 internally using a modular type with the appropriate number of bits, and
9409 thus the same lack of restriction applies. For example, if you declare:
9411 @smallexample @c ada
9412 type R is array (1 .. 49) of Boolean;
9418 then a component clause for a component of type R may start on any
9419 specified bit boundary, and may specify a value of 49 bits or greater.
9421 The rules for other types are different for GNAT 3 and GNAT 5 versions
9422 (based on GCC 2 and GCC 3 respectively). In GNAT 5, larger components
9423 may also be placed on arbitrary boundaries, so for example, the following
9426 @smallexample @c ada
9427 type R is array (1 .. 79) of Boolean;
9437 G at 0 range 0 .. 0;
9438 H at 0 range 1 .. 1;
9439 L at 0 range 2 .. 80;
9440 R at 0 range 81 .. 159;
9445 In GNAT 3, there are more severe restrictions on larger components.
9446 For non-primitive types, including packed arrays with a size greater than
9447 64 bits, component clauses must respect the alignment requirement of the
9448 type, in particular, always starting on a byte boundary, and the length
9449 must be a multiple of the storage unit.
9451 The following rules regarding tagged types are enforced in both GNAT 3 and
9454 The tag field of a tagged type always occupies an address sized field at
9455 the start of the record. No component clause may attempt to overlay this
9458 In the case of a record extension T1, of a type T, no component clause applied
9459 to the type T1 can specify a storage location that would overlap the first
9460 T'Size bytes of the record.
9462 @node Enumeration Clauses
9463 @section Enumeration Clauses
9465 The only restriction on enumeration clauses is that the range of values
9466 must be representable. For the signed case, if one or more of the
9467 representation values are negative, all values must be in the range:
9469 @smallexample @c ada
9470 System.Min_Int .. System.Max_Int
9474 For the unsigned case, where all values are non negative, the values must
9477 @smallexample @c ada
9478 0 .. System.Max_Binary_Modulus;
9482 A @emph{confirming} representation clause is one in which the values range
9483 from 0 in sequence, i.e.@: a clause that confirms the default representation
9484 for an enumeration type.
9485 Such a confirming representation
9486 is permitted by these rules, and is specially recognized by the compiler so
9487 that no extra overhead results from the use of such a clause.
9489 If an array has an index type which is an enumeration type to which an
9490 enumeration clause has been applied, then the array is stored in a compact
9491 manner. Consider the declarations:
9493 @smallexample @c ada
9494 type r is (A, B, C);
9495 for r use (A => 1, B => 5, C => 10);
9496 type t is array (r) of Character;
9500 The array type t corresponds to a vector with exactly three elements and
9501 has a default size equal to @code{3*Character'Size}. This ensures efficient
9502 use of space, but means that accesses to elements of the array will incur
9503 the overhead of converting representation values to the corresponding
9504 positional values, (i.e.@: the value delivered by the @code{Pos} attribute).
9506 @node Address Clauses
9507 @section Address Clauses
9508 @cindex Address Clause
9510 The reference manual allows a general restriction on representation clauses,
9511 as found in RM 13.1(22):
9514 An implementation need not support representation
9515 items containing nonstatic expressions, except that
9516 an implementation should support a representation item
9517 for a given entity if each nonstatic expression in the
9518 representation item is a name that statically denotes
9519 a constant declared before the entity.
9523 In practice this is applicable only to address clauses, since this is the
9524 only case in which a non-static expression is permitted by the syntax. As
9525 the AARM notes in sections 13.1 (22.a-22.h):
9528 22.a Reason: This is to avoid the following sort of thing:
9530 22.b X : Integer := F(@dots{});
9531 Y : Address := G(@dots{});
9532 for X'Address use Y;
9534 22.c In the above, we have to evaluate the
9535 initialization expression for X before we
9536 know where to put the result. This seems
9537 like an unreasonable implementation burden.
9539 22.d The above code should instead be written
9542 22.e Y : constant Address := G(@dots{});
9543 X : Integer := F(@dots{});
9544 for X'Address use Y;
9546 22.f This allows the expression ``Y'' to be safely
9547 evaluated before X is created.
9549 22.g The constant could be a formal parameter of mode in.
9551 22.h An implementation can support other nonstatic
9552 expressions if it wants to. Expressions of type
9553 Address are hardly ever static, but their value
9554 might be known at compile time anyway in many
9559 GNAT does indeed permit many additional cases of non-static expressions. In
9560 particular, if the type involved is elementary there are no restrictions
9561 (since in this case, holding a temporary copy of the initialization value,
9562 if one is present, is inexpensive). In addition, if there is no implicit or
9563 explicit initialization, then there are no restrictions. GNAT will reject
9564 only the case where all three of these conditions hold:
9569 The type of the item is non-elementary (e.g.@: a record or array).
9572 There is explicit or implicit initialization required for the object.
9573 Note that access values are always implicitly initialized, and also
9574 in GNAT, certain bit-packed arrays (those having a dynamic length or
9575 a length greater than 64) will also be implicitly initialized to zero.
9578 The address value is non-static. Here GNAT is more permissive than the
9579 RM, and allows the address value to be the address of a previously declared
9580 stand-alone variable, as long as it does not itself have an address clause.
9582 @smallexample @c ada
9583 Anchor : Some_Initialized_Type;
9584 Overlay : Some_Initialized_Type;
9585 for Overlay'Address use Anchor'Address;
9589 However, the prefix of the address clause cannot be an array component, or
9590 a component of a discriminated record.
9595 As noted above in section 22.h, address values are typically non-static. In
9596 particular the To_Address function, even if applied to a literal value, is
9597 a non-static function call. To avoid this minor annoyance, GNAT provides
9598 the implementation defined attribute 'To_Address. The following two
9599 expressions have identical values:
9603 @smallexample @c ada
9604 To_Address (16#1234_0000#)
9605 System'To_Address (16#1234_0000#);
9609 except that the second form is considered to be a static expression, and
9610 thus when used as an address clause value is always permitted.
9613 Additionally, GNAT treats as static an address clause that is an
9614 unchecked_conversion of a static integer value. This simplifies the porting
9615 of legacy code, and provides a portable equivalent to the GNAT attribute
9618 Another issue with address clauses is the interaction with alignment
9619 requirements. When an address clause is given for an object, the address
9620 value must be consistent with the alignment of the object (which is usually
9621 the same as the alignment of the type of the object). If an address clause
9622 is given that specifies an inappropriately aligned address value, then the
9623 program execution is erroneous.
9625 Since this source of erroneous behavior can have unfortunate effects, GNAT
9626 checks (at compile time if possible, generating a warning, or at execution
9627 time with a run-time check) that the alignment is appropriate. If the
9628 run-time check fails, then @code{Program_Error} is raised. This run-time
9629 check is suppressed if range checks are suppressed, or if
9630 @code{pragma Restrictions (No_Elaboration_Code)} is in effect.
9633 An address clause cannot be given for an exported object. More
9634 understandably the real restriction is that objects with an address
9635 clause cannot be exported. This is because such variables are not
9636 defined by the Ada program, so there is no external object to export.
9639 It is permissible to give an address clause and a pragma Import for the
9640 same object. In this case, the variable is not really defined by the
9641 Ada program, so there is no external symbol to be linked. The link name
9642 and the external name are ignored in this case. The reason that we allow this
9643 combination is that it provides a useful idiom to avoid unwanted
9644 initializations on objects with address clauses.
9646 When an address clause is given for an object that has implicit or
9647 explicit initialization, then by default initialization takes place. This
9648 means that the effect of the object declaration is to overwrite the
9649 memory at the specified address. This is almost always not what the
9650 programmer wants, so GNAT will output a warning:
9660 for Ext'Address use System'To_Address (16#1234_1234#);
9662 >>> warning: implicit initialization of "Ext" may
9663 modify overlaid storage
9664 >>> warning: use pragma Import for "Ext" to suppress
9665 initialization (RM B(24))
9671 As indicated by the warning message, the solution is to use a (dummy) pragma
9672 Import to suppress this initialization. The pragma tell the compiler that the
9673 object is declared and initialized elsewhere. The following package compiles
9674 without warnings (and the initialization is suppressed):
9676 @smallexample @c ada
9684 for Ext'Address use System'To_Address (16#1234_1234#);
9685 pragma Import (Ada, Ext);
9690 A final issue with address clauses involves their use for overlaying
9691 variables, as in the following example:
9692 @cindex Overlaying of objects
9694 @smallexample @c ada
9697 for B'Address use A'Address;
9701 or alternatively, using the form recommended by the RM:
9703 @smallexample @c ada
9705 Addr : constant Address := A'Address;
9707 for B'Address use Addr;
9711 In both of these cases, @code{A}
9712 and @code{B} become aliased to one another via the
9713 address clause. This use of address clauses to overlay
9714 variables, achieving an effect similar to unchecked
9715 conversion was erroneous in Ada 83, but in Ada 95
9716 the effect is implementation defined. Furthermore, the
9717 Ada 95 RM specifically recommends that in a situation
9718 like this, @code{B} should be subject to the following
9719 implementation advice (RM 13.3(19)):
9722 19 If the Address of an object is specified, or it is imported
9723 or exported, then the implementation should not perform
9724 optimizations based on assumptions of no aliases.
9728 GNAT follows this recommendation, and goes further by also applying
9729 this recommendation to the overlaid variable (@code{A}
9730 in the above example) in this case. This means that the overlay
9731 works "as expected", in that a modification to one of the variables
9732 will affect the value of the other.
9734 @node Effect of Convention on Representation
9735 @section Effect of Convention on Representation
9736 @cindex Convention, effect on representation
9739 Normally the specification of a foreign language convention for a type or
9740 an object has no effect on the chosen representation. In particular, the
9741 representation chosen for data in GNAT generally meets the standard system
9742 conventions, and for example records are laid out in a manner that is
9743 consistent with C@. This means that specifying convention C (for example)
9746 There are three exceptions to this general rule:
9750 @item Convention Fortran and array subtypes
9751 If pragma Convention Fortran is specified for an array subtype, then in
9752 accordance with the implementation advice in section 3.6.2(11) of the
9753 Ada Reference Manual, the array will be stored in a Fortran-compatible
9754 column-major manner, instead of the normal default row-major order.
9756 @item Convention C and enumeration types
9757 GNAT normally stores enumeration types in 8, 16, or 32 bits as required
9758 to accommodate all values of the type. For example, for the enumeration
9761 @smallexample @c ada
9762 type Color is (Red, Green, Blue);
9766 8 bits is sufficient to store all values of the type, so by default, objects
9767 of type @code{Color} will be represented using 8 bits. However, normal C
9768 convention is to use 32 bits for all enum values in C, since enum values
9769 are essentially of type int. If pragma @code{Convention C} is specified for an
9770 Ada enumeration type, then the size is modified as necessary (usually to
9771 32 bits) to be consistent with the C convention for enum values.
9773 @item Convention C/Fortran and Boolean types
9774 In C, the usual convention for boolean values, that is values used for
9775 conditions, is that zero represents false, and nonzero values represent
9776 true. In Ada, the normal convention is that two specific values, typically
9777 0/1, are used to represent false/true respectively.
9779 Fortran has a similar convention for @code{LOGICAL} values (any nonzero
9780 value represents true).
9782 To accommodate the Fortran and C conventions, if a pragma Convention specifies
9783 C or Fortran convention for a derived Boolean, as in the following example:
9785 @smallexample @c ada
9786 type C_Switch is new Boolean;
9787 pragma Convention (C, C_Switch);
9791 then the GNAT generated code will treat any nonzero value as true. For truth
9792 values generated by GNAT, the conventional value 1 will be used for True, but
9793 when one of these values is read, any nonzero value is treated as True.
9797 @node Determining the Representations chosen by GNAT
9798 @section Determining the Representations chosen by GNAT
9799 @cindex Representation, determination of
9800 @cindex @code{-gnatR} switch
9803 Although the descriptions in this section are intended to be complete, it is
9804 often easier to simply experiment to see what GNAT accepts and what the
9805 effect is on the layout of types and objects.
9807 As required by the Ada RM, if a representation clause is not accepted, then
9808 it must be rejected as illegal by the compiler. However, when a
9809 representation clause or pragma is accepted, there can still be questions
9810 of what the compiler actually does. For example, if a partial record
9811 representation clause specifies the location of some components and not
9812 others, then where are the non-specified components placed? Or if pragma
9813 @code{Pack} is used on a record, then exactly where are the resulting
9814 fields placed? The section on pragma @code{Pack} in this chapter can be
9815 used to answer the second question, but it is often easier to just see
9816 what the compiler does.
9818 For this purpose, GNAT provides the option @code{-gnatR}. If you compile
9819 with this option, then the compiler will output information on the actual
9820 representations chosen, in a format similar to source representation
9821 clauses. For example, if we compile the package:
9823 @smallexample @c ada
9825 type r (x : boolean) is tagged record
9827 when True => S : String (1 .. 100);
9832 type r2 is new r (false) with record
9837 y2 at 16 range 0 .. 31;
9844 type x1 is array (1 .. 10) of x;
9845 for x1'component_size use 11;
9847 type ia is access integer;
9849 type Rb1 is array (1 .. 13) of Boolean;
9852 type Rb2 is array (1 .. 65) of Boolean;
9868 using the switch @code{-gnatR} we obtain the following output:
9871 Representation information for unit q
9872 -------------------------------------
9875 for r'Alignment use 4;
9877 x at 4 range 0 .. 7;
9878 _tag at 0 range 0 .. 31;
9879 s at 5 range 0 .. 799;
9882 for r2'Size use 160;
9883 for r2'Alignment use 4;
9885 x at 4 range 0 .. 7;
9886 _tag at 0 range 0 .. 31;
9887 _parent at 0 range 0 .. 63;
9888 y2 at 16 range 0 .. 31;
9892 for x'Alignment use 1;
9894 y at 0 range 0 .. 7;
9897 for x1'Size use 112;
9898 for x1'Alignment use 1;
9899 for x1'Component_Size use 11;
9901 for rb1'Size use 13;
9902 for rb1'Alignment use 2;
9903 for rb1'Component_Size use 1;
9905 for rb2'Size use 72;
9906 for rb2'Alignment use 1;
9907 for rb2'Component_Size use 1;
9909 for x2'Size use 224;
9910 for x2'Alignment use 4;
9912 l1 at 0 range 0 .. 0;
9913 l2 at 0 range 1 .. 64;
9914 l3 at 12 range 0 .. 31;
9915 l4 at 16 range 0 .. 0;
9916 l5 at 16 range 1 .. 13;
9917 l6 at 18 range 0 .. 71;
9922 The Size values are actually the Object_Size, i.e.@: the default size that
9923 will be allocated for objects of the type.
9924 The ?? size for type r indicates that we have a variant record, and the
9925 actual size of objects will depend on the discriminant value.
9927 The Alignment values show the actual alignment chosen by the compiler
9928 for each record or array type.
9930 The record representation clause for type r shows where all fields
9931 are placed, including the compiler generated tag field (whose location
9932 cannot be controlled by the programmer).
9934 The record representation clause for the type extension r2 shows all the
9935 fields present, including the parent field, which is a copy of the fields
9936 of the parent type of r2, i.e.@: r1.
9938 The component size and size clauses for types rb1 and rb2 show
9939 the exact effect of pragma @code{Pack} on these arrays, and the record
9940 representation clause for type x2 shows how pragma @code{Pack} affects
9943 In some cases, it may be useful to cut and paste the representation clauses
9944 generated by the compiler into the original source to fix and guarantee
9945 the actual representation to be used.
9947 @node Standard Library Routines
9948 @chapter Standard Library Routines
9951 The Ada 95 Reference Manual contains in Annex A a full description of an
9952 extensive set of standard library routines that can be used in any Ada
9953 program, and which must be provided by all Ada compilers. They are
9954 analogous to the standard C library used by C programs.
9956 GNAT implements all of the facilities described in annex A, and for most
9957 purposes the description in the Ada 95
9958 reference manual, or appropriate Ada
9959 text book, will be sufficient for making use of these facilities.
9961 In the case of the input-output facilities, @xref{The Implementation of
9962 Standard I/O}, gives details on exactly how GNAT interfaces to the
9963 file system. For the remaining packages, the Ada 95 reference manual
9964 should be sufficient. The following is a list of the packages included,
9965 together with a brief description of the functionality that is provided.
9967 For completeness, references are included to other predefined library
9968 routines defined in other sections of the Ada 95 reference manual (these are
9969 cross-indexed from annex A).
9973 This is a parent package for all the standard library packages. It is
9974 usually included implicitly in your program, and itself contains no
9975 useful data or routines.
9977 @item Ada.Calendar (9.6)
9978 @code{Calendar} provides time of day access, and routines for
9979 manipulating times and durations.
9981 @item Ada.Characters (A.3.1)
9982 This is a dummy parent package that contains no useful entities
9984 @item Ada.Characters.Handling (A.3.2)
9985 This package provides some basic character handling capabilities,
9986 including classification functions for classes of characters (e.g.@: test
9987 for letters, or digits).
9989 @item Ada.Characters.Latin_1 (A.3.3)
9990 This package includes a complete set of definitions of the characters
9991 that appear in type CHARACTER@. It is useful for writing programs that
9992 will run in international environments. For example, if you want an
9993 upper case E with an acute accent in a string, it is often better to use
9994 the definition of @code{UC_E_Acute} in this package. Then your program
9995 will print in an understandable manner even if your environment does not
9996 support these extended characters.
9998 @item Ada.Command_Line (A.15)
9999 This package provides access to the command line parameters and the name
10000 of the current program (analogous to the use of @code{argc} and @code{argv}
10001 in C), and also allows the exit status for the program to be set in a
10002 system-independent manner.
10004 @item Ada.Decimal (F.2)
10005 This package provides constants describing the range of decimal numbers
10006 implemented, and also a decimal divide routine (analogous to the COBOL
10007 verb DIVIDE .. GIVING .. REMAINDER ..)
10009 @item Ada.Direct_IO (A.8.4)
10010 This package provides input-output using a model of a set of records of
10011 fixed-length, containing an arbitrary definite Ada type, indexed by an
10012 integer record number.
10014 @item Ada.Dynamic_Priorities (D.5)
10015 This package allows the priorities of a task to be adjusted dynamically
10016 as the task is running.
10018 @item Ada.Exceptions (11.4.1)
10019 This package provides additional information on exceptions, and also
10020 contains facilities for treating exceptions as data objects, and raising
10021 exceptions with associated messages.
10023 @item Ada.Finalization (7.6)
10024 This package contains the declarations and subprograms to support the
10025 use of controlled types, providing for automatic initialization and
10026 finalization (analogous to the constructors and destructors of C++)
10028 @item Ada.Interrupts (C.3.2)
10029 This package provides facilities for interfacing to interrupts, which
10030 includes the set of signals or conditions that can be raised and
10031 recognized as interrupts.
10033 @item Ada.Interrupts.Names (C.3.2)
10034 This package provides the set of interrupt names (actually signal
10035 or condition names) that can be handled by GNAT@.
10037 @item Ada.IO_Exceptions (A.13)
10038 This package defines the set of exceptions that can be raised by use of
10039 the standard IO packages.
10042 This package contains some standard constants and exceptions used
10043 throughout the numerics packages. Note that the constants pi and e are
10044 defined here, and it is better to use these definitions than rolling
10047 @item Ada.Numerics.Complex_Elementary_Functions
10048 Provides the implementation of standard elementary functions (such as
10049 log and trigonometric functions) operating on complex numbers using the
10050 standard @code{Float} and the @code{Complex} and @code{Imaginary} types
10051 created by the package @code{Numerics.Complex_Types}.
10053 @item Ada.Numerics.Complex_Types
10054 This is a predefined instantiation of
10055 @code{Numerics.Generic_Complex_Types} using @code{Standard.Float} to
10056 build the type @code{Complex} and @code{Imaginary}.
10058 @item Ada.Numerics.Discrete_Random
10059 This package provides a random number generator suitable for generating
10060 random integer values from a specified range.
10062 @item Ada.Numerics.Float_Random
10063 This package provides a random number generator suitable for generating
10064 uniformly distributed floating point values.
10066 @item Ada.Numerics.Generic_Complex_Elementary_Functions
10067 This is a generic version of the package that provides the
10068 implementation of standard elementary functions (such as log and
10069 trigonometric functions) for an arbitrary complex type.
10071 The following predefined instantiations of this package are provided:
10075 @code{Ada.Numerics.Short_Complex_Elementary_Functions}
10077 @code{Ada.Numerics.Complex_Elementary_Functions}
10079 @code{Ada.Numerics.
10080 Long_Complex_Elementary_Functions}
10083 @item Ada.Numerics.Generic_Complex_Types
10084 This is a generic package that allows the creation of complex types,
10085 with associated complex arithmetic operations.
10087 The following predefined instantiations of this package exist
10090 @code{Ada.Numerics.Short_Complex_Complex_Types}
10092 @code{Ada.Numerics.Complex_Complex_Types}
10094 @code{Ada.Numerics.Long_Complex_Complex_Types}
10097 @item Ada.Numerics.Generic_Elementary_Functions
10098 This is a generic package that provides the implementation of standard
10099 elementary functions (such as log an trigonometric functions) for an
10100 arbitrary float type.
10102 The following predefined instantiations of this package exist
10106 @code{Ada.Numerics.Short_Elementary_Functions}
10108 @code{Ada.Numerics.Elementary_Functions}
10110 @code{Ada.Numerics.Long_Elementary_Functions}
10113 @item Ada.Real_Time (D.8)
10114 This package provides facilities similar to those of @code{Calendar}, but
10115 operating with a finer clock suitable for real time control. Note that
10116 annex D requires that there be no backward clock jumps, and GNAT generally
10117 guarantees this behavior, but of course if the external clock on which
10118 the GNAT runtime depends is deliberately reset by some external event,
10119 then such a backward jump may occur.
10121 @item Ada.Sequential_IO (A.8.1)
10122 This package provides input-output facilities for sequential files,
10123 which can contain a sequence of values of a single type, which can be
10124 any Ada type, including indefinite (unconstrained) types.
10126 @item Ada.Storage_IO (A.9)
10127 This package provides a facility for mapping arbitrary Ada types to and
10128 from a storage buffer. It is primarily intended for the creation of new
10131 @item Ada.Streams (13.13.1)
10132 This is a generic package that provides the basic support for the
10133 concept of streams as used by the stream attributes (@code{Input},
10134 @code{Output}, @code{Read} and @code{Write}).
10136 @item Ada.Streams.Stream_IO (A.12.1)
10137 This package is a specialization of the type @code{Streams} defined in
10138 package @code{Streams} together with a set of operations providing
10139 Stream_IO capability. The Stream_IO model permits both random and
10140 sequential access to a file which can contain an arbitrary set of values
10141 of one or more Ada types.
10143 @item Ada.Strings (A.4.1)
10144 This package provides some basic constants used by the string handling
10147 @item Ada.Strings.Bounded (A.4.4)
10148 This package provides facilities for handling variable length
10149 strings. The bounded model requires a maximum length. It is thus
10150 somewhat more limited than the unbounded model, but avoids the use of
10151 dynamic allocation or finalization.
10153 @item Ada.Strings.Fixed (A.4.3)
10154 This package provides facilities for handling fixed length strings.
10156 @item Ada.Strings.Maps (A.4.2)
10157 This package provides facilities for handling character mappings and
10158 arbitrarily defined subsets of characters. For instance it is useful in
10159 defining specialized translation tables.
10161 @item Ada.Strings.Maps.Constants (A.4.6)
10162 This package provides a standard set of predefined mappings and
10163 predefined character sets. For example, the standard upper to lower case
10164 conversion table is found in this package. Note that upper to lower case
10165 conversion is non-trivial if you want to take the entire set of
10166 characters, including extended characters like E with an acute accent,
10167 into account. You should use the mappings in this package (rather than
10168 adding 32 yourself) to do case mappings.
10170 @item Ada.Strings.Unbounded (A.4.5)
10171 This package provides facilities for handling variable length
10172 strings. The unbounded model allows arbitrary length strings, but
10173 requires the use of dynamic allocation and finalization.
10175 @item Ada.Strings.Wide_Bounded (A.4.7)
10176 @itemx Ada.Strings.Wide_Fixed (A.4.7)
10177 @itemx Ada.Strings.Wide_Maps (A.4.7)
10178 @itemx Ada.Strings.Wide_Maps.Constants (A.4.7)
10179 @itemx Ada.Strings.Wide_Unbounded (A.4.7)
10180 These packages provide analogous capabilities to the corresponding
10181 packages without @samp{Wide_} in the name, but operate with the types
10182 @code{Wide_String} and @code{Wide_Character} instead of @code{String}
10183 and @code{Character}.
10185 @item Ada.Synchronous_Task_Control (D.10)
10186 This package provides some standard facilities for controlling task
10187 communication in a synchronous manner.
10190 This package contains definitions for manipulation of the tags of tagged
10193 @item Ada.Task_Attributes
10194 This package provides the capability of associating arbitrary
10195 task-specific data with separate tasks.
10198 This package provides basic text input-output capabilities for
10199 character, string and numeric data. The subpackages of this
10200 package are listed next.
10202 @item Ada.Text_IO.Decimal_IO
10203 Provides input-output facilities for decimal fixed-point types
10205 @item Ada.Text_IO.Enumeration_IO
10206 Provides input-output facilities for enumeration types.
10208 @item Ada.Text_IO.Fixed_IO
10209 Provides input-output facilities for ordinary fixed-point types.
10211 @item Ada.Text_IO.Float_IO
10212 Provides input-output facilities for float types. The following
10213 predefined instantiations of this generic package are available:
10217 @code{Short_Float_Text_IO}
10219 @code{Float_Text_IO}
10221 @code{Long_Float_Text_IO}
10224 @item Ada.Text_IO.Integer_IO
10225 Provides input-output facilities for integer types. The following
10226 predefined instantiations of this generic package are available:
10229 @item Short_Short_Integer
10230 @code{Ada.Short_Short_Integer_Text_IO}
10231 @item Short_Integer
10232 @code{Ada.Short_Integer_Text_IO}
10234 @code{Ada.Integer_Text_IO}
10236 @code{Ada.Long_Integer_Text_IO}
10237 @item Long_Long_Integer
10238 @code{Ada.Long_Long_Integer_Text_IO}
10241 @item Ada.Text_IO.Modular_IO
10242 Provides input-output facilities for modular (unsigned) types
10244 @item Ada.Text_IO.Complex_IO (G.1.3)
10245 This package provides basic text input-output capabilities for complex
10248 @item Ada.Text_IO.Editing (F.3.3)
10249 This package contains routines for edited output, analogous to the use
10250 of pictures in COBOL@. The picture formats used by this package are a
10251 close copy of the facility in COBOL@.
10253 @item Ada.Text_IO.Text_Streams (A.12.2)
10254 This package provides a facility that allows Text_IO files to be treated
10255 as streams, so that the stream attributes can be used for writing
10256 arbitrary data, including binary data, to Text_IO files.
10258 @item Ada.Unchecked_Conversion (13.9)
10259 This generic package allows arbitrary conversion from one type to
10260 another of the same size, providing for breaking the type safety in
10261 special circumstances.
10263 If the types have the same Size (more accurately the same Value_Size),
10264 then the effect is simply to transfer the bits from the source to the
10265 target type without any modification. This usage is well defined, and
10266 for simple types whose representation is typically the same across
10267 all implementations, gives a portable method of performing such
10270 If the types do not have the same size, then the result is implementation
10271 defined, and thus may be non-portable. The following describes how GNAT
10272 handles such unchecked conversion cases.
10274 If the types are of different sizes, and are both discrete types, then
10275 the effect is of a normal type conversion without any constraint checking.
10276 In particular if the result type has a larger size, the result will be
10277 zero or sign extended. If the result type has a smaller size, the result
10278 will be truncated by ignoring high order bits.
10280 If the types are of different sizes, and are not both discrete types,
10281 then the conversion works as though pointers were created to the source
10282 and target, and the pointer value is converted. The effect is that bits
10283 are copied from successive low order storage units and bits of the source
10284 up to the length of the target type.
10286 A warning is issued if the lengths differ, since the effect in this
10287 case is implementation dependent, and the above behavior may not match
10288 that of some other compiler.
10290 A pointer to one type may be converted to a pointer to another type using
10291 unchecked conversion. The only case in which the effect is undefined is
10292 when one or both pointers are pointers to unconstrained array types. In
10293 this case, the bounds information may get incorrectly transferred, and in
10294 particular, GNAT uses double size pointers for such types, and it is
10295 meaningless to convert between such pointer types. GNAT will issue a
10296 warning if the alignment of the target designated type is more strict
10297 than the alignment of the source designated type (since the result may
10298 be unaligned in this case).
10300 A pointer other than a pointer to an unconstrained array type may be
10301 converted to and from System.Address. Such usage is common in Ada 83
10302 programs, but note that Ada.Address_To_Access_Conversions is the
10303 preferred method of performing such conversions in Ada 95. Neither
10304 unchecked conversion nor Ada.Address_To_Access_Conversions should be
10305 used in conjunction with pointers to unconstrained objects, since
10306 the bounds information cannot be handled correctly in this case.
10308 @item Ada.Unchecked_Deallocation (13.11.2)
10309 This generic package allows explicit freeing of storage previously
10310 allocated by use of an allocator.
10312 @item Ada.Wide_Text_IO (A.11)
10313 This package is similar to @code{Ada.Text_IO}, except that the external
10314 file supports wide character representations, and the internal types are
10315 @code{Wide_Character} and @code{Wide_String} instead of @code{Character}
10316 and @code{String}. It contains generic subpackages listed next.
10318 @item Ada.Wide_Text_IO.Decimal_IO
10319 Provides input-output facilities for decimal fixed-point types
10321 @item Ada.Wide_Text_IO.Enumeration_IO
10322 Provides input-output facilities for enumeration types.
10324 @item Ada.Wide_Text_IO.Fixed_IO
10325 Provides input-output facilities for ordinary fixed-point types.
10327 @item Ada.Wide_Text_IO.Float_IO
10328 Provides input-output facilities for float types. The following
10329 predefined instantiations of this generic package are available:
10333 @code{Short_Float_Wide_Text_IO}
10335 @code{Float_Wide_Text_IO}
10337 @code{Long_Float_Wide_Text_IO}
10340 @item Ada.Wide_Text_IO.Integer_IO
10341 Provides input-output facilities for integer types. The following
10342 predefined instantiations of this generic package are available:
10345 @item Short_Short_Integer
10346 @code{Ada.Short_Short_Integer_Wide_Text_IO}
10347 @item Short_Integer
10348 @code{Ada.Short_Integer_Wide_Text_IO}
10350 @code{Ada.Integer_Wide_Text_IO}
10352 @code{Ada.Long_Integer_Wide_Text_IO}
10353 @item Long_Long_Integer
10354 @code{Ada.Long_Long_Integer_Wide_Text_IO}
10357 @item Ada.Wide_Text_IO.Modular_IO
10358 Provides input-output facilities for modular (unsigned) types
10360 @item Ada.Wide_Text_IO.Complex_IO (G.1.3)
10361 This package is similar to @code{Ada.Text_IO.Complex_IO}, except that the
10362 external file supports wide character representations.
10364 @item Ada.Wide_Text_IO.Editing (F.3.4)
10365 This package is similar to @code{Ada.Text_IO.Editing}, except that the
10366 types are @code{Wide_Character} and @code{Wide_String} instead of
10367 @code{Character} and @code{String}.
10369 @item Ada.Wide_Text_IO.Streams (A.12.3)
10370 This package is similar to @code{Ada.Text_IO.Streams}, except that the
10371 types are @code{Wide_Character} and @code{Wide_String} instead of
10372 @code{Character} and @code{String}.
10375 @node The Implementation of Standard I/O
10376 @chapter The Implementation of Standard I/O
10379 GNAT implements all the required input-output facilities described in
10380 A.6 through A.14. These sections of the Ada 95 reference manual describe the
10381 required behavior of these packages from the Ada point of view, and if
10382 you are writing a portable Ada program that does not need to know the
10383 exact manner in which Ada maps to the outside world when it comes to
10384 reading or writing external files, then you do not need to read this
10385 chapter. As long as your files are all regular files (not pipes or
10386 devices), and as long as you write and read the files only from Ada, the
10387 description in the Ada 95 reference manual is sufficient.
10389 However, if you want to do input-output to pipes or other devices, such
10390 as the keyboard or screen, or if the files you are dealing with are
10391 either generated by some other language, or to be read by some other
10392 language, then you need to know more about the details of how the GNAT
10393 implementation of these input-output facilities behaves.
10395 In this chapter we give a detailed description of exactly how GNAT
10396 interfaces to the file system. As always, the sources of the system are
10397 available to you for answering questions at an even more detailed level,
10398 but for most purposes the information in this chapter will suffice.
10400 Another reason that you may need to know more about how input-output is
10401 implemented arises when you have a program written in mixed languages
10402 where, for example, files are shared between the C and Ada sections of
10403 the same program. GNAT provides some additional facilities, in the form
10404 of additional child library packages, that facilitate this sharing, and
10405 these additional facilities are also described in this chapter.
10408 * Standard I/O Packages::
10417 * Operations on C Streams::
10418 * Interfacing to C Streams::
10421 @node Standard I/O Packages
10422 @section Standard I/O Packages
10425 The Standard I/O packages described in Annex A for
10431 Ada.Text_IO.Complex_IO
10433 Ada.Text_IO.Text_Streams,
10437 Ada.Wide_Text_IO.Complex_IO,
10439 Ada.Wide_Text_IO.Text_Streams
10449 are implemented using the C
10450 library streams facility; where
10454 All files are opened using @code{fopen}.
10456 All input/output operations use @code{fread}/@code{fwrite}.
10460 There is no internal buffering of any kind at the Ada library level. The
10461 only buffering is that provided at the system level in the
10462 implementation of the C library routines that support streams. This
10463 facilitates shared use of these streams by mixed language programs.
10466 @section FORM Strings
10469 The format of a FORM string in GNAT is:
10472 "keyword=value,keyword=value,@dots{},keyword=value"
10476 where letters may be in upper or lower case, and there are no spaces
10477 between values. The order of the entries is not important. Currently
10478 there are two keywords defined.
10486 The use of these parameters is described later in this section.
10492 Direct_IO can only be instantiated for definite types. This is a
10493 restriction of the Ada language, which means that the records are fixed
10494 length (the length being determined by @code{@var{type}'Size}, rounded
10495 up to the next storage unit boundary if necessary).
10497 The records of a Direct_IO file are simply written to the file in index
10498 sequence, with the first record starting at offset zero, and subsequent
10499 records following. There is no control information of any kind. For
10500 example, if 32-bit integers are being written, each record takes
10501 4-bytes, so the record at index @var{K} starts at offset
10502 (@var{K}@minus{}1)*4.
10504 There is no limit on the size of Direct_IO files, they are expanded as
10505 necessary to accommodate whatever records are written to the file.
10507 @node Sequential_IO
10508 @section Sequential_IO
10511 Sequential_IO may be instantiated with either a definite (constrained)
10512 or indefinite (unconstrained) type.
10514 For the definite type case, the elements written to the file are simply
10515 the memory images of the data values with no control information of any
10516 kind. The resulting file should be read using the same type, no validity
10517 checking is performed on input.
10519 For the indefinite type case, the elements written consist of two
10520 parts. First is the size of the data item, written as the memory image
10521 of a @code{Interfaces.C.size_t} value, followed by the memory image of
10522 the data value. The resulting file can only be read using the same
10523 (unconstrained) type. Normal assignment checks are performed on these
10524 read operations, and if these checks fail, @code{Data_Error} is
10525 raised. In particular, in the array case, the lengths must match, and in
10526 the variant record case, if the variable for a particular read operation
10527 is constrained, the discriminants must match.
10529 Note that it is not possible to use Sequential_IO to write variable
10530 length array items, and then read the data back into different length
10531 arrays. For example, the following will raise @code{Data_Error}:
10533 @smallexample @c ada
10534 package IO is new Sequential_IO (String);
10539 IO.Write (F, "hello!")
10540 IO.Reset (F, Mode=>In_File);
10547 On some Ada implementations, this will print @code{hell}, but the program is
10548 clearly incorrect, since there is only one element in the file, and that
10549 element is the string @code{hello!}.
10551 In Ada 95, this kind of behavior can be legitimately achieved using
10552 Stream_IO, and this is the preferred mechanism. In particular, the above
10553 program fragment rewritten to use Stream_IO will work correctly.
10559 Text_IO files consist of a stream of characters containing the following
10560 special control characters:
10563 LF (line feed, 16#0A#) Line Mark
10564 FF (form feed, 16#0C#) Page Mark
10568 A canonical Text_IO file is defined as one in which the following
10569 conditions are met:
10573 The character @code{LF} is used only as a line mark, i.e.@: to mark the end
10577 The character @code{FF} is used only as a page mark, i.e.@: to mark the
10578 end of a page and consequently can appear only immediately following a
10579 @code{LF} (line mark) character.
10582 The file ends with either @code{LF} (line mark) or @code{LF}-@code{FF}
10583 (line mark, page mark). In the former case, the page mark is implicitly
10584 assumed to be present.
10588 A file written using Text_IO will be in canonical form provided that no
10589 explicit @code{LF} or @code{FF} characters are written using @code{Put}
10590 or @code{Put_Line}. There will be no @code{FF} character at the end of
10591 the file unless an explicit @code{New_Page} operation was performed
10592 before closing the file.
10594 A canonical Text_IO file that is a regular file, i.e.@: not a device or a
10595 pipe, can be read using any of the routines in Text_IO@. The
10596 semantics in this case will be exactly as defined in the Ada 95 reference
10597 manual and all the routines in Text_IO are fully implemented.
10599 A text file that does not meet the requirements for a canonical Text_IO
10600 file has one of the following:
10604 The file contains @code{FF} characters not immediately following a
10605 @code{LF} character.
10608 The file contains @code{LF} or @code{FF} characters written by
10609 @code{Put} or @code{Put_Line}, which are not logically considered to be
10610 line marks or page marks.
10613 The file ends in a character other than @code{LF} or @code{FF},
10614 i.e.@: there is no explicit line mark or page mark at the end of the file.
10618 Text_IO can be used to read such non-standard text files but subprograms
10619 to do with line or page numbers do not have defined meanings. In
10620 particular, a @code{FF} character that does not follow a @code{LF}
10621 character may or may not be treated as a page mark from the point of
10622 view of page and line numbering. Every @code{LF} character is considered
10623 to end a line, and there is an implied @code{LF} character at the end of
10627 * Text_IO Stream Pointer Positioning::
10628 * Text_IO Reading and Writing Non-Regular Files::
10630 * Treating Text_IO Files as Streams::
10631 * Text_IO Extensions::
10632 * Text_IO Facilities for Unbounded Strings::
10635 @node Text_IO Stream Pointer Positioning
10636 @subsection Stream Pointer Positioning
10639 @code{Ada.Text_IO} has a definition of current position for a file that
10640 is being read. No internal buffering occurs in Text_IO, and usually the
10641 physical position in the stream used to implement the file corresponds
10642 to this logical position defined by Text_IO@. There are two exceptions:
10646 After a call to @code{End_Of_Page} that returns @code{True}, the stream
10647 is positioned past the @code{LF} (line mark) that precedes the page
10648 mark. Text_IO maintains an internal flag so that subsequent read
10649 operations properly handle the logical position which is unchanged by
10650 the @code{End_Of_Page} call.
10653 After a call to @code{End_Of_File} that returns @code{True}, if the
10654 Text_IO file was positioned before the line mark at the end of file
10655 before the call, then the logical position is unchanged, but the stream
10656 is physically positioned right at the end of file (past the line mark,
10657 and past a possible page mark following the line mark. Again Text_IO
10658 maintains internal flags so that subsequent read operations properly
10659 handle the logical position.
10663 These discrepancies have no effect on the observable behavior of
10664 Text_IO, but if a single Ada stream is shared between a C program and
10665 Ada program, or shared (using @samp{shared=yes} in the form string)
10666 between two Ada files, then the difference may be observable in some
10669 @node Text_IO Reading and Writing Non-Regular Files
10670 @subsection Reading and Writing Non-Regular Files
10673 A non-regular file is a device (such as a keyboard), or a pipe. Text_IO
10674 can be used for reading and writing. Writing is not affected and the
10675 sequence of characters output is identical to the normal file case, but
10676 for reading, the behavior of Text_IO is modified to avoid undesirable
10677 look-ahead as follows:
10679 An input file that is not a regular file is considered to have no page
10680 marks. Any @code{Ascii.FF} characters (the character normally used for a
10681 page mark) appearing in the file are considered to be data
10682 characters. In particular:
10686 @code{Get_Line} and @code{Skip_Line} do not test for a page mark
10687 following a line mark. If a page mark appears, it will be treated as a
10691 This avoids the need to wait for an extra character to be typed or
10692 entered from the pipe to complete one of these operations.
10695 @code{End_Of_Page} always returns @code{False}
10698 @code{End_Of_File} will return @code{False} if there is a page mark at
10699 the end of the file.
10703 Output to non-regular files is the same as for regular files. Page marks
10704 may be written to non-regular files using @code{New_Page}, but as noted
10705 above they will not be treated as page marks on input if the output is
10706 piped to another Ada program.
10708 Another important discrepancy when reading non-regular files is that the end
10709 of file indication is not ``sticky''. If an end of file is entered, e.g.@: by
10710 pressing the @key{EOT} key,
10712 is signaled once (i.e.@: the test @code{End_Of_File}
10713 will yield @code{True}, or a read will
10714 raise @code{End_Error}), but then reading can resume
10715 to read data past that end of
10716 file indication, until another end of file indication is entered.
10718 @node Get_Immediate
10719 @subsection Get_Immediate
10720 @cindex Get_Immediate
10723 Get_Immediate returns the next character (including control characters)
10724 from the input file. In particular, Get_Immediate will return LF or FF
10725 characters used as line marks or page marks. Such operations leave the
10726 file positioned past the control character, and it is thus not treated
10727 as having its normal function. This means that page, line and column
10728 counts after this kind of Get_Immediate call are set as though the mark
10729 did not occur. In the case where a Get_Immediate leaves the file
10730 positioned between the line mark and page mark (which is not normally
10731 possible), it is undefined whether the FF character will be treated as a
10734 @node Treating Text_IO Files as Streams
10735 @subsection Treating Text_IO Files as Streams
10736 @cindex Stream files
10739 The package @code{Text_IO.Streams} allows a Text_IO file to be treated
10740 as a stream. Data written to a Text_IO file in this stream mode is
10741 binary data. If this binary data contains bytes 16#0A# (@code{LF}) or
10742 16#0C# (@code{FF}), the resulting file may have non-standard
10743 format. Similarly if read operations are used to read from a Text_IO
10744 file treated as a stream, then @code{LF} and @code{FF} characters may be
10745 skipped and the effect is similar to that described above for
10746 @code{Get_Immediate}.
10748 @node Text_IO Extensions
10749 @subsection Text_IO Extensions
10750 @cindex Text_IO extensions
10753 A package GNAT.IO_Aux in the GNAT library provides some useful extensions
10754 to the standard @code{Text_IO} package:
10757 @item function File_Exists (Name : String) return Boolean;
10758 Determines if a file of the given name exists.
10760 @item function Get_Line return String;
10761 Reads a string from the standard input file. The value returned is exactly
10762 the length of the line that was read.
10764 @item function Get_Line (File : Ada.Text_IO.File_Type) return String;
10765 Similar, except that the parameter File specifies the file from which
10766 the string is to be read.
10770 @node Text_IO Facilities for Unbounded Strings
10771 @subsection Text_IO Facilities for Unbounded Strings
10772 @cindex Text_IO for unbounded strings
10773 @cindex Unbounded_String, Text_IO operations
10776 The package @code{Ada.Strings.Unbounded.Text_IO}
10777 in library files @code{a-suteio.ads/adb} contains some GNAT-specific
10778 subprograms useful for Text_IO operations on unbounded strings:
10782 @item function Get_Line (File : File_Type) return Unbounded_String;
10783 Reads a line from the specified file
10784 and returns the result as an unbounded string.
10786 @item procedure Put (File : File_Type; U : Unbounded_String);
10787 Writes the value of the given unbounded string to the specified file
10788 Similar to the effect of
10789 @code{Put (To_String (U))} except that an extra copy is avoided.
10791 @item procedure Put_Line (File : File_Type; U : Unbounded_String);
10792 Writes the value of the given unbounded string to the specified file,
10793 followed by a @code{New_Line}.
10794 Similar to the effect of @code{Put_Line (To_String (U))} except
10795 that an extra copy is avoided.
10799 In the above procedures, @code{File} is of type @code{Ada.Text_IO.File_Type}
10800 and is optional. If the parameter is omitted, then the standard input or
10801 output file is referenced as appropriate.
10803 The package @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} in library
10804 files @file{a-swuwti.ads} and @file{a-swuwti.adb} provides similar extended
10805 @code{Wide_Text_IO} functionality for unbounded wide strings.
10808 @section Wide_Text_IO
10811 @code{Wide_Text_IO} is similar in most respects to Text_IO, except that
10812 both input and output files may contain special sequences that represent
10813 wide character values. The encoding scheme for a given file may be
10814 specified using a FORM parameter:
10821 as part of the FORM string (WCEM = wide character encoding method),
10822 where @var{x} is one of the following characters
10828 Upper half encoding
10840 The encoding methods match those that
10841 can be used in a source
10842 program, but there is no requirement that the encoding method used for
10843 the source program be the same as the encoding method used for files,
10844 and different files may use different encoding methods.
10846 The default encoding method for the standard files, and for opened files
10847 for which no WCEM parameter is given in the FORM string matches the
10848 wide character encoding specified for the main program (the default
10849 being brackets encoding if no coding method was specified with -gnatW).
10853 In this encoding, a wide character is represented by a five character
10861 where @var{a}, @var{b}, @var{c}, @var{d} are the four hexadecimal
10862 characters (using upper case letters) of the wide character code. For
10863 example, ESC A345 is used to represent the wide character with code
10864 16#A345#. This scheme is compatible with use of the full
10865 @code{Wide_Character} set.
10867 @item Upper Half Coding
10868 The wide character with encoding 16#abcd#, where the upper bit is on
10869 (i.e.@: a is in the range 8-F) is represented as two bytes 16#ab# and
10870 16#cd#. The second byte may never be a format control character, but is
10871 not required to be in the upper half. This method can be also used for
10872 shift-JIS or EUC where the internal coding matches the external coding.
10874 @item Shift JIS Coding
10875 A wide character is represented by a two character sequence 16#ab# and
10876 16#cd#, with the restrictions described for upper half encoding as
10877 described above. The internal character code is the corresponding JIS
10878 character according to the standard algorithm for Shift-JIS
10879 conversion. Only characters defined in the JIS code set table can be
10880 used with this encoding method.
10883 A wide character is represented by a two character sequence 16#ab# and
10884 16#cd#, with both characters being in the upper half. The internal
10885 character code is the corresponding JIS character according to the EUC
10886 encoding algorithm. Only characters defined in the JIS code set table
10887 can be used with this encoding method.
10890 A wide character is represented using
10891 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
10892 10646-1/Am.2. Depending on the character value, the representation
10893 is a one, two, or three byte sequence:
10896 16#0000#-16#007f#: 2#0xxxxxxx#
10897 16#0080#-16#07ff#: 2#110xxxxx# 2#10xxxxxx#
10898 16#0800#-16#ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
10902 where the xxx bits correspond to the left-padded bits of the
10903 16-bit character value. Note that all lower half ASCII characters
10904 are represented as ASCII bytes and all upper half characters and
10905 other wide characters are represented as sequences of upper-half
10906 (The full UTF-8 scheme allows for encoding 31-bit characters as
10907 6-byte sequences, but in this implementation, all UTF-8 sequences
10908 of four or more bytes length will raise a Constraint_Error, as
10909 will all invalid UTF-8 sequences.)
10911 @item Brackets Coding
10912 In this encoding, a wide character is represented by the following eight
10913 character sequence:
10920 where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
10921 characters (using uppercase letters) of the wide character code. For
10922 example, @code{["A345"]} is used to represent the wide character with code
10924 This scheme is compatible with use of the full Wide_Character set.
10925 On input, brackets coding can also be used for upper half characters,
10926 e.g.@: @code{["C1"]} for lower case a. However, on output, brackets notation
10927 is only used for wide characters with a code greater than @code{16#FF#}.
10932 For the coding schemes other than Hex and Brackets encoding,
10933 not all wide character
10934 values can be represented. An attempt to output a character that cannot
10935 be represented using the encoding scheme for the file causes
10936 Constraint_Error to be raised. An invalid wide character sequence on
10937 input also causes Constraint_Error to be raised.
10940 * Wide_Text_IO Stream Pointer Positioning::
10941 * Wide_Text_IO Reading and Writing Non-Regular Files::
10944 @node Wide_Text_IO Stream Pointer Positioning
10945 @subsection Stream Pointer Positioning
10948 @code{Ada.Wide_Text_IO} is similar to @code{Ada.Text_IO} in its handling
10949 of stream pointer positioning (@pxref{Text_IO}). There is one additional
10952 If @code{Ada.Wide_Text_IO.Look_Ahead} reads a character outside the
10953 normal lower ASCII set (i.e.@: a character in the range:
10955 @smallexample @c ada
10956 Wide_Character'Val (16#0080#) .. Wide_Character'Val (16#FFFF#)
10960 then although the logical position of the file pointer is unchanged by
10961 the @code{Look_Ahead} call, the stream is physically positioned past the
10962 wide character sequence. Again this is to avoid the need for buffering
10963 or backup, and all @code{Wide_Text_IO} routines check the internal
10964 indication that this situation has occurred so that this is not visible
10965 to a normal program using @code{Wide_Text_IO}. However, this discrepancy
10966 can be observed if the wide text file shares a stream with another file.
10968 @node Wide_Text_IO Reading and Writing Non-Regular Files
10969 @subsection Reading and Writing Non-Regular Files
10972 As in the case of Text_IO, when a non-regular file is read, it is
10973 assumed that the file contains no page marks (any form characters are
10974 treated as data characters), and @code{End_Of_Page} always returns
10975 @code{False}. Similarly, the end of file indication is not sticky, so
10976 it is possible to read beyond an end of file.
10982 A stream file is a sequence of bytes, where individual elements are
10983 written to the file as described in the Ada 95 reference manual. The type
10984 @code{Stream_Element} is simply a byte. There are two ways to read or
10985 write a stream file.
10989 The operations @code{Read} and @code{Write} directly read or write a
10990 sequence of stream elements with no control information.
10993 The stream attributes applied to a stream file transfer data in the
10994 manner described for stream attributes.
10998 @section Shared Files
11001 Section A.14 of the Ada 95 Reference Manual allows implementations to
11002 provide a wide variety of behavior if an attempt is made to access the
11003 same external file with two or more internal files.
11005 To provide a full range of functionality, while at the same time
11006 minimizing the problems of portability caused by this implementation
11007 dependence, GNAT handles file sharing as follows:
11011 In the absence of a @samp{shared=@var{xxx}} form parameter, an attempt
11012 to open two or more files with the same full name is considered an error
11013 and is not supported. The exception @code{Use_Error} will be
11014 raised. Note that a file that is not explicitly closed by the program
11015 remains open until the program terminates.
11018 If the form parameter @samp{shared=no} appears in the form string, the
11019 file can be opened or created with its own separate stream identifier,
11020 regardless of whether other files sharing the same external file are
11021 opened. The exact effect depends on how the C stream routines handle
11022 multiple accesses to the same external files using separate streams.
11025 If the form parameter @samp{shared=yes} appears in the form string for
11026 each of two or more files opened using the same full name, the same
11027 stream is shared between these files, and the semantics are as described
11028 in Ada 95 Reference Manual, Section A.14.
11032 When a program that opens multiple files with the same name is ported
11033 from another Ada compiler to GNAT, the effect will be that
11034 @code{Use_Error} is raised.
11036 The documentation of the original compiler and the documentation of the
11037 program should then be examined to determine if file sharing was
11038 expected, and @samp{shared=@var{xxx}} parameters added to @code{Open}
11039 and @code{Create} calls as required.
11041 When a program is ported from GNAT to some other Ada compiler, no
11042 special attention is required unless the @samp{shared=@var{xxx}} form
11043 parameter is used in the program. In this case, you must examine the
11044 documentation of the new compiler to see if it supports the required
11045 file sharing semantics, and form strings modified appropriately. Of
11046 course it may be the case that the program cannot be ported if the
11047 target compiler does not support the required functionality. The best
11048 approach in writing portable code is to avoid file sharing (and hence
11049 the use of the @samp{shared=@var{xxx}} parameter in the form string)
11052 One common use of file sharing in Ada 83 is the use of instantiations of
11053 Sequential_IO on the same file with different types, to achieve
11054 heterogeneous input-output. Although this approach will work in GNAT if
11055 @samp{shared=yes} is specified, it is preferable in Ada 95 to use Stream_IO
11056 for this purpose (using the stream attributes)
11059 @section Open Modes
11062 @code{Open} and @code{Create} calls result in a call to @code{fopen}
11063 using the mode shown in the following table:
11066 @center @code{Open} and @code{Create} Call Modes
11068 @b{OPEN } @b{CREATE}
11069 Append_File "r+" "w+"
11071 Out_File (Direct_IO) "r+" "w"
11072 Out_File (all other cases) "w" "w"
11073 Inout_File "r+" "w+"
11077 If text file translation is required, then either @samp{b} or @samp{t}
11078 is added to the mode, depending on the setting of Text. Text file
11079 translation refers to the mapping of CR/LF sequences in an external file
11080 to LF characters internally. This mapping only occurs in DOS and
11081 DOS-like systems, and is not relevant to other systems.
11083 A special case occurs with Stream_IO@. As shown in the above table, the
11084 file is initially opened in @samp{r} or @samp{w} mode for the
11085 @code{In_File} and @code{Out_File} cases. If a @code{Set_Mode} operation
11086 subsequently requires switching from reading to writing or vice-versa,
11087 then the file is reopened in @samp{r+} mode to permit the required operation.
11089 @node Operations on C Streams
11090 @section Operations on C Streams
11091 The package @code{Interfaces.C_Streams} provides an Ada program with direct
11092 access to the C library functions for operations on C streams:
11094 @smallexample @c adanocomment
11095 package Interfaces.C_Streams is
11096 -- Note: the reason we do not use the types that are in
11097 -- Interfaces.C is that we want to avoid dragging in the
11098 -- code in this unit if possible.
11099 subtype chars is System.Address;
11100 -- Pointer to null-terminated array of characters
11101 subtype FILEs is System.Address;
11102 -- Corresponds to the C type FILE*
11103 subtype voids is System.Address;
11104 -- Corresponds to the C type void*
11105 subtype int is Integer;
11106 subtype long is Long_Integer;
11107 -- Note: the above types are subtypes deliberately, and it
11108 -- is part of this spec that the above correspondences are
11109 -- guaranteed. This means that it is legitimate to, for
11110 -- example, use Integer instead of int. We provide these
11111 -- synonyms for clarity, but in some cases it may be
11112 -- convenient to use the underlying types (for example to
11113 -- avoid an unnecessary dependency of a spec on the spec
11115 type size_t is mod 2 ** Standard'Address_Size;
11116 NULL_Stream : constant FILEs;
11117 -- Value returned (NULL in C) to indicate an
11118 -- fdopen/fopen/tmpfile error
11119 ----------------------------------
11120 -- Constants Defined in stdio.h --
11121 ----------------------------------
11122 EOF : constant int;
11123 -- Used by a number of routines to indicate error or
11125 IOFBF : constant int;
11126 IOLBF : constant int;
11127 IONBF : constant int;
11128 -- Used to indicate buffering mode for setvbuf call
11129 SEEK_CUR : constant int;
11130 SEEK_END : constant int;
11131 SEEK_SET : constant int;
11132 -- Used to indicate origin for fseek call
11133 function stdin return FILEs;
11134 function stdout return FILEs;
11135 function stderr return FILEs;
11136 -- Streams associated with standard files
11137 --------------------------
11138 -- Standard C functions --
11139 --------------------------
11140 -- The functions selected below are ones that are
11141 -- available in DOS, OS/2, UNIX and Xenix (but not
11142 -- necessarily in ANSI C). These are very thin interfaces
11143 -- which copy exactly the C headers. For more
11144 -- documentation on these functions, see the Microsoft C
11145 -- "Run-Time Library Reference" (Microsoft Press, 1990,
11146 -- ISBN 1-55615-225-6), which includes useful information
11147 -- on system compatibility.
11148 procedure clearerr (stream : FILEs);
11149 function fclose (stream : FILEs) return int;
11150 function fdopen (handle : int; mode : chars) return FILEs;
11151 function feof (stream : FILEs) return int;
11152 function ferror (stream : FILEs) return int;
11153 function fflush (stream : FILEs) return int;
11154 function fgetc (stream : FILEs) return int;
11155 function fgets (strng : chars; n : int; stream : FILEs)
11157 function fileno (stream : FILEs) return int;
11158 function fopen (filename : chars; Mode : chars)
11160 -- Note: to maintain target independence, use
11161 -- text_translation_required, a boolean variable defined in
11162 -- a-sysdep.c to deal with the target dependent text
11163 -- translation requirement. If this variable is set,
11164 -- then b/t should be appended to the standard mode
11165 -- argument to set the text translation mode off or on
11167 function fputc (C : int; stream : FILEs) return int;
11168 function fputs (Strng : chars; Stream : FILEs) return int;
11185 function ftell (stream : FILEs) return long;
11192 function isatty (handle : int) return int;
11193 procedure mktemp (template : chars);
11194 -- The return value (which is just a pointer to template)
11196 procedure rewind (stream : FILEs);
11197 function rmtmp return int;
11205 function tmpfile return FILEs;
11206 function ungetc (c : int; stream : FILEs) return int;
11207 function unlink (filename : chars) return int;
11208 ---------------------
11209 -- Extra functions --
11210 ---------------------
11211 -- These functions supply slightly thicker bindings than
11212 -- those above. They are derived from functions in the
11213 -- C Run-Time Library, but may do a bit more work than
11214 -- just directly calling one of the Library functions.
11215 function is_regular_file (handle : int) return int;
11216 -- Tests if given handle is for a regular file (result 1)
11217 -- or for a non-regular file (pipe or device, result 0).
11218 ---------------------------------
11219 -- Control of Text/Binary Mode --
11220 ---------------------------------
11221 -- If text_translation_required is true, then the following
11222 -- functions may be used to dynamically switch a file from
11223 -- binary to text mode or vice versa. These functions have
11224 -- no effect if text_translation_required is false (i.e. in
11225 -- normal UNIX mode). Use fileno to get a stream handle.
11226 procedure set_binary_mode (handle : int);
11227 procedure set_text_mode (handle : int);
11228 ----------------------------
11229 -- Full Path Name support --
11230 ----------------------------
11231 procedure full_name (nam : chars; buffer : chars);
11232 -- Given a NUL terminated string representing a file
11233 -- name, returns in buffer a NUL terminated string
11234 -- representing the full path name for the file name.
11235 -- On systems where it is relevant the drive is also
11236 -- part of the full path name. It is the responsibility
11237 -- of the caller to pass an actual parameter for buffer
11238 -- that is big enough for any full path name. Use
11239 -- max_path_len given below as the size of buffer.
11240 max_path_len : integer;
11241 -- Maximum length of an allowable full path name on the
11242 -- system, including a terminating NUL character.
11243 end Interfaces.C_Streams;
11246 @node Interfacing to C Streams
11247 @section Interfacing to C Streams
11250 The packages in this section permit interfacing Ada files to C Stream
11253 @smallexample @c ada
11254 with Interfaces.C_Streams;
11255 package Ada.Sequential_IO.C_Streams is
11256 function C_Stream (F : File_Type)
11257 return Interfaces.C_Streams.FILEs;
11259 (File : in out File_Type;
11260 Mode : in File_Mode;
11261 C_Stream : in Interfaces.C_Streams.FILEs;
11262 Form : in String := "");
11263 end Ada.Sequential_IO.C_Streams;
11265 with Interfaces.C_Streams;
11266 package Ada.Direct_IO.C_Streams is
11267 function C_Stream (F : File_Type)
11268 return Interfaces.C_Streams.FILEs;
11270 (File : in out File_Type;
11271 Mode : in File_Mode;
11272 C_Stream : in Interfaces.C_Streams.FILEs;
11273 Form : in String := "");
11274 end Ada.Direct_IO.C_Streams;
11276 with Interfaces.C_Streams;
11277 package Ada.Text_IO.C_Streams is
11278 function C_Stream (F : File_Type)
11279 return Interfaces.C_Streams.FILEs;
11281 (File : in out File_Type;
11282 Mode : in File_Mode;
11283 C_Stream : in Interfaces.C_Streams.FILEs;
11284 Form : in String := "");
11285 end Ada.Text_IO.C_Streams;
11287 with Interfaces.C_Streams;
11288 package Ada.Wide_Text_IO.C_Streams is
11289 function C_Stream (F : File_Type)
11290 return Interfaces.C_Streams.FILEs;
11292 (File : in out File_Type;
11293 Mode : in File_Mode;
11294 C_Stream : in Interfaces.C_Streams.FILEs;
11295 Form : in String := "");
11296 end Ada.Wide_Text_IO.C_Streams;
11298 with Interfaces.C_Streams;
11299 package Ada.Stream_IO.C_Streams is
11300 function C_Stream (F : File_Type)
11301 return Interfaces.C_Streams.FILEs;
11303 (File : in out File_Type;
11304 Mode : in File_Mode;
11305 C_Stream : in Interfaces.C_Streams.FILEs;
11306 Form : in String := "");
11307 end Ada.Stream_IO.C_Streams;
11311 In each of these five packages, the @code{C_Stream} function obtains the
11312 @code{FILE} pointer from a currently opened Ada file. It is then
11313 possible to use the @code{Interfaces.C_Streams} package to operate on
11314 this stream, or the stream can be passed to a C program which can
11315 operate on it directly. Of course the program is responsible for
11316 ensuring that only appropriate sequences of operations are executed.
11318 One particular use of relevance to an Ada program is that the
11319 @code{setvbuf} function can be used to control the buffering of the
11320 stream used by an Ada file. In the absence of such a call the standard
11321 default buffering is used.
11323 The @code{Open} procedures in these packages open a file giving an
11324 existing C Stream instead of a file name. Typically this stream is
11325 imported from a C program, allowing an Ada file to operate on an
11328 @node The GNAT Library
11329 @chapter The GNAT Library
11332 The GNAT library contains a number of general and special purpose packages.
11333 It represents functionality that the GNAT developers have found useful, and
11334 which is made available to GNAT users. The packages described here are fully
11335 supported, and upwards compatibility will be maintained in future releases,
11336 so you can use these facilities with the confidence that the same functionality
11337 will be available in future releases.
11339 The chapter here simply gives a brief summary of the facilities available.
11340 The full documentation is found in the spec file for the package. The full
11341 sources of these library packages, including both spec and body, are provided
11342 with all GNAT releases. For example, to find out the full specifications of
11343 the SPITBOL pattern matching capability, including a full tutorial and
11344 extensive examples, look in the @file{g-spipat.ads} file in the library.
11346 For each entry here, the package name (as it would appear in a @code{with}
11347 clause) is given, followed by the name of the corresponding spec file in
11348 parentheses. The packages are children in four hierarchies, @code{Ada},
11349 @code{Interfaces}, @code{System}, and @code{GNAT}, the latter being a
11350 GNAT-specific hierarchy.
11352 Note that an application program should only use packages in one of these
11353 four hierarchies if the package is defined in the Ada Reference Manual,
11354 or is listed in this section of the GNAT Programmers Reference Manual.
11355 All other units should be considered internal implementation units and
11356 should not be directly @code{with}'ed by application code. The use of
11357 a @code{with} statement that references one of these internal implementation
11358 units makes an application potentially dependent on changes in versions
11359 of GNAT, and will generate a warning message.
11362 * Ada.Characters.Latin_9 (a-chlat9.ads)::
11363 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
11364 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
11365 * Ada.Command_Line.Remove (a-colire.ads)::
11366 * Ada.Command_Line.Environment (a-colien.ads)::
11367 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
11368 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
11369 * Ada.Exceptions.Traceback (a-exctra.ads)::
11370 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
11371 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
11372 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
11373 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
11374 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
11375 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
11376 * GNAT.Array_Split (g-arrspl.ads)::
11377 * GNAT.AWK (g-awk.ads)::
11378 * GNAT.Bounded_Buffers (g-boubuf.ads)::
11379 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
11380 * GNAT.Bubble_Sort (g-bubsor.ads)::
11381 * GNAT.Bubble_Sort_A (g-busora.ads)::
11382 * GNAT.Bubble_Sort_G (g-busorg.ads)::
11383 * GNAT.Calendar (g-calend.ads)::
11384 * GNAT.Calendar.Time_IO (g-catiio.ads)::
11385 * GNAT.CRC32 (g-crc32.ads)::
11386 * GNAT.Case_Util (g-casuti.ads)::
11387 * GNAT.CGI (g-cgi.ads)::
11388 * GNAT.CGI.Cookie (g-cgicoo.ads)::
11389 * GNAT.CGI.Debug (g-cgideb.ads)::
11390 * GNAT.Command_Line (g-comlin.ads)::
11391 * GNAT.Compiler_Version (g-comver.ads)::
11392 * GNAT.Ctrl_C (g-ctrl_c.ads)::
11393 * GNAT.Current_Exception (g-curexc.ads)::
11394 * GNAT.Debug_Pools (g-debpoo.ads)::
11395 * GNAT.Debug_Utilities (g-debuti.ads)::
11396 * GNAT.Directory_Operations (g-dirope.ads)::
11397 * GNAT.Dynamic_HTables (g-dynhta.ads)::
11398 * GNAT.Dynamic_Tables (g-dyntab.ads)::
11399 * GNAT.Exception_Actions (g-excact.ads)::
11400 * GNAT.Exception_Traces (g-exctra.ads)::
11401 * GNAT.Exceptions (g-except.ads)::
11402 * GNAT.Expect (g-expect.ads)::
11403 * GNAT.Float_Control (g-flocon.ads)::
11404 * GNAT.Heap_Sort (g-heasor.ads)::
11405 * GNAT.Heap_Sort_A (g-hesora.ads)::
11406 * GNAT.Heap_Sort_G (g-hesorg.ads)::
11407 * GNAT.HTable (g-htable.ads)::
11408 * GNAT.IO (g-io.ads)::
11409 * GNAT.IO_Aux (g-io_aux.ads)::
11410 * GNAT.Lock_Files (g-locfil.ads)::
11411 * GNAT.MD5 (g-md5.ads)::
11412 * GNAT.Memory_Dump (g-memdum.ads)::
11413 * GNAT.Most_Recent_Exception (g-moreex.ads)::
11414 * GNAT.OS_Lib (g-os_lib.ads)::
11415 * GNAT.Perfect_Hash.Generators (g-pehage.ads)::
11416 * GNAT.Regexp (g-regexp.ads)::
11417 * GNAT.Registry (g-regist.ads)::
11418 * GNAT.Regpat (g-regpat.ads)::
11419 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
11420 * GNAT.Semaphores (g-semaph.ads)::
11421 * GNAT.Signals (g-signal.ads)::
11422 * GNAT.Sockets (g-socket.ads)::
11423 * GNAT.Source_Info (g-souinf.ads)::
11424 * GNAT.Spell_Checker (g-speche.ads)::
11425 * GNAT.Spitbol.Patterns (g-spipat.ads)::
11426 * GNAT.Spitbol (g-spitbo.ads)::
11427 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
11428 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
11429 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
11430 * GNAT.Strings (g-string.ads)::
11431 * GNAT.String_Split (g-strspl.ads)::
11432 * GNAT.Table (g-table.ads)::
11433 * GNAT.Task_Lock (g-tasloc.ads)::
11434 * GNAT.Threads (g-thread.ads)::
11435 * GNAT.Traceback (g-traceb.ads)::
11436 * GNAT.Traceback.Symbolic (g-trasym.ads)::
11437 * GNAT.Wide_String_Split (g-wistsp.ads)::
11438 * Interfaces.C.Extensions (i-cexten.ads)::
11439 * Interfaces.C.Streams (i-cstrea.ads)::
11440 * Interfaces.CPP (i-cpp.ads)::
11441 * Interfaces.Os2lib (i-os2lib.ads)::
11442 * Interfaces.Os2lib.Errors (i-os2err.ads)::
11443 * Interfaces.Os2lib.Synchronization (i-os2syn.ads)::
11444 * Interfaces.Os2lib.Threads (i-os2thr.ads)::
11445 * Interfaces.Packed_Decimal (i-pacdec.ads)::
11446 * Interfaces.VxWorks (i-vxwork.ads)::
11447 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
11448 * System.Address_Image (s-addima.ads)::
11449 * System.Assertions (s-assert.ads)::
11450 * System.Memory (s-memory.ads)::
11451 * System.Partition_Interface (s-parint.ads)::
11452 * System.Restrictions (s-restri.ads)::
11453 * System.Rident (s-rident.ads)::
11454 * System.Task_Info (s-tasinf.ads)::
11455 * System.Wch_Cnv (s-wchcnv.ads)::
11456 * System.Wch_Con (s-wchcon.ads)::
11459 @node Ada.Characters.Latin_9 (a-chlat9.ads)
11460 @section @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
11461 @cindex @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
11462 @cindex Latin_9 constants for Character
11465 This child of @code{Ada.Characters}
11466 provides a set of definitions corresponding to those in the
11467 RM-defined package @code{Ada.Characters.Latin_1} but with the
11468 few modifications required for @code{Latin-9}
11469 The provision of such a package
11470 is specifically authorized by the Ada Reference Manual
11473 @node Ada.Characters.Wide_Latin_1 (a-cwila1.ads)
11474 @section @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
11475 @cindex @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
11476 @cindex Latin_1 constants for Wide_Character
11479 This child of @code{Ada.Characters}
11480 provides a set of definitions corresponding to those in the
11481 RM-defined package @code{Ada.Characters.Latin_1} but with the
11482 types of the constants being @code{Wide_Character}
11483 instead of @code{Character}. The provision of such a package
11484 is specifically authorized by the Ada Reference Manual
11487 @node Ada.Characters.Wide_Latin_9 (a-cwila9.ads)
11488 @section @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
11489 @cindex @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
11490 @cindex Latin_9 constants for Wide_Character
11493 This child of @code{Ada.Characters}
11494 provides a set of definitions corresponding to those in the
11495 GNAT defined package @code{Ada.Characters.Latin_9} but with the
11496 types of the constants being @code{Wide_Character}
11497 instead of @code{Character}. The provision of such a package
11498 is specifically authorized by the Ada Reference Manual
11501 @node Ada.Command_Line.Remove (a-colire.ads)
11502 @section @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
11503 @cindex @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
11504 @cindex Removing command line arguments
11505 @cindex Command line, argument removal
11508 This child of @code{Ada.Command_Line}
11509 provides a mechanism for logically removing
11510 arguments from the argument list. Once removed, an argument is not visible
11511 to further calls on the subprograms in @code{Ada.Command_Line} will not
11512 see the removed argument.
11514 @node Ada.Command_Line.Environment (a-colien.ads)
11515 @section @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
11516 @cindex @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
11517 @cindex Environment entries
11520 This child of @code{Ada.Command_Line}
11521 provides a mechanism for obtaining environment values on systems
11522 where this concept makes sense.
11524 @node Ada.Direct_IO.C_Streams (a-diocst.ads)
11525 @section @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
11526 @cindex @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
11527 @cindex C Streams, Interfacing with Direct_IO
11530 This package provides subprograms that allow interfacing between
11531 C streams and @code{Direct_IO}. The stream identifier can be
11532 extracted from a file opened on the Ada side, and an Ada file
11533 can be constructed from a stream opened on the C side.
11535 @node Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)
11536 @section @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
11537 @cindex @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
11538 @cindex Null_Occurrence, testing for
11541 This child subprogram provides a way of testing for the null
11542 exception occurrence (@code{Null_Occurrence}) without raising
11545 @node Ada.Exceptions.Traceback (a-exctra.ads)
11546 @section @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
11547 @cindex @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
11548 @cindex Traceback for Exception Occurrence
11551 This child package provides the subprogram (@code{Tracebacks}) to
11552 give a traceback array of addresses based on an exception
11555 @node Ada.Sequential_IO.C_Streams (a-siocst.ads)
11556 @section @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
11557 @cindex @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
11558 @cindex C Streams, Interfacing with Sequential_IO
11561 This package provides subprograms that allow interfacing between
11562 C streams and @code{Sequential_IO}. The stream identifier can be
11563 extracted from a file opened on the Ada side, and an Ada file
11564 can be constructed from a stream opened on the C side.
11566 @node Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)
11567 @section @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
11568 @cindex @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
11569 @cindex C Streams, Interfacing with Stream_IO
11572 This package provides subprograms that allow interfacing between
11573 C streams and @code{Stream_IO}. The stream identifier can be
11574 extracted from a file opened on the Ada side, and an Ada file
11575 can be constructed from a stream opened on the C side.
11577 @node Ada.Strings.Unbounded.Text_IO (a-suteio.ads)
11578 @section @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
11579 @cindex @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
11580 @cindex @code{Unbounded_String}, IO support
11581 @cindex @code{Text_IO}, extensions for unbounded strings
11584 This package provides subprograms for Text_IO for unbounded
11585 strings, avoiding the necessity for an intermediate operation
11586 with ordinary strings.
11588 @node Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)
11589 @section @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
11590 @cindex @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
11591 @cindex @code{Unbounded_Wide_String}, IO support
11592 @cindex @code{Text_IO}, extensions for unbounded wide strings
11595 This package provides subprograms for Text_IO for unbounded
11596 wide strings, avoiding the necessity for an intermediate operation
11597 with ordinary wide strings.
11599 @node Ada.Text_IO.C_Streams (a-tiocst.ads)
11600 @section @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
11601 @cindex @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
11602 @cindex C Streams, Interfacing with @code{Text_IO}
11605 This package provides subprograms that allow interfacing between
11606 C streams and @code{Text_IO}. The stream identifier can be
11607 extracted from a file opened on the Ada side, and an Ada file
11608 can be constructed from a stream opened on the C side.
11610 @node Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)
11611 @section @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
11612 @cindex @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
11613 @cindex C Streams, Interfacing with @code{Wide_Text_IO}
11616 This package provides subprograms that allow interfacing between
11617 C streams and @code{Wide_Text_IO}. The stream identifier can be
11618 extracted from a file opened on the Ada side, and an Ada file
11619 can be constructed from a stream opened on the C side.
11621 @node GNAT.Array_Split (g-arrspl.ads)
11622 @section @code{GNAT.Array_Split} (@file{g-arrspl.ads})
11623 @cindex @code{GNAT.Array_Split} (@file{g-arrspl.ads})
11624 @cindex Array splitter
11627 Useful array-manipulation routines: given a set of separators, split
11628 an array wherever the separators appear, and provide direct access
11629 to the resulting slices.
11631 @node GNAT.AWK (g-awk.ads)
11632 @section @code{GNAT.AWK} (@file{g-awk.ads})
11633 @cindex @code{GNAT.AWK} (@file{g-awk.ads})
11638 Provides AWK-like parsing functions, with an easy interface for parsing one
11639 or more files containing formatted data. The file is viewed as a database
11640 where each record is a line and a field is a data element in this line.
11642 @node GNAT.Bounded_Buffers (g-boubuf.ads)
11643 @section @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
11644 @cindex @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
11646 @cindex Bounded Buffers
11649 Provides a concurrent generic bounded buffer abstraction. Instances are
11650 useful directly or as parts of the implementations of other abstractions,
11653 @node GNAT.Bounded_Mailboxes (g-boumai.ads)
11654 @section @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
11655 @cindex @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
11660 Provides a thread-safe asynchronous intertask mailbox communication facility.
11662 @node GNAT.Bubble_Sort (g-bubsor.ads)
11663 @section @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
11664 @cindex @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
11666 @cindex Bubble sort
11669 Provides a general implementation of bubble sort usable for sorting arbitrary
11670 data items. Exchange and comparison procedures are provided by passing
11671 access-to-procedure values.
11673 @node GNAT.Bubble_Sort_A (g-busora.ads)
11674 @section @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
11675 @cindex @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
11677 @cindex Bubble sort
11680 Provides a general implementation of bubble sort usable for sorting arbitrary
11681 data items. Move and comparison procedures are provided by passing
11682 access-to-procedure values. This is an older version, retained for
11683 compatibility. Usually @code{GNAT.Bubble_Sort} will be preferable.
11685 @node GNAT.Bubble_Sort_G (g-busorg.ads)
11686 @section @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
11687 @cindex @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
11689 @cindex Bubble sort
11692 Similar to @code{Bubble_Sort_A} except that the move and sorting procedures
11693 are provided as generic parameters, this improves efficiency, especially
11694 if the procedures can be inlined, at the expense of duplicating code for
11695 multiple instantiations.
11697 @node GNAT.Calendar (g-calend.ads)
11698 @section @code{GNAT.Calendar} (@file{g-calend.ads})
11699 @cindex @code{GNAT.Calendar} (@file{g-calend.ads})
11700 @cindex @code{Calendar}
11703 Extends the facilities provided by @code{Ada.Calendar} to include handling
11704 of days of the week, an extended @code{Split} and @code{Time_Of} capability.
11705 Also provides conversion of @code{Ada.Calendar.Time} values to and from the
11706 C @code{timeval} format.
11708 @node GNAT.Calendar.Time_IO (g-catiio.ads)
11709 @section @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
11710 @cindex @code{Calendar}
11712 @cindex @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
11714 @node GNAT.CRC32 (g-crc32.ads)
11715 @section @code{GNAT.CRC32} (@file{g-crc32.ads})
11716 @cindex @code{GNAT.CRC32} (@file{g-crc32.ads})
11718 @cindex Cyclic Redundancy Check
11721 This package implements the CRC-32 algorithm. For a full description
11722 of this algorithm see
11723 ``Computation of Cyclic Redundancy Checks via Table Look-Up'',
11724 @cite{Communications of the ACM}, Vol.@: 31 No.@: 8, pp.@: 1008-1013,
11725 Aug.@: 1988. Sarwate, D.V@.
11728 Provides an extended capability for formatted output of time values with
11729 full user control over the format. Modeled on the GNU Date specification.
11731 @node GNAT.Case_Util (g-casuti.ads)
11732 @section @code{GNAT.Case_Util} (@file{g-casuti.ads})
11733 @cindex @code{GNAT.Case_Util} (@file{g-casuti.ads})
11734 @cindex Casing utilities
11735 @cindex Character handling (@code{GNAT.Case_Util})
11738 A set of simple routines for handling upper and lower casing of strings
11739 without the overhead of the full casing tables
11740 in @code{Ada.Characters.Handling}.
11742 @node GNAT.CGI (g-cgi.ads)
11743 @section @code{GNAT.CGI} (@file{g-cgi.ads})
11744 @cindex @code{GNAT.CGI} (@file{g-cgi.ads})
11745 @cindex CGI (Common Gateway Interface)
11748 This is a package for interfacing a GNAT program with a Web server via the
11749 Common Gateway Interface (CGI)@. Basically this package parses the CGI
11750 parameters, which are a set of key/value pairs sent by the Web server. It
11751 builds a table whose index is the key and provides some services to deal
11754 @node GNAT.CGI.Cookie (g-cgicoo.ads)
11755 @section @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
11756 @cindex @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
11757 @cindex CGI (Common Gateway Interface) cookie support
11758 @cindex Cookie support in CGI
11761 This is a package to interface a GNAT program with a Web server via the
11762 Common Gateway Interface (CGI). It exports services to deal with Web
11763 cookies (piece of information kept in the Web client software).
11765 @node GNAT.CGI.Debug (g-cgideb.ads)
11766 @section @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
11767 @cindex @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
11768 @cindex CGI (Common Gateway Interface) debugging
11771 This is a package to help debugging CGI (Common Gateway Interface)
11772 programs written in Ada.
11774 @node GNAT.Command_Line (g-comlin.ads)
11775 @section @code{GNAT.Command_Line} (@file{g-comlin.ads})
11776 @cindex @code{GNAT.Command_Line} (@file{g-comlin.ads})
11777 @cindex Command line
11780 Provides a high level interface to @code{Ada.Command_Line} facilities,
11781 including the ability to scan for named switches with optional parameters
11782 and expand file names using wild card notations.
11784 @node GNAT.Compiler_Version (g-comver.ads)
11785 @section @code{GNAT.Compiler_Version} (@file{g-comver.ads})
11786 @cindex @code{GNAT.Compiler_Version} (@file{g-comver.ads})
11787 @cindex Compiler Version
11788 @cindex Version, of compiler
11791 Provides a routine for obtaining the version of the compiler used to
11792 compile the program. More accurately this is the version of the binder
11793 used to bind the program (this will normally be the same as the version
11794 of the compiler if a consistent tool set is used to compile all units
11797 @node GNAT.Ctrl_C (g-ctrl_c.ads)
11798 @section @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
11799 @cindex @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
11803 Provides a simple interface to handle Ctrl-C keyboard events.
11805 @node GNAT.Current_Exception (g-curexc.ads)
11806 @section @code{GNAT.Current_Exception} (@file{g-curexc.ads})
11807 @cindex @code{GNAT.Current_Exception} (@file{g-curexc.ads})
11808 @cindex Current exception
11809 @cindex Exception retrieval
11812 Provides access to information on the current exception that has been raised
11813 without the need for using the Ada-95 exception choice parameter specification
11814 syntax. This is particularly useful in simulating typical facilities for
11815 obtaining information about exceptions provided by Ada 83 compilers.
11817 @node GNAT.Debug_Pools (g-debpoo.ads)
11818 @section @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
11819 @cindex @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
11821 @cindex Debug pools
11822 @cindex Memory corruption debugging
11825 Provide a debugging storage pools that helps tracking memory corruption
11826 problems. See section ``Finding memory problems with GNAT Debug Pool'' in
11827 the @cite{GNAT User's Guide}.
11829 @node GNAT.Debug_Utilities (g-debuti.ads)
11830 @section @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
11831 @cindex @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
11835 Provides a few useful utilities for debugging purposes, including conversion
11836 to and from string images of address values. Supports both C and Ada formats
11837 for hexadecimal literals.
11839 @node GNAT.Directory_Operations (g-dirope.ads)
11840 @section @code{GNAT.Directory_Operations} (g-dirope.ads)
11841 @cindex @code{GNAT.Directory_Operations} (g-dirope.ads)
11842 @cindex Directory operations
11845 Provides a set of routines for manipulating directories, including changing
11846 the current directory, making new directories, and scanning the files in a
11849 @node GNAT.Dynamic_HTables (g-dynhta.ads)
11850 @section @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
11851 @cindex @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
11852 @cindex Hash tables
11855 A generic implementation of hash tables that can be used to hash arbitrary
11856 data. Provided in two forms, a simple form with built in hash functions,
11857 and a more complex form in which the hash function is supplied.
11860 This package provides a facility similar to that of @code{GNAT.HTable},
11861 except that this package declares a type that can be used to define
11862 dynamic instances of the hash table, while an instantiation of
11863 @code{GNAT.HTable} creates a single instance of the hash table.
11865 @node GNAT.Dynamic_Tables (g-dyntab.ads)
11866 @section @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
11867 @cindex @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
11868 @cindex Table implementation
11869 @cindex Arrays, extendable
11872 A generic package providing a single dimension array abstraction where the
11873 length of the array can be dynamically modified.
11876 This package provides a facility similar to that of @code{GNAT.Table},
11877 except that this package declares a type that can be used to define
11878 dynamic instances of the table, while an instantiation of
11879 @code{GNAT.Table} creates a single instance of the table type.
11881 @node GNAT.Exception_Actions (g-excact.ads)
11882 @section @code{GNAT.Exception_Actions} (@file{g-excact.ads})
11883 @cindex @code{GNAT.Exception_Actions} (@file{g-excact.ads})
11884 @cindex Exception actions
11887 Provides callbacks when an exception is raised. Callbacks can be registered
11888 for specific exceptions, or when any exception is raised. This
11889 can be used for instance to force a core dump to ease debugging.
11891 @node GNAT.Exception_Traces (g-exctra.ads)
11892 @section @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
11893 @cindex @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
11894 @cindex Exception traces
11898 Provides an interface allowing to control automatic output upon exception
11901 @node GNAT.Exceptions (g-except.ads)
11902 @section @code{GNAT.Exceptions} (@file{g-expect.ads})
11903 @cindex @code{GNAT.Exceptions} (@file{g-expect.ads})
11904 @cindex Exceptions, Pure
11905 @cindex Pure packages, exceptions
11908 Normally it is not possible to raise an exception with
11909 a message from a subprogram in a pure package, since the
11910 necessary types and subprograms are in @code{Ada.Exceptions}
11911 which is not a pure unit. @code{GNAT.Exceptions} provides a
11912 facility for getting around this limitation for a few
11913 predefined exceptions, and for example allow raising
11914 @code{Constraint_Error} with a message from a pure subprogram.
11916 @node GNAT.Expect (g-expect.ads)
11917 @section @code{GNAT.Expect} (@file{g-expect.ads})
11918 @cindex @code{GNAT.Expect} (@file{g-expect.ads})
11921 Provides a set of subprograms similar to what is available
11922 with the standard Tcl Expect tool.
11923 It allows you to easily spawn and communicate with an external process.
11924 You can send commands or inputs to the process, and compare the output
11925 with some expected regular expression. Currently @code{GNAT.Expect}
11926 is implemented on all native GNAT ports except for OpenVMS@.
11927 It is not implemented for cross ports, and in particular is not
11928 implemented for VxWorks or LynxOS@.
11930 @node GNAT.Float_Control (g-flocon.ads)
11931 @section @code{GNAT.Float_Control} (@file{g-flocon.ads})
11932 @cindex @code{GNAT.Float_Control} (@file{g-flocon.ads})
11933 @cindex Floating-Point Processor
11936 Provides an interface for resetting the floating-point processor into the
11937 mode required for correct semantic operation in Ada. Some third party
11938 library calls may cause this mode to be modified, and the Reset procedure
11939 in this package can be used to reestablish the required mode.
11941 @node GNAT.Heap_Sort (g-heasor.ads)
11942 @section @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
11943 @cindex @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
11947 Provides a general implementation of heap sort usable for sorting arbitrary
11948 data items. Exchange and comparison procedures are provided by passing
11949 access-to-procedure values. The algorithm used is a modified heap sort
11950 that performs approximately N*log(N) comparisons in the worst case.
11952 @node GNAT.Heap_Sort_A (g-hesora.ads)
11953 @section @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
11954 @cindex @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
11958 Provides a general implementation of heap sort usable for sorting arbitrary
11959 data items. Move and comparison procedures are provided by passing
11960 access-to-procedure values. The algorithm used is a modified heap sort
11961 that performs approximately N*log(N) comparisons in the worst case.
11962 This differs from @code{GNAT.Heap_Sort} in having a less convenient
11963 interface, but may be slightly more efficient.
11965 @node GNAT.Heap_Sort_G (g-hesorg.ads)
11966 @section @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
11967 @cindex @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
11971 Similar to @code{Heap_Sort_A} except that the move and sorting procedures
11972 are provided as generic parameters, this improves efficiency, especially
11973 if the procedures can be inlined, at the expense of duplicating code for
11974 multiple instantiations.
11976 @node GNAT.HTable (g-htable.ads)
11977 @section @code{GNAT.HTable} (@file{g-htable.ads})
11978 @cindex @code{GNAT.HTable} (@file{g-htable.ads})
11979 @cindex Hash tables
11982 A generic implementation of hash tables that can be used to hash arbitrary
11983 data. Provides two approaches, one a simple static approach, and the other
11984 allowing arbitrary dynamic hash tables.
11986 @node GNAT.IO (g-io.ads)
11987 @section @code{GNAT.IO} (@file{g-io.ads})
11988 @cindex @code{GNAT.IO} (@file{g-io.ads})
11990 @cindex Input/Output facilities
11993 A simple preelaborable input-output package that provides a subset of
11994 simple Text_IO functions for reading characters and strings from
11995 Standard_Input, and writing characters, strings and integers to either
11996 Standard_Output or Standard_Error.
11998 @node GNAT.IO_Aux (g-io_aux.ads)
11999 @section @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
12000 @cindex @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
12002 @cindex Input/Output facilities
12004 Provides some auxiliary functions for use with Text_IO, including a test
12005 for whether a file exists, and functions for reading a line of text.
12007 @node GNAT.Lock_Files (g-locfil.ads)
12008 @section @code{GNAT.Lock_Files} (@file{g-locfil.ads})
12009 @cindex @code{GNAT.Lock_Files} (@file{g-locfil.ads})
12010 @cindex File locking
12011 @cindex Locking using files
12014 Provides a general interface for using files as locks. Can be used for
12015 providing program level synchronization.
12017 @node GNAT.MD5 (g-md5.ads)
12018 @section @code{GNAT.MD5} (@file{g-md5.ads})
12019 @cindex @code{GNAT.MD5} (@file{g-md5.ads})
12020 @cindex Message Digest MD5
12023 Implements the MD5 Message-Digest Algorithm as described in RFC 1321.
12025 @node GNAT.Memory_Dump (g-memdum.ads)
12026 @section @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
12027 @cindex @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
12028 @cindex Dump Memory
12031 Provides a convenient routine for dumping raw memory to either the
12032 standard output or standard error files. Uses GNAT.IO for actual
12035 @node GNAT.Most_Recent_Exception (g-moreex.ads)
12036 @section @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
12037 @cindex @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
12038 @cindex Exception, obtaining most recent
12041 Provides access to the most recently raised exception. Can be used for
12042 various logging purposes, including duplicating functionality of some
12043 Ada 83 implementation dependent extensions.
12045 @node GNAT.OS_Lib (g-os_lib.ads)
12046 @section @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
12047 @cindex @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
12048 @cindex Operating System interface
12049 @cindex Spawn capability
12052 Provides a range of target independent operating system interface functions,
12053 including time/date management, file operations, subprocess management,
12054 including a portable spawn procedure, and access to environment variables
12055 and error return codes.
12057 @node GNAT.Perfect_Hash.Generators (g-pehage.ads)
12058 @section @code{GNAT.Perfect_Hash.Generators} (@file{g-pehage.ads})
12059 @cindex @code{GNAT.Perfect_Hash.Generators} (@file{g-pehage.ads})
12060 @cindex Hash functions
12063 Provides a generator of static minimal perfect hash functions. No
12064 collisions occur and each item can be retrieved from the table in one
12065 probe (perfect property). The hash table size corresponds to the exact
12066 size of the key set and no larger (minimal property). The key set has to
12067 be know in advance (static property). The hash functions are also order
12068 preservering. If w2 is inserted after w1 in the generator, their
12069 hashcode are in the same order. These hashing functions are very
12070 convenient for use with realtime applications.
12072 @node GNAT.Regexp (g-regexp.ads)
12073 @section @code{GNAT.Regexp} (@file{g-regexp.ads})
12074 @cindex @code{GNAT.Regexp} (@file{g-regexp.ads})
12075 @cindex Regular expressions
12076 @cindex Pattern matching
12079 A simple implementation of regular expressions, using a subset of regular
12080 expression syntax copied from familiar Unix style utilities. This is the
12081 simples of the three pattern matching packages provided, and is particularly
12082 suitable for ``file globbing'' applications.
12084 @node GNAT.Registry (g-regist.ads)
12085 @section @code{GNAT.Registry} (@file{g-regist.ads})
12086 @cindex @code{GNAT.Registry} (@file{g-regist.ads})
12087 @cindex Windows Registry
12090 This is a high level binding to the Windows registry. It is possible to
12091 do simple things like reading a key value, creating a new key. For full
12092 registry API, but at a lower level of abstraction, refer to the Win32.Winreg
12093 package provided with the Win32Ada binding
12095 @node GNAT.Regpat (g-regpat.ads)
12096 @section @code{GNAT.Regpat} (@file{g-regpat.ads})
12097 @cindex @code{GNAT.Regpat} (@file{g-regpat.ads})
12098 @cindex Regular expressions
12099 @cindex Pattern matching
12102 A complete implementation of Unix-style regular expression matching, copied
12103 from the original V7 style regular expression library written in C by
12104 Henry Spencer (and binary compatible with this C library).
12106 @node GNAT.Secondary_Stack_Info (g-sestin.ads)
12107 @section @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
12108 @cindex @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
12109 @cindex Secondary Stack Info
12112 Provide the capability to query the high water mark of the current task's
12115 @node GNAT.Semaphores (g-semaph.ads)
12116 @section @code{GNAT.Semaphores} (@file{g-semaph.ads})
12117 @cindex @code{GNAT.Semaphores} (@file{g-semaph.ads})
12121 Provides classic counting and binary semaphores using protected types.
12123 @node GNAT.Signals (g-signal.ads)
12124 @section @code{GNAT.Signals} (@file{g-signal.ads})
12125 @cindex @code{GNAT.Signals} (@file{g-signal.ads})
12129 Provides the ability to manipulate the blocked status of signals on supported
12132 @node GNAT.Sockets (g-socket.ads)
12133 @section @code{GNAT.Sockets} (@file{g-socket.ads})
12134 @cindex @code{GNAT.Sockets} (@file{g-socket.ads})
12138 A high level and portable interface to develop sockets based applications.
12139 This package is based on the sockets thin binding found in
12140 @code{GNAT.Sockets.Thin}. Currently @code{GNAT.Sockets} is implemented
12141 on all native GNAT ports except for OpenVMS@. It is not implemented
12142 for the LynxOS@ cross port.
12144 @node GNAT.Source_Info (g-souinf.ads)
12145 @section @code{GNAT.Source_Info} (@file{g-souinf.ads})
12146 @cindex @code{GNAT.Source_Info} (@file{g-souinf.ads})
12147 @cindex Source Information
12150 Provides subprograms that give access to source code information known at
12151 compile time, such as the current file name and line number.
12153 @node GNAT.Spell_Checker (g-speche.ads)
12154 @section @code{GNAT.Spell_Checker} (@file{g-speche.ads})
12155 @cindex @code{GNAT.Spell_Checker} (@file{g-speche.ads})
12156 @cindex Spell checking
12159 Provides a function for determining whether one string is a plausible
12160 near misspelling of another string.
12162 @node GNAT.Spitbol.Patterns (g-spipat.ads)
12163 @section @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
12164 @cindex @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
12165 @cindex SPITBOL pattern matching
12166 @cindex Pattern matching
12169 A complete implementation of SNOBOL4 style pattern matching. This is the
12170 most elaborate of the pattern matching packages provided. It fully duplicates
12171 the SNOBOL4 dynamic pattern construction and matching capabilities, using the
12172 efficient algorithm developed by Robert Dewar for the SPITBOL system.
12174 @node GNAT.Spitbol (g-spitbo.ads)
12175 @section @code{GNAT.Spitbol} (@file{g-spitbo.ads})
12176 @cindex @code{GNAT.Spitbol} (@file{g-spitbo.ads})
12177 @cindex SPITBOL interface
12180 The top level package of the collection of SPITBOL-style functionality, this
12181 package provides basic SNOBOL4 string manipulation functions, such as
12182 Pad, Reverse, Trim, Substr capability, as well as a generic table function
12183 useful for constructing arbitrary mappings from strings in the style of
12184 the SNOBOL4 TABLE function.
12186 @node GNAT.Spitbol.Table_Boolean (g-sptabo.ads)
12187 @section @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
12188 @cindex @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
12189 @cindex Sets of strings
12190 @cindex SPITBOL Tables
12193 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
12194 for type @code{Standard.Boolean}, giving an implementation of sets of
12197 @node GNAT.Spitbol.Table_Integer (g-sptain.ads)
12198 @section @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
12199 @cindex @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
12200 @cindex Integer maps
12202 @cindex SPITBOL Tables
12205 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
12206 for type @code{Standard.Integer}, giving an implementation of maps
12207 from string to integer values.
12209 @node GNAT.Spitbol.Table_VString (g-sptavs.ads)
12210 @section @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
12211 @cindex @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
12212 @cindex String maps
12214 @cindex SPITBOL Tables
12217 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table} for
12218 a variable length string type, giving an implementation of general
12219 maps from strings to strings.
12221 @node GNAT.Strings (g-string.ads)
12222 @section @code{GNAT.Strings} (@file{g-string.ads})
12223 @cindex @code{GNAT.Strings} (@file{g-string.ads})
12226 Common String access types and related subprograms. Basically it
12227 defines a string access and an array of string access types.
12229 @node GNAT.String_Split (g-strspl.ads)
12230 @section @code{GNAT.String_Split} (@file{g-strspl.ads})
12231 @cindex @code{GNAT.String_Split} (@file{g-strspl.ads})
12232 @cindex String splitter
12235 Useful string-manipulation routines: given a set of separators, split
12236 a string wherever the separators appear, and provide direct access
12237 to the resulting slices. This package is instantiated from
12238 @code{GNAT.Array_Split}.
12240 @node GNAT.Table (g-table.ads)
12241 @section @code{GNAT.Table} (@file{g-table.ads})
12242 @cindex @code{GNAT.Table} (@file{g-table.ads})
12243 @cindex Table implementation
12244 @cindex Arrays, extendable
12247 A generic package providing a single dimension array abstraction where the
12248 length of the array can be dynamically modified.
12251 This package provides a facility similar to that of @code{GNAT.Dynamic_Tables},
12252 except that this package declares a single instance of the table type,
12253 while an instantiation of @code{GNAT.Dynamic_Tables} creates a type that can be
12254 used to define dynamic instances of the table.
12256 @node GNAT.Task_Lock (g-tasloc.ads)
12257 @section @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
12258 @cindex @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
12259 @cindex Task synchronization
12260 @cindex Task locking
12264 A very simple facility for locking and unlocking sections of code using a
12265 single global task lock. Appropriate for use in situations where contention
12266 between tasks is very rarely expected.
12268 @node GNAT.Threads (g-thread.ads)
12269 @section @code{GNAT.Threads} (@file{g-thread.ads})
12270 @cindex @code{GNAT.Threads} (@file{g-thread.ads})
12271 @cindex Foreign threads
12272 @cindex Threads, foreign
12275 Provides facilities for creating and destroying threads with explicit calls.
12276 These threads are known to the GNAT run-time system. These subprograms are
12277 exported C-convention procedures intended to be called from foreign code.
12278 By using these primitives rather than directly calling operating systems
12279 routines, compatibility with the Ada tasking runt-time is provided.
12281 @node GNAT.Traceback (g-traceb.ads)
12282 @section @code{GNAT.Traceback} (@file{g-traceb.ads})
12283 @cindex @code{GNAT.Traceback} (@file{g-traceb.ads})
12284 @cindex Trace back facilities
12287 Provides a facility for obtaining non-symbolic traceback information, useful
12288 in various debugging situations.
12290 @node GNAT.Traceback.Symbolic (g-trasym.ads)
12291 @section @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
12292 @cindex @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
12293 @cindex Trace back facilities
12296 Provides symbolic traceback information that includes the subprogram
12297 name and line number information.
12299 @node GNAT.Wide_String_Split (g-wistsp.ads)
12300 @section @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
12301 @cindex @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
12302 @cindex Wide_String splitter
12305 Useful wide_string-manipulation routines: given a set of separators, split
12306 a wide_string wherever the separators appear, and provide direct access
12307 to the resulting slices. This package is instantiated from
12308 @code{GNAT.Array_Split}.
12310 @node Interfaces.C.Extensions (i-cexten.ads)
12311 @section @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
12312 @cindex @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
12315 This package contains additional C-related definitions, intended
12316 for use with either manually or automatically generated bindings
12319 @node Interfaces.C.Streams (i-cstrea.ads)
12320 @section @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
12321 @cindex @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
12322 @cindex C streams, interfacing
12325 This package is a binding for the most commonly used operations
12328 @node Interfaces.CPP (i-cpp.ads)
12329 @section @code{Interfaces.CPP} (@file{i-cpp.ads})
12330 @cindex @code{Interfaces.CPP} (@file{i-cpp.ads})
12331 @cindex C++ interfacing
12332 @cindex Interfacing, to C++
12335 This package provides facilities for use in interfacing to C++. It
12336 is primarily intended to be used in connection with automated tools
12337 for the generation of C++ interfaces.
12339 @node Interfaces.Os2lib (i-os2lib.ads)
12340 @section @code{Interfaces.Os2lib} (@file{i-os2lib.ads})
12341 @cindex @code{Interfaces.Os2lib} (@file{i-os2lib.ads})
12342 @cindex Interfacing, to OS/2
12343 @cindex OS/2 interfacing
12346 This package provides interface definitions to the OS/2 library.
12347 It is a thin binding which is a direct translation of the
12348 various @file{<bse@.h>} files.
12350 @node Interfaces.Os2lib.Errors (i-os2err.ads)
12351 @section @code{Interfaces.Os2lib.Errors} (@file{i-os2err.ads})
12352 @cindex @code{Interfaces.Os2lib.Errors} (@file{i-os2err.ads})
12353 @cindex OS/2 Error codes
12354 @cindex Interfacing, to OS/2
12355 @cindex OS/2 interfacing
12358 This package provides definitions of the OS/2 error codes.
12360 @node Interfaces.Os2lib.Synchronization (i-os2syn.ads)
12361 @section @code{Interfaces.Os2lib.Synchronization} (@file{i-os2syn.ads})
12362 @cindex @code{Interfaces.Os2lib.Synchronization} (@file{i-os2syn.ads})
12363 @cindex Interfacing, to OS/2
12364 @cindex Synchronization, OS/2
12365 @cindex OS/2 synchronization primitives
12368 This is a child package that provides definitions for interfacing
12369 to the @code{OS/2} synchronization primitives.
12371 @node Interfaces.Os2lib.Threads (i-os2thr.ads)
12372 @section @code{Interfaces.Os2lib.Threads} (@file{i-os2thr.ads})
12373 @cindex @code{Interfaces.Os2lib.Threads} (@file{i-os2thr.ads})
12374 @cindex Interfacing, to OS/2
12375 @cindex Thread control, OS/2
12376 @cindex OS/2 thread interfacing
12379 This is a child package that provides definitions for interfacing
12380 to the @code{OS/2} thread primitives.
12382 @node Interfaces.Packed_Decimal (i-pacdec.ads)
12383 @section @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
12384 @cindex @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
12385 @cindex IBM Packed Format
12386 @cindex Packed Decimal
12389 This package provides a set of routines for conversions to and
12390 from a packed decimal format compatible with that used on IBM
12393 @node Interfaces.VxWorks (i-vxwork.ads)
12394 @section @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
12395 @cindex @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
12396 @cindex Interfacing to VxWorks
12397 @cindex VxWorks, interfacing
12400 This package provides a limited binding to the VxWorks API.
12401 In particular, it interfaces with the
12402 VxWorks hardware interrupt facilities.
12404 @node Interfaces.VxWorks.IO (i-vxwoio.ads)
12405 @section @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
12406 @cindex @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
12407 @cindex Interfacing to VxWorks' I/O
12408 @cindex VxWorks, I/O interfacing
12409 @cindex VxWorks, Get_Immediate
12410 @cindex Get_Immediate, VxWorks
12413 This package provides a binding to the ioctl (IO/Control)
12414 function of VxWorks, defining a set of option values and
12415 function codes. A particular use of this package is
12416 to enable the use of Get_Immediate under VxWorks.
12418 @node System.Address_Image (s-addima.ads)
12419 @section @code{System.Address_Image} (@file{s-addima.ads})
12420 @cindex @code{System.Address_Image} (@file{s-addima.ads})
12421 @cindex Address image
12422 @cindex Image, of an address
12425 This function provides a useful debugging
12426 function that gives an (implementation dependent)
12427 string which identifies an address.
12429 @node System.Assertions (s-assert.ads)
12430 @section @code{System.Assertions} (@file{s-assert.ads})
12431 @cindex @code{System.Assertions} (@file{s-assert.ads})
12433 @cindex Assert_Failure, exception
12436 This package provides the declaration of the exception raised
12437 by an run-time assertion failure, as well as the routine that
12438 is used internally to raise this assertion.
12440 @node System.Memory (s-memory.ads)
12441 @section @code{System.Memory} (@file{s-memory.ads})
12442 @cindex @code{System.Memory} (@file{s-memory.ads})
12443 @cindex Memory allocation
12446 This package provides the interface to the low level routines used
12447 by the generated code for allocation and freeing storage for the
12448 default storage pool (analogous to the C routines malloc and free.
12449 It also provides a reallocation interface analogous to the C routine
12450 realloc. The body of this unit may be modified to provide alternative
12451 allocation mechanisms for the default pool, and in addition, direct
12452 calls to this unit may be made for low level allocation uses (for
12453 example see the body of @code{GNAT.Tables}).
12455 @node System.Partition_Interface (s-parint.ads)
12456 @section @code{System.Partition_Interface} (@file{s-parint.ads})
12457 @cindex @code{System.Partition_Interface} (@file{s-parint.ads})
12458 @cindex Partition intefacing functions
12461 This package provides facilities for partition interfacing. It
12462 is used primarily in a distribution context when using Annex E
12465 @node System.Restrictions (s-restri.ads)
12466 @section @code{System.Restrictions} (@file{s-restri.ads})
12467 @cindex @code{System.Restrictions} (@file{s-restri.ads})
12468 @cindex Run-time restrictions access
12471 This package provides facilities for accessing at run-time
12472 the status of restrictions specified at compile time for
12473 the partition. Information is available both with regard
12474 to actual restrictions specified, and with regard to
12475 compiler determined information on which restrictions
12476 are violated by one or more packages in the partition.
12478 @node System.Rident (s-rident.ads)
12479 @section @code{System.Rident} (@file{s-rident.ads})
12480 @cindex @code{System.Rident} (@file{s-rident.ads})
12481 @cindex Restrictions definitions
12484 This package provides definitions of the restrictions
12485 identifiers supported by GNAT, and also the format of
12486 the restrictions provided in package System.Restrictions.
12487 It is not normally necessary to @code{with} this generic package
12488 since the necessary instantiation is included in
12489 package System.Restrictions.
12491 @node System.Task_Info (s-tasinf.ads)
12492 @section @code{System.Task_Info} (@file{s-tasinf.ads})
12493 @cindex @code{System.Task_Info} (@file{s-tasinf.ads})
12494 @cindex Task_Info pragma
12497 This package provides target dependent functionality that is used
12498 to support the @code{Task_Info} pragma
12500 @node System.Wch_Cnv (s-wchcnv.ads)
12501 @section @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
12502 @cindex @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
12503 @cindex Wide Character, Representation
12504 @cindex Wide String, Conversion
12505 @cindex Representation of wide characters
12508 This package provides routines for converting between
12509 wide characters and a representation as a value of type
12510 @code{Standard.String}, using a specified wide character
12511 encoding method. It uses definitions in
12512 package @code{System.Wch_Con}.
12514 @node System.Wch_Con (s-wchcon.ads)
12515 @section @code{System.Wch_Con} (@file{s-wchcon.ads})
12516 @cindex @code{System.Wch_Con} (@file{s-wchcon.ads})
12519 This package provides definitions and descriptions of
12520 the various methods used for encoding wide characters
12521 in ordinary strings. These definitions are used by
12522 the package @code{System.Wch_Cnv}.
12524 @node Interfacing to Other Languages
12525 @chapter Interfacing to Other Languages
12527 The facilities in annex B of the Ada 95 Reference Manual are fully
12528 implemented in GNAT, and in addition, a full interface to C++ is
12532 * Interfacing to C::
12533 * Interfacing to C++::
12534 * Interfacing to COBOL::
12535 * Interfacing to Fortran::
12536 * Interfacing to non-GNAT Ada code::
12539 @node Interfacing to C
12540 @section Interfacing to C
12543 Interfacing to C with GNAT can use one of two approaches:
12547 The types in the package @code{Interfaces.C} may be used.
12549 Standard Ada types may be used directly. This may be less portable to
12550 other compilers, but will work on all GNAT compilers, which guarantee
12551 correspondence between the C and Ada types.
12555 Pragma @code{Convention C} may be applied to Ada types, but mostly has no
12556 effect, since this is the default. The following table shows the
12557 correspondence between Ada scalar types and the corresponding C types.
12562 @item Short_Integer
12564 @item Short_Short_Integer
12568 @item Long_Long_Integer
12576 @item Long_Long_Float
12577 This is the longest floating-point type supported by the hardware.
12581 Additionally, there are the following general correspondences between Ada
12585 Ada enumeration types map to C enumeration types directly if pragma
12586 @code{Convention C} is specified, which causes them to have int
12587 length. Without pragma @code{Convention C}, Ada enumeration types map to
12588 8, 16, or 32 bits (i.e.@: C types @code{signed char}, @code{short},
12589 @code{int}, respectively) depending on the number of values passed.
12590 This is the only case in which pragma @code{Convention C} affects the
12591 representation of an Ada type.
12594 Ada access types map to C pointers, except for the case of pointers to
12595 unconstrained types in Ada, which have no direct C equivalent.
12598 Ada arrays map directly to C arrays.
12601 Ada records map directly to C structures.
12604 Packed Ada records map to C structures where all members are bit fields
12605 of the length corresponding to the @code{@var{type}'Size} value in Ada.
12608 @node Interfacing to C++
12609 @section Interfacing to C++
12612 The interface to C++ makes use of the following pragmas, which are
12613 primarily intended to be constructed automatically using a binding generator
12614 tool, although it is possible to construct them by hand. No suitable binding
12615 generator tool is supplied with GNAT though.
12617 Using these pragmas it is possible to achieve complete
12618 inter-operability between Ada tagged types and C class definitions.
12619 See @ref{Implementation Defined Pragmas}, for more details.
12622 @item pragma CPP_Class ([Entity =>] @var{local_name})
12623 The argument denotes an entity in the current declarative region that is
12624 declared as a tagged or untagged record type. It indicates that the type
12625 corresponds to an externally declared C++ class type, and is to be laid
12626 out the same way that C++ would lay out the type.
12628 @item pragma CPP_Constructor ([Entity =>] @var{local_name})
12629 This pragma identifies an imported function (imported in the usual way
12630 with pragma @code{Import}) as corresponding to a C++ constructor.
12632 @item pragma CPP_Vtable @dots{}
12633 One @code{CPP_Vtable} pragma can be present for each component of type
12634 @code{CPP.Interfaces.Vtable_Ptr} in a record to which pragma @code{CPP_Class}
12638 @node Interfacing to COBOL
12639 @section Interfacing to COBOL
12642 Interfacing to COBOL is achieved as described in section B.4 of
12643 the Ada 95 reference manual.
12645 @node Interfacing to Fortran
12646 @section Interfacing to Fortran
12649 Interfacing to Fortran is achieved as described in section B.5 of the
12650 reference manual. The pragma @code{Convention Fortran}, applied to a
12651 multi-dimensional array causes the array to be stored in column-major
12652 order as required for convenient interface to Fortran.
12654 @node Interfacing to non-GNAT Ada code
12655 @section Interfacing to non-GNAT Ada code
12657 It is possible to specify the convention @code{Ada} in a pragma
12658 @code{Import} or pragma @code{Export}. However this refers to
12659 the calling conventions used by GNAT, which may or may not be
12660 similar enough to those used by some other Ada 83 or Ada 95
12661 compiler to allow interoperation.
12663 If arguments types are kept simple, and if the foreign compiler generally
12664 follows system calling conventions, then it may be possible to integrate
12665 files compiled by other Ada compilers, provided that the elaboration
12666 issues are adequately addressed (for example by eliminating the
12667 need for any load time elaboration).
12669 In particular, GNAT running on VMS is designed to
12670 be highly compatible with the DEC Ada 83 compiler, so this is one
12671 case in which it is possible to import foreign units of this type,
12672 provided that the data items passed are restricted to simple scalar
12673 values or simple record types without variants, or simple array
12674 types with fixed bounds.
12676 @node Specialized Needs Annexes
12677 @chapter Specialized Needs Annexes
12680 Ada 95 defines a number of specialized needs annexes, which are not
12681 required in all implementations. However, as described in this chapter,
12682 GNAT implements all of these special needs annexes:
12685 @item Systems Programming (Annex C)
12686 The Systems Programming Annex is fully implemented.
12688 @item Real-Time Systems (Annex D)
12689 The Real-Time Systems Annex is fully implemented.
12691 @item Distributed Systems (Annex E)
12692 Stub generation is fully implemented in the GNAT compiler. In addition,
12693 a complete compatible PCS is available as part of the GLADE system,
12694 a separate product. When the two
12695 products are used in conjunction, this annex is fully implemented.
12697 @item Information Systems (Annex F)
12698 The Information Systems annex is fully implemented.
12700 @item Numerics (Annex G)
12701 The Numerics Annex is fully implemented.
12703 @item Safety and Security (Annex H)
12704 The Safety and Security annex is fully implemented.
12707 @node Implementation of Specific Ada Features
12708 @chapter Implementation of Specific Ada Features
12711 This chapter describes the GNAT implementation of several Ada language
12715 * Machine Code Insertions::
12716 * GNAT Implementation of Tasking::
12717 * GNAT Implementation of Shared Passive Packages::
12718 * Code Generation for Array Aggregates::
12721 @node Machine Code Insertions
12722 @section Machine Code Insertions
12725 Package @code{Machine_Code} provides machine code support as described
12726 in the Ada 95 Reference Manual in two separate forms:
12729 Machine code statements, consisting of qualified expressions that
12730 fit the requirements of RM section 13.8.
12732 An intrinsic callable procedure, providing an alternative mechanism of
12733 including machine instructions in a subprogram.
12737 The two features are similar, and both are closely related to the mechanism
12738 provided by the asm instruction in the GNU C compiler. Full understanding
12739 and use of the facilities in this package requires understanding the asm
12740 instruction as described in @cite{Using the GNU Compiler Collection (GCC)}
12741 by Richard Stallman. The relevant section is titled ``Extensions to the C
12742 Language Family'' -> ``Assembler Instructions with C Expression Operands''.
12744 Calls to the function @code{Asm} and the procedure @code{Asm} have identical
12745 semantic restrictions and effects as described below. Both are provided so
12746 that the procedure call can be used as a statement, and the function call
12747 can be used to form a code_statement.
12749 The first example given in the GCC documentation is the C @code{asm}
12752 asm ("fsinx %1 %0" : "=f" (result) : "f" (angle));
12756 The equivalent can be written for GNAT as:
12758 @smallexample @c ada
12759 Asm ("fsinx %1 %0",
12760 My_Float'Asm_Output ("=f", result),
12761 My_Float'Asm_Input ("f", angle));
12765 The first argument to @code{Asm} is the assembler template, and is
12766 identical to what is used in GNU C@. This string must be a static
12767 expression. The second argument is the output operand list. It is
12768 either a single @code{Asm_Output} attribute reference, or a list of such
12769 references enclosed in parentheses (technically an array aggregate of
12772 The @code{Asm_Output} attribute denotes a function that takes two
12773 parameters. The first is a string, the second is the name of a variable
12774 of the type designated by the attribute prefix. The first (string)
12775 argument is required to be a static expression and designates the
12776 constraint for the parameter (e.g.@: what kind of register is
12777 required). The second argument is the variable to be updated with the
12778 result. The possible values for constraint are the same as those used in
12779 the RTL, and are dependent on the configuration file used to build the
12780 GCC back end. If there are no output operands, then this argument may
12781 either be omitted, or explicitly given as @code{No_Output_Operands}.
12783 The second argument of @code{@var{my_float}'Asm_Output} functions as
12784 though it were an @code{out} parameter, which is a little curious, but
12785 all names have the form of expressions, so there is no syntactic
12786 irregularity, even though normally functions would not be permitted
12787 @code{out} parameters. The third argument is the list of input
12788 operands. It is either a single @code{Asm_Input} attribute reference, or
12789 a list of such references enclosed in parentheses (technically an array
12790 aggregate of such references).
12792 The @code{Asm_Input} attribute denotes a function that takes two
12793 parameters. The first is a string, the second is an expression of the
12794 type designated by the prefix. The first (string) argument is required
12795 to be a static expression, and is the constraint for the parameter,
12796 (e.g.@: what kind of register is required). The second argument is the
12797 value to be used as the input argument. The possible values for the
12798 constant are the same as those used in the RTL, and are dependent on
12799 the configuration file used to built the GCC back end.
12801 If there are no input operands, this argument may either be omitted, or
12802 explicitly given as @code{No_Input_Operands}. The fourth argument, not
12803 present in the above example, is a list of register names, called the
12804 @dfn{clobber} argument. This argument, if given, must be a static string
12805 expression, and is a space or comma separated list of names of registers
12806 that must be considered destroyed as a result of the @code{Asm} call. If
12807 this argument is the null string (the default value), then the code
12808 generator assumes that no additional registers are destroyed.
12810 The fifth argument, not present in the above example, called the
12811 @dfn{volatile} argument, is by default @code{False}. It can be set to
12812 the literal value @code{True} to indicate to the code generator that all
12813 optimizations with respect to the instruction specified should be
12814 suppressed, and that in particular, for an instruction that has outputs,
12815 the instruction will still be generated, even if none of the outputs are
12816 used. See the full description in the GCC manual for further details.
12818 The @code{Asm} subprograms may be used in two ways. First the procedure
12819 forms can be used anywhere a procedure call would be valid, and
12820 correspond to what the RM calls ``intrinsic'' routines. Such calls can
12821 be used to intersperse machine instructions with other Ada statements.
12822 Second, the function forms, which return a dummy value of the limited
12823 private type @code{Asm_Insn}, can be used in code statements, and indeed
12824 this is the only context where such calls are allowed. Code statements
12825 appear as aggregates of the form:
12827 @smallexample @c ada
12828 Asm_Insn'(Asm (@dots{}));
12829 Asm_Insn'(Asm_Volatile (@dots{}));
12833 In accordance with RM rules, such code statements are allowed only
12834 within subprograms whose entire body consists of such statements. It is
12835 not permissible to intermix such statements with other Ada statements.
12837 Typically the form using intrinsic procedure calls is more convenient
12838 and more flexible. The code statement form is provided to meet the RM
12839 suggestion that such a facility should be made available. The following
12840 is the exact syntax of the call to @code{Asm}. As usual, if named notation
12841 is used, the arguments may be given in arbitrary order, following the
12842 normal rules for use of positional and named arguments)
12846 [Template =>] static_string_EXPRESSION
12847 [,[Outputs =>] OUTPUT_OPERAND_LIST ]
12848 [,[Inputs =>] INPUT_OPERAND_LIST ]
12849 [,[Clobber =>] static_string_EXPRESSION ]
12850 [,[Volatile =>] static_boolean_EXPRESSION] )
12852 OUTPUT_OPERAND_LIST ::=
12853 [PREFIX.]No_Output_Operands
12854 | OUTPUT_OPERAND_ATTRIBUTE
12855 | (OUTPUT_OPERAND_ATTRIBUTE @{,OUTPUT_OPERAND_ATTRIBUTE@})
12857 OUTPUT_OPERAND_ATTRIBUTE ::=
12858 SUBTYPE_MARK'Asm_Output (static_string_EXPRESSION, NAME)
12860 INPUT_OPERAND_LIST ::=
12861 [PREFIX.]No_Input_Operands
12862 | INPUT_OPERAND_ATTRIBUTE
12863 | (INPUT_OPERAND_ATTRIBUTE @{,INPUT_OPERAND_ATTRIBUTE@})
12865 INPUT_OPERAND_ATTRIBUTE ::=
12866 SUBTYPE_MARK'Asm_Input (static_string_EXPRESSION, EXPRESSION)
12870 The identifiers @code{No_Input_Operands} and @code{No_Output_Operands}
12871 are declared in the package @code{Machine_Code} and must be referenced
12872 according to normal visibility rules. In particular if there is no
12873 @code{use} clause for this package, then appropriate package name
12874 qualification is required.
12876 @node GNAT Implementation of Tasking
12877 @section GNAT Implementation of Tasking
12880 This chapter outlines the basic GNAT approach to tasking (in particular,
12881 a multi-layered library for portability) and discusses issues related
12882 to compliance with the Real-Time Systems Annex.
12885 * Mapping Ada Tasks onto the Underlying Kernel Threads::
12886 * Ensuring Compliance with the Real-Time Annex::
12889 @node Mapping Ada Tasks onto the Underlying Kernel Threads
12890 @subsection Mapping Ada Tasks onto the Underlying Kernel Threads
12893 GNAT's run-time support comprises two layers:
12896 @item GNARL (GNAT Run-time Layer)
12897 @item GNULL (GNAT Low-level Library)
12901 In GNAT, Ada's tasking services rely on a platform and OS independent
12902 layer known as GNARL@. This code is responsible for implementing the
12903 correct semantics of Ada's task creation, rendezvous, protected
12906 GNARL decomposes Ada's tasking semantics into simpler lower level
12907 operations such as create a thread, set the priority of a thread,
12908 yield, create a lock, lock/unlock, etc. The spec for these low-level
12909 operations constitutes GNULLI, the GNULL Interface. This interface is
12910 directly inspired from the POSIX real-time API@.
12912 If the underlying executive or OS implements the POSIX standard
12913 faithfully, the GNULL Interface maps as is to the services offered by
12914 the underlying kernel. Otherwise, some target dependent glue code maps
12915 the services offered by the underlying kernel to the semantics expected
12918 Whatever the underlying OS (VxWorks, UNIX, OS/2, Windows NT, etc.) the
12919 key point is that each Ada task is mapped on a thread in the underlying
12920 kernel. For example, in the case of VxWorks, one Ada task = one VxWorks task.
12922 In addition Ada task priorities map onto the underlying thread priorities.
12923 Mapping Ada tasks onto the underlying kernel threads has several advantages:
12927 The underlying scheduler is used to schedule the Ada tasks. This
12928 makes Ada tasks as efficient as kernel threads from a scheduling
12932 Interaction with code written in C containing threads is eased
12933 since at the lowest level Ada tasks and C threads map onto the same
12934 underlying kernel concept.
12937 When an Ada task is blocked during I/O the remaining Ada tasks are
12941 On multiprocessor systems Ada tasks can execute in parallel.
12945 Some threads libraries offer a mechanism to fork a new process, with the
12946 child process duplicating the threads from the parent.
12948 support this functionality when the parent contains more than one task.
12949 @cindex Forking a new process
12951 @node Ensuring Compliance with the Real-Time Annex
12952 @subsection Ensuring Compliance with the Real-Time Annex
12953 @cindex Real-Time Systems Annex compliance
12956 Although mapping Ada tasks onto
12957 the underlying threads has significant advantages, it does create some
12958 complications when it comes to respecting the scheduling semantics
12959 specified in the real-time annex (Annex D).
12961 For instance the Annex D requirement for the @code{FIFO_Within_Priorities}
12962 scheduling policy states:
12965 @emph{When the active priority of a ready task that is not running
12966 changes, or the setting of its base priority takes effect, the
12967 task is removed from the ready queue for its old active priority
12968 and is added at the tail of the ready queue for its new active
12969 priority, except in the case where the active priority is lowered
12970 due to the loss of inherited priority, in which case the task is
12971 added at the head of the ready queue for its new active priority.}
12975 While most kernels do put tasks at the end of the priority queue when
12976 a task changes its priority, (which respects the main
12977 FIFO_Within_Priorities requirement), almost none keep a thread at the
12978 beginning of its priority queue when its priority drops from the loss
12979 of inherited priority.
12981 As a result most vendors have provided incomplete Annex D implementations.
12983 The GNAT run-time, has a nice cooperative solution to this problem
12984 which ensures that accurate FIFO_Within_Priorities semantics are
12987 The principle is as follows. When an Ada task T is about to start
12988 running, it checks whether some other Ada task R with the same
12989 priority as T has been suspended due to the loss of priority
12990 inheritance. If this is the case, T yields and is placed at the end of
12991 its priority queue. When R arrives at the front of the queue it
12994 Note that this simple scheme preserves the relative order of the tasks
12995 that were ready to execute in the priority queue where R has been
12998 @node GNAT Implementation of Shared Passive Packages
12999 @section GNAT Implementation of Shared Passive Packages
13000 @cindex Shared passive packages
13003 GNAT fully implements the pragma @code{Shared_Passive} for
13004 @cindex pragma @code{Shared_Passive}
13005 the purpose of designating shared passive packages.
13006 This allows the use of passive partitions in the
13007 context described in the Ada Reference Manual; i.e. for communication
13008 between separate partitions of a distributed application using the
13009 features in Annex E.
13011 @cindex Distribution Systems Annex
13013 However, the implementation approach used by GNAT provides for more
13014 extensive usage as follows:
13017 @item Communication between separate programs
13019 This allows separate programs to access the data in passive
13020 partitions, using protected objects for synchronization where
13021 needed. The only requirement is that the two programs have a
13022 common shared file system. It is even possible for programs
13023 running on different machines with different architectures
13024 (e.g. different endianness) to communicate via the data in
13025 a passive partition.
13027 @item Persistence between program runs
13029 The data in a passive package can persist from one run of a
13030 program to another, so that a later program sees the final
13031 values stored by a previous run of the same program.
13036 The implementation approach used is to store the data in files. A
13037 separate stream file is created for each object in the package, and
13038 an access to an object causes the corresponding file to be read or
13041 The environment variable @code{SHARED_MEMORY_DIRECTORY} should be
13042 @cindex @code{SHARED_MEMORY_DIRECTORY} environment variable
13043 set to the directory to be used for these files.
13044 The files in this directory
13045 have names that correspond to their fully qualified names. For
13046 example, if we have the package
13048 @smallexample @c ada
13050 pragma Shared_Passive (X);
13057 and the environment variable is set to @code{/stemp/}, then the files created
13058 will have the names:
13066 These files are created when a value is initially written to the object, and
13067 the files are retained until manually deleted. This provides the persistence
13068 semantics. If no file exists, it means that no partition has assigned a value
13069 to the variable; in this case the initial value declared in the package
13070 will be used. This model ensures that there are no issues in synchronizing
13071 the elaboration process, since elaboration of passive packages elaborates the
13072 initial values, but does not create the files.
13074 The files are written using normal @code{Stream_IO} access.
13075 If you want to be able
13076 to communicate between programs or partitions running on different
13077 architectures, then you should use the XDR versions of the stream attribute
13078 routines, since these are architecture independent.
13080 If active synchronization is required for access to the variables in the
13081 shared passive package, then as described in the Ada Reference Manual, the
13082 package may contain protected objects used for this purpose. In this case
13083 a lock file (whose name is @file{___lock} (three underscores)
13084 is created in the shared memory directory.
13085 @cindex @file{___lock} file (for shared passive packages)
13086 This is used to provide the required locking
13087 semantics for proper protected object synchronization.
13089 As of January 2003, GNAT supports shared passive packages on all platforms
13090 except for OpenVMS.
13092 @node Code Generation for Array Aggregates
13093 @section Code Generation for Array Aggregates
13096 * Static constant aggregates with static bounds::
13097 * Constant aggregates with an unconstrained nominal types::
13098 * Aggregates with static bounds::
13099 * Aggregates with non-static bounds::
13100 * Aggregates in assignment statements::
13104 Aggregate have a rich syntax and allow the user to specify the values of
13105 complex data structures by means of a single construct. As a result, the
13106 code generated for aggregates can be quite complex and involve loops, case
13107 statements and multiple assignments. In the simplest cases, however, the
13108 compiler will recognize aggregates whose components and constraints are
13109 fully static, and in those cases the compiler will generate little or no
13110 executable code. The following is an outline of the code that GNAT generates
13111 for various aggregate constructs. For further details, the user will find it
13112 useful to examine the output produced by the -gnatG flag to see the expanded
13113 source that is input to the code generator. The user will also want to examine
13114 the assembly code generated at various levels of optimization.
13116 The code generated for aggregates depends on the context, the component values,
13117 and the type. In the context of an object declaration the code generated is
13118 generally simpler than in the case of an assignment. As a general rule, static
13119 component values and static subtypes also lead to simpler code.
13121 @node Static constant aggregates with static bounds
13122 @subsection Static constant aggregates with static bounds
13125 For the declarations:
13126 @smallexample @c ada
13127 type One_Dim is array (1..10) of integer;
13128 ar0 : constant One_Dim := ( 1, 2, 3, 4, 5, 6, 7, 8, 9, 0);
13132 GNAT generates no executable code: the constant ar0 is placed in static memory.
13133 The same is true for constant aggregates with named associations:
13135 @smallexample @c ada
13136 Cr1 : constant One_Dim := (4 => 16, 2 => 4, 3 => 9, 1=> 1);
13137 Cr3 : constant One_Dim := (others => 7777);
13141 The same is true for multidimensional constant arrays such as:
13143 @smallexample @c ada
13144 type two_dim is array (1..3, 1..3) of integer;
13145 Unit : constant two_dim := ( (1,0,0), (0,1,0), (0,0,1));
13149 The same is true for arrays of one-dimensional arrays: the following are
13152 @smallexample @c ada
13153 type ar1b is array (1..3) of boolean;
13154 type ar_ar is array (1..3) of ar1b;
13155 None : constant ar1b := (others => false); -- fully static
13156 None2 : constant ar_ar := (1..3 => None); -- fully static
13160 However, for multidimensional aggregates with named associations, GNAT will
13161 generate assignments and loops, even if all associations are static. The
13162 following two declarations generate a loop for the first dimension, and
13163 individual component assignments for the second dimension:
13165 @smallexample @c ada
13166 Zero1: constant two_dim := (1..3 => (1..3 => 0));
13167 Zero2: constant two_dim := (others => (others => 0));
13170 @node Constant aggregates with an unconstrained nominal types
13171 @subsection Constant aggregates with an unconstrained nominal types
13174 In such cases the aggregate itself establishes the subtype, so that
13175 associations with @code{others} cannot be used. GNAT determines the
13176 bounds for the actual subtype of the aggregate, and allocates the
13177 aggregate statically as well. No code is generated for the following:
13179 @smallexample @c ada
13180 type One_Unc is array (natural range <>) of integer;
13181 Cr_Unc : constant One_Unc := (12,24,36);
13184 @node Aggregates with static bounds
13185 @subsection Aggregates with static bounds
13188 In all previous examples the aggregate was the initial (and immutable) value
13189 of a constant. If the aggregate initializes a variable, then code is generated
13190 for it as a combination of individual assignments and loops over the target
13191 object. The declarations
13193 @smallexample @c ada
13194 Cr_Var1 : One_Dim := (2, 5, 7, 11);
13195 Cr_Var2 : One_Dim := (others > -1);
13199 generate the equivalent of
13201 @smallexample @c ada
13207 for I in Cr_Var2'range loop
13208 Cr_Var2 (I) := =-1;
13212 @node Aggregates with non-static bounds
13213 @subsection Aggregates with non-static bounds
13216 If the bounds of the aggregate are not statically compatible with the bounds
13217 of the nominal subtype of the target, then constraint checks have to be
13218 generated on the bounds. For a multidimensional array, constraint checks may
13219 have to be applied to sub-arrays individually, if they do not have statically
13220 compatible subtypes.
13222 @node Aggregates in assignment statements
13223 @subsection Aggregates in assignment statements
13226 In general, aggregate assignment requires the construction of a temporary,
13227 and a copy from the temporary to the target of the assignment. This is because
13228 it is not always possible to convert the assignment into a series of individual
13229 component assignments. For example, consider the simple case:
13231 @smallexample @c ada
13236 This cannot be converted into:
13238 @smallexample @c ada
13244 So the aggregate has to be built first in a separate location, and then
13245 copied into the target. GNAT recognizes simple cases where this intermediate
13246 step is not required, and the assignments can be performed in place, directly
13247 into the target. The following sufficient criteria are applied:
13251 The bounds of the aggregate are static, and the associations are static.
13253 The components of the aggregate are static constants, names of
13254 simple variables that are not renamings, or expressions not involving
13255 indexed components whose operands obey these rules.
13259 If any of these conditions are violated, the aggregate will be built in
13260 a temporary (created either by the front-end or the code generator) and then
13261 that temporary will be copied onto the target.
13263 @node Project File Reference
13264 @chapter Project File Reference
13267 This chapter describes the syntax and semantics of project files.
13268 Project files specify the options to be used when building a system.
13269 Project files can specify global settings for all tools,
13270 as well as tool-specific settings.
13271 See the chapter on project files in the GNAT Users guide for examples of use.
13275 * Lexical Elements::
13277 * Typed string declarations::
13281 * Project Attributes::
13282 * Attribute References::
13283 * External Values::
13284 * Case Construction::
13286 * Package Renamings::
13288 * Project Extensions::
13289 * Project File Elaboration::
13292 @node Reserved Words
13293 @section Reserved Words
13296 All Ada95 reserved words are reserved in project files, and cannot be used
13297 as variable names or project names. In addition, the following are
13298 also reserved in project files:
13301 @item @code{extends}
13303 @item @code{external}
13305 @item @code{project}
13309 @node Lexical Elements
13310 @section Lexical Elements
13313 Rules for identifiers are the same as in Ada95. Identifiers
13314 are case-insensitive. Strings are case sensitive, except where noted.
13315 Comments have the same form as in Ada95.
13325 simple_name @{. simple_name@}
13329 @section Declarations
13332 Declarations introduce new entities that denote types, variables, attributes,
13333 and packages. Some declarations can only appear immediately within a project
13334 declaration. Others can appear within a project or within a package.
13338 declarative_item ::=
13339 simple_declarative_item |
13340 typed_string_declaration |
13341 package_declaration
13343 simple_declarative_item ::=
13344 variable_declaration |
13345 typed_variable_declaration |
13346 attribute_declaration |
13350 @node Typed string declarations
13351 @section Typed string declarations
13354 Typed strings are sequences of string literals. Typed strings are the only
13355 named types in project files. They are used in case constructions, where they
13356 provide support for conditional attribute definitions.
13360 typed_string_declaration ::=
13361 @b{type} <typed_string_>_simple_name @b{is}
13362 ( string_literal @{, string_literal@} );
13366 A typed string declaration can only appear immediately within a project
13369 All the string literals in a typed string declaration must be distinct.
13375 Variables denote values, and appear as constituents of expressions.
13378 typed_variable_declaration ::=
13379 <typed_variable_>simple_name : <typed_string_>name := string_expression ;
13381 variable_declaration ::=
13382 <variable_>simple_name := expression;
13386 The elaboration of a variable declaration introduces the variable and
13387 assigns to it the value of the expression. The name of the variable is
13388 available after the assignment symbol.
13391 A typed_variable can only be declare once.
13394 a non typed variable can be declared multiple times.
13397 Before the completion of its first declaration, the value of variable
13398 is the null string.
13401 @section Expressions
13404 An expression is a formula that defines a computation or retrieval of a value.
13405 In a project file the value of an expression is either a string or a list
13406 of strings. A string value in an expression is either a literal, the current
13407 value of a variable, an external value, an attribute reference, or a
13408 concatenation operation.
13421 attribute_reference
13427 ( <string_>expression @{ , <string_>expression @} )
13430 @subsection Concatenation
13432 The following concatenation functions are defined:
13434 @smallexample @c ada
13435 function "&" (X : String; Y : String) return String;
13436 function "&" (X : String_List; Y : String) return String_List;
13437 function "&" (X : String_List; Y : String_List) return String_List;
13441 @section Attributes
13444 An attribute declaration defines a property of a project or package. This
13445 property can later be queried by means of an attribute reference.
13446 Attribute values are strings or string lists.
13448 Some attributes are associative arrays. These attributes are mappings whose
13449 domain is a set of strings. These attributes are declared one association
13450 at a time, by specifying a point in the domain and the corresponding image
13451 of the attribute. They may also be declared as a full associative array,
13452 getting the same associations as the corresponding attribute in an imported
13453 or extended project.
13455 Attributes that are not associative arrays are called simple attributes.
13459 attribute_declaration ::=
13460 full_associative_array_declaration |
13461 @b{for} attribute_designator @b{use} expression ;
13463 full_associative_array_declaration ::=
13464 @b{for} <associative_array_attribute_>simple_name @b{use}
13465 <project_>simple_name [ . <package_>simple_Name ] ' <attribute_>simple_name ;
13467 attribute_designator ::=
13468 <simple_attribute_>simple_name |
13469 <associative_array_attribute_>simple_name ( string_literal )
13473 Some attributes are project-specific, and can only appear immediately within
13474 a project declaration. Others are package-specific, and can only appear within
13475 the proper package.
13477 The expression in an attribute definition must be a string or a string_list.
13478 The string literal appearing in the attribute_designator of an associative
13479 array attribute is case-insensitive.
13481 @node Project Attributes
13482 @section Project Attributes
13485 The following attributes apply to a project. All of them are simple
13490 Expression must be a path name. The attribute defines the
13491 directory in which the object files created by the build are to be placed. If
13492 not specified, object files are placed in the project directory.
13495 Expression must be a path name. The attribute defines the
13496 directory in which the executables created by the build are to be placed.
13497 If not specified, executables are placed in the object directory.
13500 Expression must be a list of path names. The attribute
13501 defines the directories in which the source files for the project are to be
13502 found. If not specified, source files are found in the project directory.
13505 Expression must be a list of file names. The attribute
13506 defines the individual files, in the project directory, which are to be used
13507 as sources for the project. File names are path_names that contain no directory
13508 information. If the project has no sources the attribute must be declared
13509 explicitly with an empty list.
13511 @item Source_List_File
13512 Expression must a single path name. The attribute
13513 defines a text file that contains a list of source file names to be used
13514 as sources for the project
13517 Expression must be a path name. The attribute defines the
13518 directory in which a library is to be built. The directory must exist, must
13519 be distinct from the project's object directory, and must be writable.
13522 Expression must be a string that is a legal file name,
13523 without extension. The attribute defines a string that is used to generate
13524 the name of the library to be built by the project.
13527 Argument must be a string value that must be one of the
13528 following @code{"static"}, @code{"dynamic"} or @code{"relocatable"}. This
13529 string is case-insensitive. If this attribute is not specified, the library is
13530 a static library. Otherwise, the library may be dynamic or relocatable. This
13531 distinction is operating-system dependent.
13533 @item Library_Version
13534 Expression must be a string value whose interpretation
13535 is platform dependent. On UNIX, it is used only for dynamic/relocatable
13536 libraries as the internal name of the library (the @code{"soname"}). If the
13537 library file name (built from the @code{Library_Name}) is different from the
13538 @code{Library_Version}, then the library file will be a symbolic link to the
13539 actual file whose name will be @code{Library_Version}.
13541 @item Library_Interface
13542 Expression must be a string list. Each element of the string list
13543 must designate a unit of the project.
13544 If this attribute is present in a Library Project File, then the project
13545 file is a Stand-alone Library_Project_File.
13547 @item Library_Auto_Init
13548 Expression must be a single string "true" or "false", case-insensitive.
13549 If this attribute is present in a Stand-alone Library Project File,
13550 it indicates if initialization is automatic when the dynamic library
13553 @item Library_Options
13554 Expression must be a string list. Indicates additional switches that
13555 are to be used when building a shared library.
13558 Expression must be a single string. Designates an alternative to "gcc"
13559 for building shared libraries.
13561 @item Library_Src_Dir
13562 Expression must be a path name. The attribute defines the
13563 directory in which the sources of the interfaces of a Stand-alone Library will
13564 be copied. The directory must exist, must be distinct from the project's
13565 object directory and source directories, and must be writable.
13568 Expression must be a list of strings that are legal file names.
13569 These file names designate existing compilation units in the source directory
13570 that are legal main subprograms.
13572 When a project file is elaborated, as part of the execution of a gnatmake
13573 command, one or several executables are built and placed in the Exec_Dir.
13574 If the gnatmake command does not include explicit file names, the executables
13575 that are built correspond to the files specified by this attribute.
13577 @item Main_Language
13578 This is a simple attribute. Its value is a string that specifies the
13579 language of the main program.
13582 Expression must be a string list. Each string designates
13583 a programming language that is known to GNAT. The strings are case-insensitive.
13585 @item Locally_Removed_Files
13586 This attribute is legal only in a project file that extends another.
13587 Expression must be a list of strings that are legal file names.
13588 Each file name must designate a source that would normally be inherited
13589 by the current project file. It cannot designate an immediate source that is
13590 not inherited. Each of the source files in the list are not considered to
13591 be sources of the project file: they are not inherited.
13594 @node Attribute References
13595 @section Attribute References
13598 Attribute references are used to retrieve the value of previously defined
13599 attribute for a package or project.
13602 attribute_reference ::=
13603 attribute_prefix ' <simple_attribute_>simple_name [ ( string_literal ) ]
13605 attribute_prefix ::=
13607 <project_simple_name | package_identifier |
13608 <project_>simple_name . package_identifier
13612 If an attribute has not been specified for a given package or project, its
13613 value is the null string or the empty list.
13615 @node External Values
13616 @section External Values
13619 An external value is an expression whose value is obtained from the command
13620 that invoked the processing of the current project file (typically a
13626 @b{external} ( string_literal [, string_literal] )
13630 The first string_literal is the string to be used on the command line or
13631 in the environment to specify the external value. The second string_literal,
13632 if present, is the default to use if there is no specification for this
13633 external value either on the command line or in the environment.
13635 @node Case Construction
13636 @section Case Construction
13639 A case construction supports attribute declarations that depend on the value of
13640 a previously declared variable.
13644 case_construction ::=
13645 @b{case} <typed_variable_>name @b{is}
13650 @b{when} discrete_choice_list =>
13651 @{case_construction | attribute_declaration@}
13653 discrete_choice_list ::=
13654 string_literal @{| string_literal@} |
13659 All choices in a choice list must be distinct. The choice lists of two
13660 distinct alternatives must be disjoint. Unlike Ada, the choice lists of all
13661 alternatives do not need to include all values of the type. An @code{others}
13662 choice must appear last in the list of alternatives.
13668 A package provides a grouping of variable declarations and attribute
13669 declarations to be used when invoking various GNAT tools. The name of
13670 the package indicates the tool(s) to which it applies.
13674 package_declaration ::=
13675 package_specification | package_renaming
13677 package_specification ::=
13678 @b{package} package_identifier @b{is}
13679 @{simple_declarative_item@}
13680 @b{end} package_identifier ;
13682 package_identifier ::=
13683 @code{Naming} | @code{Builder} | @code{Compiler} | @code{Binder} |
13684 @code{Linker} | @code{Finder} | @code{Cross_Reference} |
13685 @code{gnatls} | @code{IDE} | @code{Pretty_Printer}
13688 @subsection Package Naming
13691 The attributes of a @code{Naming} package specifies the naming conventions
13692 that apply to the source files in a project. When invoking other GNAT tools,
13693 they will use the sources in the source directories that satisfy these
13694 naming conventions.
13696 The following attributes apply to a @code{Naming} package:
13700 This is a simple attribute whose value is a string. Legal values of this
13701 string are @code{"lowercase"}, @code{"uppercase"} or @code{"mixedcase"}.
13702 These strings are themselves case insensitive.
13705 If @code{Casing} is not specified, then the default is @code{"lowercase"}.
13707 @item Dot_Replacement
13708 This is a simple attribute whose string value satisfies the following
13712 @item It must not be empty
13713 @item It cannot start or end with an alphanumeric character
13714 @item It cannot be a single underscore
13715 @item It cannot start with an underscore followed by an alphanumeric
13716 @item It cannot contain a dot @code{'.'} if longer than one character
13720 If @code{Dot_Replacement} is not specified, then the default is @code{"-"}.
13723 This is an associative array attribute, defined on language names,
13724 whose image is a string that must satisfy the following
13728 @item It must not be empty
13729 @item It cannot start with an alphanumeric character
13730 @item It cannot start with an underscore followed by an alphanumeric character
13734 For Ada, the attribute denotes the suffix used in file names that contain
13735 library unit declarations, that is to say units that are package and
13736 subprogram declarations. If @code{Spec_Suffix ("Ada")} is not
13737 specified, then the default is @code{".ads"}.
13739 For C and C++, the attribute denotes the suffix used in file names that
13740 contain prototypes.
13743 This is an associative array attribute defined on language names,
13744 whose image is a string that must satisfy the following
13748 @item It must not be empty
13749 @item It cannot start with an alphanumeric character
13750 @item It cannot start with an underscore followed by an alphanumeric character
13751 @item It cannot be a suffix of @code{Spec_Suffix}
13755 For Ada, the attribute denotes the suffix used in file names that contain
13756 library bodies, that is to say units that are package and subprogram bodies.
13757 If @code{Body_Suffix ("Ada")} is not specified, then the default is
13760 For C and C++, the attribute denotes the suffix used in file names that contain
13763 @item Separate_Suffix
13764 This is a simple attribute whose value satisfies the same conditions as
13765 @code{Body_Suffix}.
13767 This attribute is specific to Ada. It denotes the suffix used in file names
13768 that contain separate bodies. If it is not specified, then it defaults to same
13769 value as @code{Body_Suffix ("Ada")}.
13772 This is an associative array attribute, specific to Ada, defined over
13773 compilation unit names. The image is a string that is the name of the file
13774 that contains that library unit. The file name is case sensitive if the
13775 conventions of the host operating system require it.
13778 This is an associative array attribute, specific to Ada, defined over
13779 compilation unit names. The image is a string that is the name of the file
13780 that contains the library unit body for the named unit. The file name is case
13781 sensitive if the conventions of the host operating system require it.
13783 @item Specification_Exceptions
13784 This is an associative array attribute defined on language names,
13785 whose value is a list of strings.
13787 This attribute is not significant for Ada.
13789 For C and C++, each string in the list denotes the name of a file that
13790 contains prototypes, but whose suffix is not necessarily the
13791 @code{Spec_Suffix} for the language.
13793 @item Implementation_Exceptions
13794 This is an associative array attribute defined on language names,
13795 whose value is a list of strings.
13797 This attribute is not significant for Ada.
13799 For C and C++, each string in the list denotes the name of a file that
13800 contains source code, but whose suffix is not necessarily the
13801 @code{Body_Suffix} for the language.
13804 The following attributes of package @code{Naming} are obsolescent. They are
13805 kept as synonyms of other attributes for compatibility with previous versions
13806 of the Project Manager.
13809 @item Specification_Suffix
13810 This is a synonym of @code{Spec_Suffix}.
13812 @item Implementation_Suffix
13813 This is a synonym of @code{Body_Suffix}.
13815 @item Specification
13816 This is a synonym of @code{Spec}.
13818 @item Implementation
13819 This is a synonym of @code{Body}.
13822 @subsection package Compiler
13825 The attributes of the @code{Compiler} package specify the compilation options
13826 to be used by the underlying compiler.
13829 @item Default_Switches
13830 This is an associative array attribute. Its
13831 domain is a set of language names. Its range is a string list that
13832 specifies the compilation options to be used when compiling a component
13833 written in that language, for which no file-specific switches have been
13837 This is an associative array attribute. Its domain is
13838 a set of file names. Its range is a string list that specifies the
13839 compilation options to be used when compiling the named file. If a file
13840 is not specified in the Switches attribute, it is compiled with the
13841 settings specified by Default_Switches.
13843 @item Local_Configuration_Pragmas.
13844 This is a simple attribute, whose
13845 value is a path name that designates a file containing configuration pragmas
13846 to be used for all invocations of the compiler for immediate sources of the
13850 This is an associative array attribute. Its domain is
13851 a set of main source file names. Its range is a simple string that specifies
13852 the executable file name to be used when linking the specified main source.
13853 If a main source is not specified in the Executable attribute, the executable
13854 file name is deducted from the main source file name.
13857 @subsection package Builder
13860 The attributes of package @code{Builder} specify the compilation, binding, and
13861 linking options to be used when building an executable for a project. The
13862 following attributes apply to package @code{Builder}:
13865 @item Default_Switches
13871 @item Global_Configuration_Pragmas
13872 This is a simple attribute, whose
13873 value is a path name that designates a file that contains configuration pragmas
13874 to be used in every build of an executable. If both local and global
13875 configuration pragmas are specified, a compilation makes use of both sets.
13878 This is an associative array attribute, defined over
13879 compilation unit names. The image is a string that is the name of the
13880 executable file corresponding to the main source file index.
13881 This attribute has no effect if its value is the empty string.
13883 @item Executable_Suffix
13884 This is a simple attribute whose value is a suffix to be added to
13885 the executables that don't have an attribute Executable specified.
13888 @subsection package Gnatls
13891 The attributes of package @code{Gnatls} specify the tool options to be used
13892 when invoking the library browser @command{gnatls}.
13893 The following attributes apply to package @code{Gnatls}:
13900 @subsection package Binder
13903 The attributes of package @code{Binder} specify the options to be used
13904 when invoking the binder in the construction of an executable.
13905 The following attributes apply to package @code{Binder}:
13908 @item Default_Switches
13914 @subsection package Linker
13917 The attributes of package @code{Linker} specify the options to be used when
13918 invoking the linker in the construction of an executable.
13919 The following attributes apply to package @code{Linker}:
13922 @item Default_Switches
13928 @subsection package Cross_Reference
13931 The attributes of package @code{Cross_Reference} specify the tool options
13933 when invoking the library tool @command{gnatxref}.
13934 The following attributes apply to package @code{Cross_Reference}:
13937 @item Default_Switches
13943 @subsection package Finder
13946 The attributes of package @code{Finder} specify the tool options to be used
13947 when invoking the search tool @command{gnatfind}.
13948 The following attributes apply to package @code{Finder}:
13951 @item Default_Switches
13957 @subsection package Pretty_Printer
13960 The attributes of package @code{Pretty_Printer}
13961 specify the tool options to be used
13962 when invoking the formatting tool @command{gnatpp}.
13963 The following attributes apply to package @code{Pretty_Printer}:
13966 @item Default_switches
13972 @subsection package IDE
13975 The attributes of package @code{IDE} specify the options to be used when using
13976 an Integrated Development Environment such as @command{GPS}.
13980 This is a simple attribute. Its value is a string that designates the remote
13981 host in a cross-compilation environment, to be used for remote compilation and
13982 debugging. This field should not be specified when running on the local
13986 This is a simple attribute. Its value is a string that specifies the
13987 name of IP address of the embedded target in a cross-compilation environment,
13988 on which the program should execute.
13990 @item Communication_Protocol
13991 This is a simple string attribute. Its value is the name of the protocol
13992 to use to communicate with the target in a cross-compilation environment,
13993 e.g. @code{"wtx"} or @code{"vxworks"}.
13995 @item Compiler_Command
13996 This is an associative array attribute, whose domain is a language name. Its
13997 value is string that denotes the command to be used to invoke the compiler.
13998 The value of @code{Compiler_Command ("Ada")} is expected to be compatible with
13999 gnatmake, in particular in the handling of switches.
14001 @item Debugger_Command
14002 This is simple attribute, Its value is a string that specifies the name of
14003 the debugger to be used, such as gdb, powerpc-wrs-vxworks-gdb or gdb-4.
14005 @item Default_Switches
14006 This is an associative array attribute. Its indexes are the name of the
14007 external tools that the GNAT Programming System (GPS) is supporting. Its
14008 value is a list of switches to use when invoking that tool.
14011 This is a simple attribute. Its value is a string that specifies the name
14012 of the @command{gnatls} utility to be used to retrieve information about the
14013 predefined path; e.g., @code{"gnatls"}, @code{"powerpc-wrs-vxworks-gnatls"}.
14016 This is a simple atribute. Is value is a string used to specify the
14017 Version Control System (VCS) to be used for this project, e.g CVS, RCS
14018 ClearCase or Perforce.
14020 @item VCS_File_Check
14021 This is a simple attribute. Its value is a string that specifies the
14022 command used by the VCS to check the validity of a file, either
14023 when the user explicitly asks for a check, or as a sanity check before
14024 doing the check-in.
14026 @item VCS_Log_Check
14027 This is a simple attribute. Its value is a string that specifies
14028 the command used by the VCS to check the validity of a log file.
14032 @node Package Renamings
14033 @section Package Renamings
14036 A package can be defined by a renaming declaration. The new package renames
14037 a package declared in a different project file, and has the same attributes
14038 as the package it renames.
14041 package_renaming ::==
14042 @b{package} package_identifier @b{renames}
14043 <project_>simple_name.package_identifier ;
14047 The package_identifier of the renamed package must be the same as the
14048 package_identifier. The project whose name is the prefix of the renamed
14049 package must contain a package declaration with this name. This project
14050 must appear in the context_clause of the enclosing project declaration,
14051 or be the parent project of the enclosing child project.
14057 A project file specifies a set of rules for constructing a software system.
14058 A project file can be self-contained, or depend on other project files.
14059 Dependencies are expressed through a context clause that names other projects.
14065 context_clause project_declaration
14067 project_declaration ::=
14068 simple_project_declaration | project_extension
14070 simple_project_declaration ::=
14071 @b{project} <project_>simple_name @b{is}
14072 @{declarative_item@}
14073 @b{end} <project_>simple_name;
14079 [@b{limited}] @b{with} path_name @{ , path_name @} ;
14086 A path name denotes a project file. A path name can be absolute or relative.
14087 An absolute path name includes a sequence of directories, in the syntax of
14088 the host operating system, that identifies uniquely the project file in the
14089 file system. A relative path name identifies the project file, relative
14090 to the directory that contains the current project, or relative to a
14091 directory listed in the environment variable ADA_PROJECT_PATH.
14092 Path names are case sensitive if file names in the host operating system
14093 are case sensitive.
14095 The syntax of the environment variable ADA_PROJECT_PATH is a list of
14096 directory names separated by colons (semicolons on Windows).
14098 A given project name can appear only once in a context_clause.
14100 It is illegal for a project imported by a context clause to refer, directly
14101 or indirectly, to the project in which this context clause appears (the
14102 dependency graph cannot contain cycles), except when one of the with_clause
14103 in the cycle is a @code{limited with}.
14105 @node Project Extensions
14106 @section Project Extensions
14109 A project extension introduces a new project, which inherits the declarations
14110 of another project.
14114 project_extension ::=
14115 @b{project} <project_>simple_name @b{extends} path_name @b{is}
14116 @{declarative_item@}
14117 @b{end} <project_>simple_name;
14121 The project extension declares a child project. The child project inherits
14122 all the declarations and all the files of the parent project, These inherited
14123 declaration can be overridden in the child project, by means of suitable
14126 @node Project File Elaboration
14127 @section Project File Elaboration
14130 A project file is processed as part of the invocation of a gnat tool that
14131 uses the project option. Elaboration of the process file consists in the
14132 sequential elaboration of all its declarations. The computed values of
14133 attributes and variables in the project are then used to establish the
14134 environment in which the gnat tool will execute.
14137 @c GNU Free Documentation License
14139 @node Index,,GNU Free Documentation License, Top