1 \input texinfo @c -*-texinfo-*-
5 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
7 @c GNAT DOCUMENTATION o
11 @c GNAT is maintained by Ada Core Technologies Inc (http://www.gnat.com). o
13 @c oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
15 @setfilename gnat_rm.info
18 Copyright @copyright{} 1995-2012, Free Software Foundation, Inc.
20 Permission is granted to copy, distribute and/or modify this document
21 under the terms of the GNU Free Documentation License, Version 1.3 or
22 any later version published by the Free Software Foundation; with no
23 Invariant Sections, with the Front-Cover Texts being ``GNAT Reference
24 Manual'', and with no Back-Cover Texts. A copy of the license is
25 included in the section entitled ``GNU Free Documentation License''.
30 @settitle GNAT Reference Manual
32 @setchapternewpage odd
35 @include gcc-common.texi
37 @dircategory GNU Ada tools
39 * GNAT Reference Manual: (gnat_rm). Reference Manual for GNU Ada tools.
43 @title GNAT Reference Manual
44 @subtitle GNAT, The GNU Ada Development Environment
48 @vskip 0pt plus 1filll
55 @node Top, About This Guide, (dir), (dir)
56 @top GNAT Reference Manual
62 GNAT, The GNU Ada Development Environment@*
63 GCC version @value{version-GCC}@*
70 * Implementation Defined Pragmas::
71 * Implementation Defined Aspects::
72 * Implementation Defined Attributes::
73 * Standard and Implementation Defined Restrictions::
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 * Implementation of Ada 2012 Features::
85 * Obsolescent Features::
86 * GNU Free Documentation License::
89 --- The Detailed Node Listing ---
93 * What This Reference Manual Contains::
94 * Related Information::
96 Implementation Defined Pragmas
98 * Pragma Abort_Defer::
99 * Pragma Abstract_State::
106 * Pragma Allow_Integer_Address::
109 * Pragma Assert_And_Cut::
110 * Pragma Assertion_Policy::
112 * Pragma Assume_No_Invalid_Values::
113 * Pragma Attribute_Definition::
115 * Pragma C_Pass_By_Copy::
117 * Pragma Check_Float_Overflow::
118 * Pragma Check_Name::
119 * Pragma Check_Policy::
120 * Pragma CIL_Constructor::
122 * Pragma Common_Object::
123 * Pragma Compile_Time_Error::
124 * Pragma Compile_Time_Warning::
125 * Pragma Compiler_Unit::
126 * Pragma Compiler_Unit_Warning::
127 * Pragma Complete_Representation::
128 * Pragma Complex_Representation::
129 * Pragma Component_Alignment::
130 * Pragma Contract_Cases::
131 * Pragma Convention_Identifier::
133 * Pragma CPP_Constructor::
134 * Pragma CPP_Virtual::
135 * Pragma CPP_Vtable::
138 * Pragma Debug_Policy::
139 * Pragma Default_Storage_Pool::
141 * Pragma Detect_Blocking::
142 * Pragma Disable_Atomic_Synchronization::
143 * Pragma Dispatching_Domain::
144 * Pragma Elaboration_Checks::
146 * Pragma Enable_Atomic_Synchronization::
147 * Pragma Export_Exception::
148 * Pragma Export_Function::
149 * Pragma Export_Object::
150 * Pragma Export_Procedure::
151 * Pragma Export_Value::
152 * Pragma Export_Valued_Procedure::
153 * Pragma Extend_System::
154 * Pragma Extensions_Allowed::
156 * Pragma External_Name_Casing::
158 * Pragma Favor_Top_Level::
159 * Pragma Finalize_Storage_Only::
160 * Pragma Float_Representation::
163 * Pragma Implementation_Defined::
164 * Pragma Implemented::
165 * Pragma Implicit_Packing::
166 * Pragma Import_Exception::
167 * Pragma Import_Function::
168 * Pragma Import_Object::
169 * Pragma Import_Procedure::
170 * Pragma Import_Valued_Procedure::
171 * Pragma Independent::
172 * Pragma Independent_Components::
173 * Pragma Initial_Condition::
174 * Pragma Initialize_Scalars::
175 * Pragma Initializes::
176 * Pragma Inline_Always::
177 * Pragma Inline_Generic::
179 * Pragma Interface_Name::
180 * Pragma Interrupt_Handler::
181 * Pragma Interrupt_State::
183 * Pragma Java_Constructor::
184 * Pragma Java_Interface::
185 * Pragma Keep_Names::
188 * Pragma Linker_Alias::
189 * Pragma Linker_Constructor::
190 * Pragma Linker_Destructor::
191 * Pragma Linker_Section::
192 * Pragma Long_Float::
193 * Pragma Loop_Invariant::
194 * Pragma Loop_Optimize::
195 * Pragma Loop_Variant::
196 * Pragma Machine_Attribute::
198 * Pragma Main_Storage::
202 * Pragma No_Run_Time::
203 * Pragma No_Strict_Aliasing ::
204 * Pragma Normalize_Scalars::
205 * Pragma Obsolescent::
206 * Pragma Optimize_Alignment::
208 * Pragma Overflow_Mode::
209 * Pragma Overriding_Renamings::
210 * Pragma Partition_Elaboration_Policy::
212 * Pragma Persistent_BSS::
215 * Pragma Postcondition::
216 * Pragma Post_Class::
218 * Pragma Precondition::
220 * Pragma Preelaborable_Initialization::
221 * Pragma Preelaborate_05::
223 * Pragma Priority_Specific_Dispatching::
225 * Pragma Profile_Warnings::
226 * Pragma Propagate_Exceptions::
227 * Pragma Provide_Shift_Operators::
228 * Pragma Psect_Object::
231 * Pragma Pure_Function::
233 * Pragma Refined_State::
234 * Pragma Relative_Deadline::
235 * Pragma Remote_Access_Type::
236 * Pragma Restricted_Run_Time::
237 * Pragma Restriction_Warnings::
238 * Pragma Reviewable::
239 * Pragma Share_Generic::
241 * Pragma Short_Circuit_And_Or::
242 * Pragma Short_Descriptors::
243 * Pragma Simple_Storage_Pool_Type::
244 * Pragma Source_File_Name::
245 * Pragma Source_File_Name_Project::
246 * Pragma Source_Reference::
247 * Pragma SPARK_Mode::
248 * Pragma Static_Elaboration_Desired::
249 * Pragma Stream_Convert::
250 * Pragma Style_Checks::
253 * Pragma Suppress_All::
254 * Pragma Suppress_Debug_Info::
255 * Pragma Suppress_Exception_Locations::
256 * Pragma Suppress_Initialization::
259 * Pragma Task_Storage::
261 * Pragma Thread_Local_Storage::
262 * Pragma Time_Slice::
264 * Pragma Type_Invariant::
265 * Pragma Type_Invariant_Class::
266 * Pragma Unchecked_Union::
267 * Pragma Unimplemented_Unit::
268 * Pragma Universal_Aliasing ::
269 * Pragma Universal_Data::
270 * Pragma Unmodified::
271 * Pragma Unreferenced::
272 * Pragma Unreferenced_Objects::
273 * Pragma Unreserve_All_Interrupts::
274 * Pragma Unsuppress::
275 * Pragma Use_VADS_Size::
276 * Pragma Validity_Checks::
278 * Pragma Warning_As_Error::
280 * Pragma Weak_External::
281 * Pragma Wide_Character_Encoding::
283 Implementation Defined Aspects
285 * Aspect Abstract_State::
286 * Aspect Contract_Cases::
289 * Aspect Dimension_System::
290 * Aspect Favor_Top_Level::
292 * Aspect Initial_Condition::
293 * Aspect Initializes::
294 * Aspect Inline_Always::
296 * Aspect Linker_Section::
297 * Aspect Object_Size::
298 * Aspect Persistent_BSS::
300 * Aspect Preelaborate_05::
303 * Aspect Pure_Function::
304 * Aspect Refined_State::
305 * Aspect Remote_Access_Type::
306 * Aspect Scalar_Storage_Order::
308 * Aspect Simple_Storage_Pool::
309 * Aspect Simple_Storage_Pool_Type::
310 * Aspect SPARK_Mode::
311 * Aspect Suppress_Debug_Info::
313 * Aspect Universal_Aliasing::
314 * Aspect Universal_Data::
315 * Aspect Unmodified::
316 * Aspect Unreferenced::
317 * Aspect Unreferenced_Objects::
318 * Aspect Value_Size::
321 Implementation Defined Attributes
323 * Attribute Abort_Signal::
324 * Attribute Address_Size::
325 * Attribute Asm_Input::
326 * Attribute Asm_Output::
327 * Attribute AST_Entry::
329 * Attribute Bit_Position::
330 * Attribute Compiler_Version::
331 * Attribute Code_Address::
332 * Attribute Default_Bit_Order::
333 * Attribute Descriptor_Size::
334 * Attribute Elaborated::
335 * Attribute Elab_Body::
336 * Attribute Elab_Spec::
337 * Attribute Elab_Subp_Body::
339 * Attribute Enabled::
340 * Attribute Enum_Rep::
341 * Attribute Enum_Val::
342 * Attribute Epsilon::
343 * Attribute Fixed_Value::
344 * Attribute Has_Access_Values::
345 * Attribute Has_Discriminants::
347 * Attribute Integer_Value::
348 * Attribute Invalid_Value::
350 * Attribute Library_Level::
351 * Attribute Loop_Entry::
352 * Attribute Machine_Size::
353 * Attribute Mantissa::
354 * Attribute Max_Interrupt_Priority::
355 * Attribute Max_Priority::
356 * Attribute Maximum_Alignment::
357 * Attribute Mechanism_Code::
358 * Attribute Null_Parameter::
359 * Attribute Object_Size::
360 * Attribute Passed_By_Reference::
361 * Attribute Pool_Address::
362 * Attribute Range_Length::
364 * Attribute Restriction_Set::
366 * Attribute Safe_Emax::
367 * Attribute Safe_Large::
368 * Attribute Scalar_Storage_Order::
369 * Attribute Simple_Storage_Pool::
371 * Attribute Storage_Unit::
372 * Attribute Stub_Type::
373 * Attribute System_Allocator_Alignment::
374 * Attribute Target_Name::
376 * Attribute To_Address::
377 * Attribute Type_Class::
378 * Attribute UET_Address::
379 * Attribute Unconstrained_Array::
380 * Attribute Universal_Literal_String::
381 * Attribute Unrestricted_Access::
383 * Attribute Valid_Scalars::
384 * Attribute VADS_Size::
385 * Attribute Value_Size::
386 * Attribute Wchar_T_Size::
387 * Attribute Word_Size::
389 Standard and Implementation Defined Restrictions
391 * Partition-Wide Restrictions::
392 * Program Unit Level Restrictions::
394 Partition-Wide Restrictions
396 * Immediate_Reclamation::
397 * Max_Asynchronous_Select_Nesting::
398 * Max_Entry_Queue_Length::
399 * Max_Protected_Entries::
400 * Max_Select_Alternatives::
401 * Max_Storage_At_Blocking::
404 * No_Abort_Statements::
405 * No_Access_Parameter_Allocators::
406 * No_Access_Subprograms::
408 * No_Anonymous_Allocators::
411 * No_Default_Initialization::
414 * No_Direct_Boolean_Operators::
416 * No_Dispatching_Calls::
417 * No_Dynamic_Attachment::
418 * No_Dynamic_Priorities::
419 * No_Entry_Calls_In_Elaboration_Code::
420 * No_Enumeration_Maps::
421 * No_Exception_Handlers::
422 * No_Exception_Propagation::
423 * No_Exception_Registration::
427 * No_Floating_Point::
428 * No_Implicit_Conditionals::
429 * No_Implicit_Dynamic_Code::
430 * No_Implicit_Heap_Allocations::
431 * No_Implicit_Loops::
432 * No_Initialize_Scalars::
434 * No_Local_Allocators::
435 * No_Local_Protected_Objects::
436 * No_Local_Timing_Events::
437 * No_Nested_Finalization::
438 * No_Protected_Type_Allocators::
439 * No_Protected_Types::
442 * No_Relative_Delay::
443 * No_Requeue_Statements::
444 * No_Secondary_Stack::
445 * No_Select_Statements::
446 * No_Specific_Termination_Handlers::
447 * No_Specification_of_Aspect::
448 * No_Standard_Allocators_After_Elaboration::
449 * No_Standard_Storage_Pools::
450 * No_Stream_Optimizations::
452 * No_Task_Allocators::
453 * No_Task_Attributes_Package::
454 * No_Task_Hierarchy::
455 * No_Task_Termination::
457 * No_Terminate_Alternatives::
458 * No_Unchecked_Access::
460 * Static_Priorities::
461 * Static_Storage_Size::
463 Program Unit Level Restrictions
465 * No_Elaboration_Code::
467 * No_Implementation_Aspect_Specifications::
468 * No_Implementation_Attributes::
469 * No_Implementation_Identifiers::
470 * No_Implementation_Pragmas::
471 * No_Implementation_Restrictions::
472 * No_Implementation_Units::
473 * No_Implicit_Aliasing::
474 * No_Obsolescent_Features::
475 * No_Wide_Characters::
478 The Implementation of Standard I/O
480 * Standard I/O Packages::
486 * Wide_Wide_Text_IO::
490 * Filenames encoding::
492 * Operations on C Streams::
493 * Interfacing to C Streams::
497 * Ada.Characters.Latin_9 (a-chlat9.ads)::
498 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
499 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
500 * Ada.Characters.Wide_Wide_Latin_1 (a-chzla1.ads)::
501 * Ada.Characters.Wide_Wide_Latin_9 (a-chzla9.ads)::
502 * Ada.Containers.Formal_Doubly_Linked_Lists (a-cfdlli.ads)::
503 * Ada.Containers.Formal_Hashed_Maps (a-cfhama.ads)::
504 * Ada.Containers.Formal_Hashed_Sets (a-cfhase.ads)::
505 * Ada.Containers.Formal_Ordered_Maps (a-cforma.ads)::
506 * Ada.Containers.Formal_Ordered_Sets (a-cforse.ads)::
507 * Ada.Containers.Formal_Vectors (a-cofove.ads)::
508 * Ada.Command_Line.Environment (a-colien.ads)::
509 * Ada.Command_Line.Remove (a-colire.ads)::
510 * Ada.Command_Line.Response_File (a-clrefi.ads)::
511 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
512 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
513 * Ada.Exceptions.Last_Chance_Handler (a-elchha.ads)::
514 * Ada.Exceptions.Traceback (a-exctra.ads)::
515 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
516 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
517 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
518 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
519 * Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO (a-szuzti.ads)::
520 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
521 * Ada.Text_IO.Reset_Standard_Files (a-tirsfi.ads)::
522 * Ada.Wide_Characters.Unicode (a-wichun.ads)::
523 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
524 * Ada.Wide_Text_IO.Reset_Standard_Files (a-wrstfi.ads)::
525 * Ada.Wide_Wide_Characters.Unicode (a-zchuni.ads)::
526 * Ada.Wide_Wide_Text_IO.C_Streams (a-ztcstr.ads)::
527 * Ada.Wide_Wide_Text_IO.Reset_Standard_Files (a-zrstfi.ads)::
528 * GNAT.Altivec (g-altive.ads)::
529 * GNAT.Altivec.Conversions (g-altcon.ads)::
530 * GNAT.Altivec.Vector_Operations (g-alveop.ads)::
531 * GNAT.Altivec.Vector_Types (g-alvety.ads)::
532 * GNAT.Altivec.Vector_Views (g-alvevi.ads)::
533 * GNAT.Array_Split (g-arrspl.ads)::
534 * GNAT.AWK (g-awk.ads)::
535 * GNAT.Bounded_Buffers (g-boubuf.ads)::
536 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
537 * GNAT.Bubble_Sort (g-bubsor.ads)::
538 * GNAT.Bubble_Sort_A (g-busora.ads)::
539 * GNAT.Bubble_Sort_G (g-busorg.ads)::
540 * GNAT.Byte_Order_Mark (g-byorma.ads)::
541 * GNAT.Byte_Swapping (g-bytswa.ads)::
542 * GNAT.Calendar (g-calend.ads)::
543 * GNAT.Calendar.Time_IO (g-catiio.ads)::
544 * GNAT.Case_Util (g-casuti.ads)::
545 * GNAT.CGI (g-cgi.ads)::
546 * GNAT.CGI.Cookie (g-cgicoo.ads)::
547 * GNAT.CGI.Debug (g-cgideb.ads)::
548 * GNAT.Command_Line (g-comlin.ads)::
549 * GNAT.Compiler_Version (g-comver.ads)::
550 * GNAT.Ctrl_C (g-ctrl_c.ads)::
551 * GNAT.CRC32 (g-crc32.ads)::
552 * GNAT.Current_Exception (g-curexc.ads)::
553 * GNAT.Debug_Pools (g-debpoo.ads)::
554 * GNAT.Debug_Utilities (g-debuti.ads)::
555 * GNAT.Decode_String (g-decstr.ads)::
556 * GNAT.Decode_UTF8_String (g-deutst.ads)::
557 * GNAT.Directory_Operations (g-dirope.ads)::
558 * GNAT.Directory_Operations.Iteration (g-diopit.ads)::
559 * GNAT.Dynamic_HTables (g-dynhta.ads)::
560 * GNAT.Dynamic_Tables (g-dyntab.ads)::
561 * GNAT.Encode_String (g-encstr.ads)::
562 * GNAT.Encode_UTF8_String (g-enutst.ads)::
563 * GNAT.Exception_Actions (g-excact.ads)::
564 * GNAT.Exception_Traces (g-exctra.ads)::
565 * GNAT.Exceptions (g-except.ads)::
566 * GNAT.Expect (g-expect.ads)::
567 * GNAT.Expect.TTY (g-exptty.ads)::
568 * GNAT.Float_Control (g-flocon.ads)::
569 * GNAT.Heap_Sort (g-heasor.ads)::
570 * GNAT.Heap_Sort_A (g-hesora.ads)::
571 * GNAT.Heap_Sort_G (g-hesorg.ads)::
572 * GNAT.HTable (g-htable.ads)::
573 * GNAT.IO (g-io.ads)::
574 * GNAT.IO_Aux (g-io_aux.ads)::
575 * GNAT.Lock_Files (g-locfil.ads)::
576 * GNAT.MBBS_Discrete_Random (g-mbdira.ads)::
577 * GNAT.MBBS_Float_Random (g-mbflra.ads)::
578 * GNAT.MD5 (g-md5.ads)::
579 * GNAT.Memory_Dump (g-memdum.ads)::
580 * GNAT.Most_Recent_Exception (g-moreex.ads)::
581 * GNAT.OS_Lib (g-os_lib.ads)::
582 * GNAT.Perfect_Hash_Generators (g-pehage.ads)::
583 * GNAT.Random_Numbers (g-rannum.ads)::
584 * GNAT.Regexp (g-regexp.ads)::
585 * GNAT.Registry (g-regist.ads)::
586 * GNAT.Regpat (g-regpat.ads)::
587 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
588 * GNAT.Semaphores (g-semaph.ads)::
589 * GNAT.Serial_Communications (g-sercom.ads)::
590 * GNAT.SHA1 (g-sha1.ads)::
591 * GNAT.SHA224 (g-sha224.ads)::
592 * GNAT.SHA256 (g-sha256.ads)::
593 * GNAT.SHA384 (g-sha384.ads)::
594 * GNAT.SHA512 (g-sha512.ads)::
595 * GNAT.Signals (g-signal.ads)::
596 * GNAT.Sockets (g-socket.ads)::
597 * GNAT.Source_Info (g-souinf.ads)::
598 * GNAT.Spelling_Checker (g-speche.ads)::
599 * GNAT.Spelling_Checker_Generic (g-spchge.ads)::
600 * GNAT.Spitbol.Patterns (g-spipat.ads)::
601 * GNAT.Spitbol (g-spitbo.ads)::
602 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
603 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
604 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
605 * GNAT.SSE (g-sse.ads)::
606 * GNAT.SSE.Vector_Types (g-ssvety.ads)::
607 * GNAT.Strings (g-string.ads)::
608 * GNAT.String_Split (g-strspl.ads)::
609 * GNAT.Table (g-table.ads)::
610 * GNAT.Task_Lock (g-tasloc.ads)::
611 * GNAT.Threads (g-thread.ads)::
612 * GNAT.Time_Stamp (g-timsta.ads)::
613 * GNAT.Traceback (g-traceb.ads)::
614 * GNAT.Traceback.Symbolic (g-trasym.ads)::
615 * GNAT.UTF_32 (g-utf_32.ads)::
616 * GNAT.UTF_32_Spelling_Checker (g-u3spch.ads)::
617 * GNAT.Wide_Spelling_Checker (g-wispch.ads)::
618 * GNAT.Wide_String_Split (g-wistsp.ads)::
619 * GNAT.Wide_Wide_Spelling_Checker (g-zspche.ads)::
620 * GNAT.Wide_Wide_String_Split (g-zistsp.ads)::
621 * Interfaces.C.Extensions (i-cexten.ads)::
622 * Interfaces.C.Streams (i-cstrea.ads)::
623 * Interfaces.CPP (i-cpp.ads)::
624 * Interfaces.Packed_Decimal (i-pacdec.ads)::
625 * Interfaces.VxWorks (i-vxwork.ads)::
626 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
627 * System.Address_Image (s-addima.ads)::
628 * System.Assertions (s-assert.ads)::
629 * System.Memory (s-memory.ads)::
630 * System.Multiprocessors (s-multip.ads)::
631 * System.Multiprocessors.Dispatching_Domains (s-mudido.ads)::
632 * System.Partition_Interface (s-parint.ads)::
633 * System.Pool_Global (s-pooglo.ads)::
634 * System.Pool_Local (s-pooloc.ads)::
635 * System.Restrictions (s-restri.ads)::
636 * System.Rident (s-rident.ads)::
637 * System.Strings.Stream_Ops (s-ststop.ads)::
638 * System.Task_Info (s-tasinf.ads)::
639 * System.Wch_Cnv (s-wchcnv.ads)::
640 * System.Wch_Con (s-wchcon.ads)::
644 * Text_IO Stream Pointer Positioning::
645 * Text_IO Reading and Writing Non-Regular Files::
647 * Treating Text_IO Files as Streams::
648 * Text_IO Extensions::
649 * Text_IO Facilities for Unbounded Strings::
653 * Wide_Text_IO Stream Pointer Positioning::
654 * Wide_Text_IO Reading and Writing Non-Regular Files::
658 * Wide_Wide_Text_IO Stream Pointer Positioning::
659 * Wide_Wide_Text_IO Reading and Writing Non-Regular Files::
661 Interfacing to Other Languages
664 * Interfacing to C++::
665 * Interfacing to COBOL::
666 * Interfacing to Fortran::
667 * Interfacing to non-GNAT Ada code::
669 Specialized Needs Annexes
671 Implementation of Specific Ada Features
672 * Machine Code Insertions::
673 * GNAT Implementation of Tasking::
674 * GNAT Implementation of Shared Passive Packages::
675 * Code Generation for Array Aggregates::
676 * The Size of Discriminated Records with Default Discriminants::
677 * Strict Conformance to the Ada Reference Manual::
679 Implementation of Ada 2012 Features
683 GNU Free Documentation License
690 @node About This Guide
691 @unnumbered About This Guide
694 This manual contains useful information in writing programs using the
695 @value{EDITION} compiler. It includes information on implementation dependent
696 characteristics of @value{EDITION}, including all the information required by
697 Annex M of the Ada language standard.
699 @value{EDITION} implements Ada 95, Ada 2005 and Ada 2012, and it may also be
700 invoked in Ada 83 compatibility mode.
701 By default, @value{EDITION} assumes Ada 2012,
702 but you can override with a compiler switch
703 to explicitly specify the language version.
704 (Please refer to @ref{Compiling Different Versions of Ada,,, gnat_ugn,
705 @value{EDITION} User's Guide}, for details on these switches.)
706 Throughout this manual, references to ``Ada'' without a year suffix
707 apply to all the Ada versions of the language.
709 Ada is designed to be highly portable.
710 In general, a program will have the same effect even when compiled by
711 different compilers on different platforms.
712 However, since Ada is designed to be used in a
713 wide variety of applications, it also contains a number of system
714 dependent features to be used in interfacing to the external world.
715 @cindex Implementation-dependent features
718 Note: Any program that makes use of implementation-dependent features
719 may be non-portable. You should follow good programming practice and
720 isolate and clearly document any sections of your program that make use
721 of these features in a non-portable manner.
724 For ease of exposition, ``@value{EDITION}'' will be referred to simply as
725 ``GNAT'' in the remainder of this document.
729 * What This Reference Manual Contains::
731 * Related Information::
734 @node What This Reference Manual Contains
735 @unnumberedsec What This Reference Manual Contains
738 This reference manual contains the following chapters:
742 @ref{Implementation Defined Pragmas}, lists GNAT implementation-dependent
743 pragmas, which can be used to extend and enhance the functionality of the
747 @ref{Implementation Defined Attributes}, lists GNAT
748 implementation-dependent attributes, which can be used to extend and
749 enhance the functionality of the compiler.
752 @ref{Standard and Implementation Defined Restrictions}, lists GNAT
753 implementation-dependent restrictions, which can be used to extend and
754 enhance the functionality of the compiler.
757 @ref{Implementation Advice}, provides information on generally
758 desirable behavior which are not requirements that all compilers must
759 follow since it cannot be provided on all systems, or which may be
760 undesirable on some systems.
763 @ref{Implementation Defined Characteristics}, provides a guide to
764 minimizing implementation dependent features.
767 @ref{Intrinsic Subprograms}, describes the intrinsic subprograms
768 implemented by GNAT, and how they can be imported into user
769 application programs.
772 @ref{Representation Clauses and Pragmas}, describes in detail the
773 way that GNAT represents data, and in particular the exact set
774 of representation clauses and pragmas that is accepted.
777 @ref{Standard Library Routines}, provides a listing of packages and a
778 brief description of the functionality that is provided by Ada's
779 extensive set of standard library routines as implemented by GNAT@.
782 @ref{The Implementation of Standard I/O}, details how the GNAT
783 implementation of the input-output facilities.
786 @ref{The GNAT Library}, is a catalog of packages that complement
787 the Ada predefined library.
790 @ref{Interfacing to Other Languages}, describes how programs
791 written in Ada using GNAT can be interfaced to other programming
794 @ref{Specialized Needs Annexes}, describes the GNAT implementation of all
795 of the specialized needs annexes.
798 @ref{Implementation of Specific Ada Features}, discusses issues related
799 to GNAT's implementation of machine code insertions, tasking, and several
803 @ref{Implementation of Ada 2012 Features}, describes the status of the
804 GNAT implementation of the Ada 2012 language standard.
807 @ref{Obsolescent Features} documents implementation dependent features,
808 including pragmas and attributes, which are considered obsolescent, since
809 there are other preferred ways of achieving the same results. These
810 obsolescent forms are retained for backwards compatibility.
814 @cindex Ada 95 Language Reference Manual
815 @cindex Ada 2005 Language Reference Manual
817 This reference manual assumes a basic familiarity with the Ada 95 language, as
818 described in the International Standard ANSI/ISO/IEC-8652:1995,
820 It does not require knowledge of the new features introduced by Ada 2005,
821 (officially known as ISO/IEC 8652:1995 with Technical Corrigendum 1
823 Both reference manuals are included in the GNAT documentation
827 @unnumberedsec Conventions
828 @cindex Conventions, typographical
829 @cindex Typographical conventions
832 Following are examples of the typographical and graphic conventions used
837 @code{Functions}, @code{utility program names}, @code{standard names},
844 @file{File names}, @samp{button names}, and @samp{field names}.
847 @code{Variables}, @env{environment variables}, and @var{metasyntactic
854 [optional information or parameters]
857 Examples are described by text
859 and then shown this way.
864 Commands that are entered by the user are preceded in this manual by the
865 characters @samp{$ } (dollar sign followed by space). If your system uses this
866 sequence as a prompt, then the commands will appear exactly as you see them
867 in the manual. If your system uses some other prompt, then the command will
868 appear with the @samp{$} replaced by whatever prompt character you are using.
870 @node Related Information
871 @unnumberedsec Related Information
873 See the following documents for further information on GNAT:
877 @xref{Top, @value{EDITION} User's Guide, About This Guide, gnat_ugn,
878 @value{EDITION} User's Guide}, which provides information on how to use the
879 GNAT compiler system.
882 @cite{Ada 95 Reference Manual}, which contains all reference
883 material for the Ada 95 programming language.
886 @cite{Ada 95 Annotated Reference Manual}, which is an annotated version
887 of the Ada 95 standard. The annotations describe
888 detailed aspects of the design decision, and in particular contain useful
889 sections on Ada 83 compatibility.
892 @cite{Ada 2005 Reference Manual}, which contains all reference
893 material for the Ada 2005 programming language.
896 @cite{Ada 2005 Annotated Reference Manual}, which is an annotated version
897 of the Ada 2005 standard. The annotations describe
898 detailed aspects of the design decision, and in particular contain useful
899 sections on Ada 83 and Ada 95 compatibility.
902 @cite{DEC Ada, Technical Overview and Comparison on DIGITAL Platforms},
903 which contains specific information on compatibility between GNAT and
907 @cite{DEC Ada, Language Reference Manual, part number AA-PYZAB-TK} which
908 describes in detail the pragmas and attributes provided by the DEC Ada 83
913 @node Implementation Defined Pragmas
914 @chapter Implementation Defined Pragmas
917 Ada defines a set of pragmas that can be used to supply additional
918 information to the compiler. These language defined pragmas are
919 implemented in GNAT and work as described in the Ada Reference Manual.
921 In addition, Ada allows implementations to define additional pragmas
922 whose meaning is defined by the implementation. GNAT provides a number
923 of these implementation-defined pragmas, which can be used to extend
924 and enhance the functionality of the compiler. This section of the GNAT
925 Reference Manual describes these additional pragmas.
927 Note that any program using these pragmas might not be portable to other
928 compilers (although GNAT implements this set of pragmas on all
929 platforms). Therefore if portability to other compilers is an important
930 consideration, the use of these pragmas should be minimized.
933 * Pragma Abort_Defer::
934 * Pragma Abstract_State::
941 * Pragma Allow_Integer_Address::
944 * Pragma Assert_And_Cut::
945 * Pragma Assertion_Policy::
947 * Pragma Assume_No_Invalid_Values::
948 * Pragma Attribute_Definition::
950 * Pragma C_Pass_By_Copy::
952 * Pragma Check_Float_Overflow::
953 * Pragma Check_Name::
954 * Pragma Check_Policy::
955 * Pragma CIL_Constructor::
957 * Pragma Common_Object::
958 * Pragma Compile_Time_Error::
959 * Pragma Compile_Time_Warning::
960 * Pragma Compiler_Unit::
961 * Pragma Compiler_Unit_Warning::
962 * Pragma Complete_Representation::
963 * Pragma Complex_Representation::
964 * Pragma Component_Alignment::
965 * Pragma Contract_Cases::
966 * Pragma Convention_Identifier::
968 * Pragma CPP_Constructor::
969 * Pragma CPP_Virtual::
970 * Pragma CPP_Vtable::
973 * Pragma Debug_Policy::
974 * Pragma Default_Storage_Pool::
976 * Pragma Detect_Blocking::
977 * Pragma Disable_Atomic_Synchronization::
978 * Pragma Dispatching_Domain::
979 * Pragma Elaboration_Checks::
981 * Pragma Enable_Atomic_Synchronization::
982 * Pragma Export_Exception::
983 * Pragma Export_Function::
984 * Pragma Export_Object::
985 * Pragma Export_Procedure::
986 * Pragma Export_Value::
987 * Pragma Export_Valued_Procedure::
988 * Pragma Extend_System::
989 * Pragma Extensions_Allowed::
991 * Pragma External_Name_Casing::
993 * Pragma Favor_Top_Level::
994 * Pragma Finalize_Storage_Only::
995 * Pragma Float_Representation::
998 * Pragma Implementation_Defined::
999 * Pragma Implemented::
1000 * Pragma Implicit_Packing::
1001 * Pragma Import_Exception::
1002 * Pragma Import_Function::
1003 * Pragma Import_Object::
1004 * Pragma Import_Procedure::
1005 * Pragma Import_Valued_Procedure::
1006 * Pragma Independent::
1007 * Pragma Independent_Components::
1008 * Pragma Initial_Condition::
1009 * Pragma Initialize_Scalars::
1010 * Pragma Initializes::
1011 * Pragma Inline_Always::
1012 * Pragma Inline_Generic::
1013 * Pragma Interface::
1014 * Pragma Interface_Name::
1015 * Pragma Interrupt_Handler::
1016 * Pragma Interrupt_State::
1017 * Pragma Invariant::
1018 * Pragma Java_Constructor::
1019 * Pragma Java_Interface::
1020 * Pragma Keep_Names::
1022 * Pragma Link_With::
1023 * Pragma Linker_Alias::
1024 * Pragma Linker_Constructor::
1025 * Pragma Linker_Destructor::
1026 * Pragma Linker_Section::
1027 * Pragma Long_Float::
1028 * Pragma Loop_Invariant::
1029 * Pragma Loop_Optimize::
1030 * Pragma Loop_Variant::
1031 * Pragma Machine_Attribute::
1033 * Pragma Main_Storage::
1035 * Pragma No_Inline::
1036 * Pragma No_Return::
1037 * Pragma No_Run_Time::
1038 * Pragma No_Strict_Aliasing::
1039 * Pragma Normalize_Scalars::
1040 * Pragma Obsolescent::
1041 * Pragma Optimize_Alignment::
1043 * Pragma Overflow_Mode::
1044 * Pragma Overriding_Renamings::
1045 * Pragma Partition_Elaboration_Policy::
1047 * Pragma Persistent_BSS::
1050 * Pragma Postcondition::
1051 * Pragma Post_Class::
1053 * Pragma Precondition::
1054 * Pragma Predicate::
1055 * Pragma Preelaborable_Initialization::
1056 * Pragma Preelaborate_05::
1057 * Pragma Pre_Class::
1058 * Pragma Priority_Specific_Dispatching::
1060 * Pragma Profile_Warnings::
1061 * Pragma Propagate_Exceptions::
1062 * Pragma Provide_Shift_Operators::
1063 * Pragma Psect_Object::
1066 * Pragma Pure_Function::
1067 * Pragma Ravenscar::
1068 * Pragma Refined_State::
1069 * Pragma Relative_Deadline::
1070 * Pragma Remote_Access_Type::
1071 * Pragma Restricted_Run_Time::
1072 * Pragma Restriction_Warnings::
1073 * Pragma Reviewable::
1074 * Pragma Share_Generic::
1076 * Pragma Short_Circuit_And_Or::
1077 * Pragma Short_Descriptors::
1078 * Pragma Simple_Storage_Pool_Type::
1079 * Pragma Source_File_Name::
1080 * Pragma Source_File_Name_Project::
1081 * Pragma Source_Reference::
1082 * Pragma SPARK_Mode::
1083 * Pragma Static_Elaboration_Desired::
1084 * Pragma Stream_Convert::
1085 * Pragma Style_Checks::
1088 * Pragma Suppress_All::
1089 * Pragma Suppress_Debug_Info::
1090 * Pragma Suppress_Exception_Locations::
1091 * Pragma Suppress_Initialization::
1092 * Pragma Task_Info::
1093 * Pragma Task_Name::
1094 * Pragma Task_Storage::
1095 * Pragma Test_Case::
1096 * Pragma Thread_Local_Storage::
1097 * Pragma Time_Slice::
1099 * Pragma Type_Invariant::
1100 * Pragma Type_Invariant_Class::
1101 * Pragma Unchecked_Union::
1102 * Pragma Unimplemented_Unit::
1103 * Pragma Universal_Aliasing ::
1104 * Pragma Universal_Data::
1105 * Pragma Unmodified::
1106 * Pragma Unreferenced::
1107 * Pragma Unreferenced_Objects::
1108 * Pragma Unreserve_All_Interrupts::
1109 * Pragma Unsuppress::
1110 * Pragma Use_VADS_Size::
1111 * Pragma Validity_Checks::
1113 * Pragma Warning_As_Error::
1115 * Pragma Weak_External::
1116 * Pragma Wide_Character_Encoding::
1119 @node Pragma Abort_Defer
1120 @unnumberedsec Pragma Abort_Defer
1122 @cindex Deferring aborts
1130 This pragma must appear at the start of the statement sequence of a
1131 handled sequence of statements (right after the @code{begin}). It has
1132 the effect of deferring aborts for the sequence of statements (but not
1133 for the declarations or handlers, if any, associated with this statement
1136 @node Pragma Abstract_State
1137 @unnumberedsec Pragma Abstract_State
1138 @findex Abstract_State
1140 For the description of this pragma, see SPARK 2014 Reference Manual,
1144 @unnumberedsec Pragma Ada_83
1148 @smallexample @c ada
1153 A configuration pragma that establishes Ada 83 mode for the unit to
1154 which it applies, regardless of the mode set by the command line
1155 switches. In Ada 83 mode, GNAT attempts to be as compatible with
1156 the syntax and semantics of Ada 83, as defined in the original Ada
1157 83 Reference Manual as possible. In particular, the keywords added by Ada 95
1158 and Ada 2005 are not recognized, optional package bodies are allowed,
1159 and generics may name types with unknown discriminants without using
1160 the @code{(<>)} notation. In addition, some but not all of the additional
1161 restrictions of Ada 83 are enforced.
1163 Ada 83 mode is intended for two purposes. Firstly, it allows existing
1164 Ada 83 code to be compiled and adapted to GNAT with less effort.
1165 Secondly, it aids in keeping code backwards compatible with Ada 83.
1166 However, there is no guarantee that code that is processed correctly
1167 by GNAT in Ada 83 mode will in fact compile and execute with an Ada
1168 83 compiler, since GNAT does not enforce all the additional checks
1172 @unnumberedsec Pragma Ada_95
1176 @smallexample @c ada
1181 A configuration pragma that establishes Ada 95 mode for the unit to which
1182 it applies, regardless of the mode set by the command line switches.
1183 This mode is set automatically for the @code{Ada} and @code{System}
1184 packages and their children, so you need not specify it in these
1185 contexts. This pragma is useful when writing a reusable component that
1186 itself uses Ada 95 features, but which is intended to be usable from
1187 either Ada 83 or Ada 95 programs.
1190 @unnumberedsec Pragma Ada_05
1194 @smallexample @c ada
1196 pragma Ada_05 (local_NAME);
1200 A configuration pragma that establishes Ada 2005 mode for the unit to which
1201 it applies, regardless of the mode set by the command line switches.
1202 This pragma is useful when writing a reusable component that
1203 itself uses Ada 2005 features, but which is intended to be usable from
1204 either Ada 83 or Ada 95 programs.
1206 The one argument form (which is not a configuration pragma)
1207 is used for managing the transition from
1208 Ada 95 to Ada 2005 in the run-time library. If an entity is marked
1209 as Ada_2005 only, then referencing the entity in Ada_83 or Ada_95
1210 mode will generate a warning. In addition, in Ada_83 or Ada_95
1211 mode, a preference rule is established which does not choose
1212 such an entity unless it is unambiguously specified. This avoids
1213 extra subprograms marked this way from generating ambiguities in
1214 otherwise legal pre-Ada_2005 programs. The one argument form is
1215 intended for exclusive use in the GNAT run-time library.
1217 @node Pragma Ada_2005
1218 @unnumberedsec Pragma Ada_2005
1222 @smallexample @c ada
1227 This configuration pragma is a synonym for pragma Ada_05 and has the
1228 same syntax and effect.
1231 @unnumberedsec Pragma Ada_12
1235 @smallexample @c ada
1237 pragma Ada_12 (local_NAME);
1241 A configuration pragma that establishes Ada 2012 mode for the unit to which
1242 it applies, regardless of the mode set by the command line switches.
1243 This mode is set automatically for the @code{Ada} and @code{System}
1244 packages and their children, so you need not specify it in these
1245 contexts. This pragma is useful when writing a reusable component that
1246 itself uses Ada 2012 features, but which is intended to be usable from
1247 Ada 83, Ada 95, or Ada 2005 programs.
1249 The one argument form, which is not a configuration pragma,
1250 is used for managing the transition from Ada
1251 2005 to Ada 2012 in the run-time library. If an entity is marked
1252 as Ada_201 only, then referencing the entity in any pre-Ada_2012
1253 mode will generate a warning. In addition, in any pre-Ada_2012
1254 mode, a preference rule is established which does not choose
1255 such an entity unless it is unambiguously specified. This avoids
1256 extra subprograms marked this way from generating ambiguities in
1257 otherwise legal pre-Ada_2012 programs. The one argument form is
1258 intended for exclusive use in the GNAT run-time library.
1260 @node Pragma Ada_2012
1261 @unnumberedsec Pragma Ada_2012
1265 @smallexample @c ada
1270 This configuration pragma is a synonym for pragma Ada_12 and has the
1271 same syntax and effect.
1273 @node Pragma Allow_Integer_Address
1274 @unnumberedsec Pragma Allow_Integer_Address
1275 @findex Allow_Integer_Address
1278 @smallexample @c ada
1279 pragma Allow_Integer_Address;
1283 In almost all versions of GNAT, @code{System.Address} is a private
1284 type in accordance with the implementation advice in the RM. This
1285 means that integer values,
1286 in particular integer literals, are not allowed as address values.
1287 If the configuration pragma
1288 @code{Allow_Integer_Address} is given, then integer expressions may
1289 be used anywhere a value of type @code{System.Address} is required.
1290 The effect is to introduce an implicit unchecked conversion from the
1291 integer value to type @code{System.Address}. The reverse case of using
1292 an address where an integer type is required is handled analogously.
1293 The following example compiles without errors:
1295 @smallexample @c ada
1296 pragma Allow_Integer_Address;
1297 with System; use System;
1298 package AddrAsInt is
1301 for X'Address use 16#1240#;
1302 for Y use at 16#3230#;
1303 m : Address := 16#4000#;
1304 n : constant Address := 4000;
1305 p : constant Address := Address (X + Y);
1306 v : Integer := y'Address;
1307 w : constant Integer := Integer (Y'Address);
1308 type R is new integer;
1311 for Z'Address use RR;
1316 Note that pragma @code{Allow_Integer_Address} is ignored if
1317 @code{System.Address}
1318 is not a private type. In implementations of @code{GNAT} where
1319 System.Address is a visible integer type (notably the implementations
1320 for @code{OpenVMS}), this pragma serves no purpose but is ignored
1321 rather than rejected to allow common sets of sources to be used
1322 in the two situations.
1324 @node Pragma Annotate
1325 @unnumberedsec Pragma Annotate
1329 @smallexample @c ada
1330 pragma Annotate (IDENTIFIER [,IDENTIFIER @{, ARG@}]);
1332 ARG ::= NAME | EXPRESSION
1336 This pragma is used to annotate programs. @var{identifier} identifies
1337 the type of annotation. GNAT verifies that it is an identifier, but does
1338 not otherwise analyze it. The second optional identifier is also left
1339 unanalyzed, and by convention is used to control the action of the tool to
1340 which the annotation is addressed. The remaining @var{arg} arguments
1341 can be either string literals or more generally expressions.
1342 String literals are assumed to be either of type
1343 @code{Standard.String} or else @code{Wide_String} or @code{Wide_Wide_String}
1344 depending on the character literals they contain.
1345 All other kinds of arguments are analyzed as expressions, and must be
1348 The analyzed pragma is retained in the tree, but not otherwise processed
1349 by any part of the GNAT compiler, except to generate corresponding note
1350 lines in the generated ALI file. For the format of these note lines, see
1351 the compiler source file lib-writ.ads. This pragma is intended for use by
1352 external tools, including ASIS@. The use of pragma Annotate does not
1353 affect the compilation process in any way. This pragma may be used as
1354 a configuration pragma.
1357 @unnumberedsec Pragma Assert
1361 @smallexample @c ada
1364 [, string_EXPRESSION]);
1368 The effect of this pragma depends on whether the corresponding command
1369 line switch is set to activate assertions. The pragma expands into code
1370 equivalent to the following:
1372 @smallexample @c ada
1373 if assertions-enabled then
1374 if not boolean_EXPRESSION then
1375 System.Assertions.Raise_Assert_Failure
1376 (string_EXPRESSION);
1382 The string argument, if given, is the message that will be associated
1383 with the exception occurrence if the exception is raised. If no second
1384 argument is given, the default message is @samp{@var{file}:@var{nnn}},
1385 where @var{file} is the name of the source file containing the assert,
1386 and @var{nnn} is the line number of the assert. A pragma is not a
1387 statement, so if a statement sequence contains nothing but a pragma
1388 assert, then a null statement is required in addition, as in:
1390 @smallexample @c ada
1393 pragma Assert (K > 3, "Bad value for K");
1399 Note that, as with the @code{if} statement to which it is equivalent, the
1400 type of the expression is either @code{Standard.Boolean}, or any type derived
1401 from this standard type.
1403 Assert checks can be either checked or ignored. By default they are ignored.
1404 They will be checked if either the command line switch @option{-gnata} is
1405 used, or if an @code{Assertion_Policy} or @code{Check_Policy} pragma is used
1406 to enable @code{Assert_Checks}.
1408 If assertions are ignored, then there
1409 is no run-time effect (and in particular, any side effects from the
1410 expression will not occur at run time). (The expression is still
1411 analyzed at compile time, and may cause types to be frozen if they are
1412 mentioned here for the first time).
1414 If assertions are checked, then the given expression is tested, and if
1415 it is @code{False} then @code{System.Assertions.Raise_Assert_Failure} is called
1416 which results in the raising of @code{Assert_Failure} with the given message.
1418 You should generally avoid side effects in the expression arguments of
1419 this pragma, because these side effects will turn on and off with the
1420 setting of the assertions mode, resulting in assertions that have an
1421 effect on the program. However, the expressions are analyzed for
1422 semantic correctness whether or not assertions are enabled, so turning
1423 assertions on and off cannot affect the legality of a program.
1425 Note that the implementation defined policy @code{DISABLE}, given in a
1426 pragma @code{Assertion_Policy}, can be used to suppress this semantic analysis.
1428 Note: this is a standard language-defined pragma in versions
1429 of Ada from 2005 on. In GNAT, it is implemented in all versions
1430 of Ada, and the DISABLE policy is an implementation-defined
1433 @node Pragma Assert_And_Cut
1434 @unnumberedsec Pragma Assert_And_Cut
1435 @findex Assert_And_Cut
1438 @smallexample @c ada
1439 pragma Assert_And_Cut (
1441 [, string_EXPRESSION]);
1445 The effect of this pragma is identical to that of pragma @code{Assert},
1446 except that in an @code{Assertion_Policy} pragma, the identifier
1447 @code{Assert_And_Cut} is used to control whether it is ignored or checked
1450 The intention is that this be used within a subprogram when the
1451 given test expresion sums up all the work done so far in the
1452 subprogram, so that the rest of the subprogram can be verified
1453 (informally or formally) using only the entry preconditions,
1454 and the expression in this pragma. This allows dividing up
1455 a subprogram into sections for the purposes of testing or
1456 formal verification. The pragma also serves as useful
1459 @node Pragma Assertion_Policy
1460 @unnumberedsec Pragma Assertion_Policy
1461 @findex Assertion_Policy
1464 @smallexample @c ada
1465 pragma Assertion_Policy (CHECK | DISABLE | IGNORE);
1467 pragma Assertion_Policy (
1468 ASSERTION_KIND => POLICY_IDENTIFIER
1469 @{, ASSERTION_KIND => POLICY_IDENTIFIER@});
1471 ASSERTION_KIND ::= RM_ASSERTION_KIND | ID_ASSERTION_KIND
1473 RM_ASSERTION_KIND ::= Assert |
1481 Type_Invariant'Class
1483 ID_ASSERTION_KIND ::= Assertions |
1496 Statement_Assertions
1498 POLICY_IDENTIFIER ::= Check | Disable | Ignore
1502 This is a standard Ada 2012 pragma that is available as an
1503 implementation-defined pragma in earlier versions of Ada.
1504 The assertion kinds @code{RM_ASSERTION_KIND} are those defined in
1505 the Ada standard. The assertion kinds @code{ID_ASSERTION_KIND}
1506 are implementation defined additions recognized by the GNAT compiler.
1508 The pragma applies in both cases to pragmas and aspects with matching
1509 names, e.g. @code{Pre} applies to the Pre aspect, and @code{Precondition}
1510 applies to both the @code{Precondition} pragma
1511 and the aspect @code{Precondition}. Note that the identifiers for
1512 pragmas Pre_Class and Post_Class are Pre'Class and Post'Class (not
1513 Pre_Class and Post_Class), since these pragmas are intended to be
1514 identical to the corresponding aspects).
1516 If the policy is @code{CHECK}, then assertions are enabled, i.e.
1517 the corresponding pragma or aspect is activated.
1518 If the policy is @code{IGNORE}, then assertions are ignored, i.e.
1519 the corresponding pragma or aspect is deactivated.
1520 This pragma overrides the effect of the @option{-gnata} switch on the
1523 The implementation defined policy @code{DISABLE} is like
1524 @code{IGNORE} except that it completely disables semantic
1525 checking of the corresponding pragma or aspect. This is
1526 useful when the pragma or aspect argument references subprograms
1527 in a with'ed package which is replaced by a dummy package
1528 for the final build.
1530 The implementation defined policy @code{Assertions} applies to all
1531 assertion kinds. The form with no assertion kind given implies this
1532 choice, so it applies to all assertion kinds (RM defined, and
1533 implementation defined).
1535 The implementation defined policy @code{Statement_Assertions}
1536 applies to @code{Assert}, @code{Assert_And_Cut},
1537 @code{Assume}, @code{Loop_Invariant}, and @code{Loop_Variant}.
1540 @unnumberedsec Pragma Assume
1544 @smallexample @c ada
1547 [, string_EXPRESSION]);
1551 The effect of this pragma is identical to that of pragma @code{Assert},
1552 except that in an @code{Assertion_Policy} pragma, the identifier
1553 @code{Assume} is used to control whether it is ignored or checked
1556 The intention is that this be used for assumptions about the
1557 external environment. So you cannot expect to verify formally
1558 or informally that the condition is met, this must be
1559 established by examining things outside the program itself.
1560 For example, we may have code that depends on the size of
1561 @code{Long_Long_Integer} being at least 64. So we could write:
1563 @smallexample @c ada
1564 pragma Assume (Long_Long_Integer'Size >= 64);
1568 This assumption cannot be proved from the program itself,
1569 but it acts as a useful run-time check that the assumption
1570 is met, and documents the need to ensure that it is met by
1571 reference to information outside the program.
1573 @node Pragma Assume_No_Invalid_Values
1574 @unnumberedsec Pragma Assume_No_Invalid_Values
1575 @findex Assume_No_Invalid_Values
1576 @cindex Invalid representations
1577 @cindex Invalid values
1580 @smallexample @c ada
1581 pragma Assume_No_Invalid_Values (On | Off);
1585 This is a configuration pragma that controls the assumptions made by the
1586 compiler about the occurrence of invalid representations (invalid values)
1589 The default behavior (corresponding to an Off argument for this pragma), is
1590 to assume that values may in general be invalid unless the compiler can
1591 prove they are valid. Consider the following example:
1593 @smallexample @c ada
1594 V1 : Integer range 1 .. 10;
1595 V2 : Integer range 11 .. 20;
1597 for J in V2 .. V1 loop
1603 if V1 and V2 have valid values, then the loop is known at compile
1604 time not to execute since the lower bound must be greater than the
1605 upper bound. However in default mode, no such assumption is made,
1606 and the loop may execute. If @code{Assume_No_Invalid_Values (On)}
1607 is given, the compiler will assume that any occurrence of a variable
1608 other than in an explicit @code{'Valid} test always has a valid
1609 value, and the loop above will be optimized away.
1611 The use of @code{Assume_No_Invalid_Values (On)} is appropriate if
1612 you know your code is free of uninitialized variables and other
1613 possible sources of invalid representations, and may result in
1614 more efficient code. A program that accesses an invalid representation
1615 with this pragma in effect is erroneous, so no guarantees can be made
1618 It is peculiar though permissible to use this pragma in conjunction
1619 with validity checking (-gnatVa). In such cases, accessing invalid
1620 values will generally give an exception, though formally the program
1621 is erroneous so there are no guarantees that this will always be the
1622 case, and it is recommended that these two options not be used together.
1624 @node Pragma Ast_Entry
1625 @unnumberedsec Pragma Ast_Entry
1630 @smallexample @c ada
1631 pragma AST_Entry (entry_IDENTIFIER);
1635 This pragma is implemented only in the OpenVMS implementation of GNAT@. The
1636 argument is the simple name of a single entry; at most one @code{AST_Entry}
1637 pragma is allowed for any given entry. This pragma must be used in
1638 conjunction with the @code{AST_Entry} attribute, and is only allowed after
1639 the entry declaration and in the same task type specification or single task
1640 as the entry to which it applies. This pragma specifies that the given entry
1641 may be used to handle an OpenVMS asynchronous system trap (@code{AST})
1642 resulting from an OpenVMS system service call. The pragma does not affect
1643 normal use of the entry. For further details on this pragma, see the
1644 DEC Ada Language Reference Manual, section 9.12a.
1646 @node Pragma Attribute_Definition
1647 @unnumberedsec Pragma Attribute_Definition
1648 @findex Attribute_Definition
1651 @smallexample @c ada
1652 pragma Attribute_Definition
1653 ([Attribute =>] ATTRIBUTE_DESIGNATOR,
1654 [Entity =>] LOCAL_NAME,
1655 [Expression =>] EXPRESSION | NAME);
1659 If @code{Attribute} is a known attribute name, this pragma is equivalent to
1660 the attribute definition clause:
1662 @smallexample @c ada
1663 for Entity'Attribute use Expression;
1666 If @code{Attribute} is not a recognized attribute name, the pragma is
1667 ignored, and a warning is emitted. This allows source
1668 code to be written that takes advantage of some new attribute, while remaining
1669 compilable with earlier compilers.
1671 @node Pragma C_Pass_By_Copy
1672 @unnumberedsec Pragma C_Pass_By_Copy
1673 @cindex Passing by copy
1674 @findex C_Pass_By_Copy
1677 @smallexample @c ada
1678 pragma C_Pass_By_Copy
1679 ([Max_Size =>] static_integer_EXPRESSION);
1683 Normally the default mechanism for passing C convention records to C
1684 convention subprograms is to pass them by reference, as suggested by RM
1685 B.3(69). Use the configuration pragma @code{C_Pass_By_Copy} to change
1686 this default, by requiring that record formal parameters be passed by
1687 copy if all of the following conditions are met:
1691 The size of the record type does not exceed the value specified for
1694 The record type has @code{Convention C}.
1696 The formal parameter has this record type, and the subprogram has a
1697 foreign (non-Ada) convention.
1701 If these conditions are met the argument is passed by copy, i.e.@: in a
1702 manner consistent with what C expects if the corresponding formal in the
1703 C prototype is a struct (rather than a pointer to a struct).
1705 You can also pass records by copy by specifying the convention
1706 @code{C_Pass_By_Copy} for the record type, or by using the extended
1707 @code{Import} and @code{Export} pragmas, which allow specification of
1708 passing mechanisms on a parameter by parameter basis.
1711 @unnumberedsec Pragma Check
1713 @cindex Named assertions
1717 @smallexample @c ada
1719 [Name =>] CHECK_KIND,
1720 [Check =>] Boolean_EXPRESSION
1721 [, [Message =>] string_EXPRESSION] );
1723 CHECK_KIND ::= IDENTIFIER |
1726 Type_Invariant'Class |
1731 This pragma is similar to the predefined pragma @code{Assert} except that an
1732 extra identifier argument is present. In conjunction with pragma
1733 @code{Check_Policy}, this can be used to define groups of assertions that can
1734 be independently controlled. The identifier @code{Assertion} is special, it
1735 refers to the normal set of pragma @code{Assert} statements.
1737 Checks introduced by this pragma are normally deactivated by default. They can
1738 be activated either by the command line option @option{-gnata}, which turns on
1739 all checks, or individually controlled using pragma @code{Check_Policy}.
1741 The identifiers @code{Assertions} and @code{Statement_Assertions} are not
1742 permitted as check kinds, since this would cause confusion with the use
1743 of these identifiers in @code{Assertion_Policy} and @code{Check_Policy}
1744 pragmas, where they are used to refer to sets of assertions.
1746 @node Pragma Check_Float_Overflow
1747 @unnumberedsec Pragma Check_Float_Overflow
1748 @cindex Floating-point overflow
1749 @findex Check_Float_Overflow
1752 @smallexample @c ada
1753 pragma Check_Float_Overflow;
1757 In Ada, the predefined floating-point types (@code{Short_Float},
1758 @code{Float}, @code{Long_Float}, @code{Long_Long_Float}) are
1759 defined to be @emph{unconstrained}. This means that even though each
1760 has a well-defined base range, an operation that delivers a result
1761 outside this base range is not required to raise an exception.
1762 This implementation permission accommodates the notion
1763 of infinities in IEEE floating-point, and corresponds to the
1764 efficient execution mode on most machines. GNAT will not raise
1765 overflow exceptions on these machines; instead it will generate
1766 infinities and NaN's as defined in the IEEE standard.
1768 Generating infinities, although efficient, is not always desirable.
1769 Often the preferable approach is to check for overflow, even at the
1770 (perhaps considerable) expense of run-time performance.
1771 This can be accomplished by defining your own constrained floating-point subtypes -- i.e., by supplying explicit
1772 range constraints -- and indeed such a subtype
1773 can have the same base range as its base type. For example:
1775 @smallexample @c ada
1776 subtype My_Float is Float range Float'Range;
1780 Here @code{My_Float} has the same range as
1781 @code{Float} but is constrained, so operations on
1782 @code{My_Float} values will be checked for overflow
1785 This style will achieve the desired goal, but
1786 it is often more convenient to be able to simply use
1787 the standard predefined floating-point types as long
1788 as overflow checking could be guaranteed.
1789 The @code{Check_Float_Overflow}
1790 configuration pragma achieves this effect. If a unit is compiled
1791 subject to this configuration pragma, then all operations
1792 on predefined floating-point types will be treated as
1793 though those types were constrained, and overflow checks
1794 will be generated. The @code{Constraint_Error}
1795 exception is raised if the result is out of range.
1797 This mode can also be set by use of the compiler
1798 switch @option{-gnateF}.
1800 @node Pragma Check_Name
1801 @unnumberedsec Pragma Check_Name
1802 @cindex Defining check names
1803 @cindex Check names, defining
1807 @smallexample @c ada
1808 pragma Check_Name (check_name_IDENTIFIER);
1812 This is a configuration pragma that defines a new implementation
1813 defined check name (unless IDENTIFIER matches one of the predefined
1814 check names, in which case the pragma has no effect). Check names
1815 are global to a partition, so if two or more configuration pragmas
1816 are present in a partition mentioning the same name, only one new
1817 check name is introduced.
1819 An implementation defined check name introduced with this pragma may
1820 be used in only three contexts: @code{pragma Suppress},
1821 @code{pragma Unsuppress},
1822 and as the prefix of a @code{Check_Name'Enabled} attribute reference. For
1823 any of these three cases, the check name must be visible. A check
1824 name is visible if it is in the configuration pragmas applying to
1825 the current unit, or if it appears at the start of any unit that
1826 is part of the dependency set of the current unit (e.g., units that
1827 are mentioned in @code{with} clauses).
1829 Check names introduced by this pragma are subject to control by compiler
1830 switches (in particular -gnatp) in the usual manner.
1832 @node Pragma Check_Policy
1833 @unnumberedsec Pragma Check_Policy
1834 @cindex Controlling assertions
1835 @cindex Assertions, control
1836 @cindex Check pragma control
1837 @cindex Named assertions
1841 @smallexample @c ada
1843 ([Name =>] CHECK_KIND,
1844 [Policy =>] POLICY_IDENTIFIER);
1846 pragma Check_Policy (
1847 CHECK_KIND => POLICY_IDENTIFIER
1848 @{, CHECK_KIND => POLICY_IDENTIFIER@});
1850 ASSERTION_KIND ::= RM_ASSERTION_KIND | ID_ASSERTION_KIND
1852 CHECK_KIND ::= IDENTIFIER |
1855 Type_Invariant'Class |
1858 The identifiers Name and Policy are not allowed as CHECK_KIND values. This
1859 avoids confusion between the two possible syntax forms for this pragma.
1861 POLICY_IDENTIFIER ::= ON | OFF | CHECK | DISABLE | IGNORE
1865 This pragma is used to set the checking policy for assertions (specified
1866 by aspects or pragmas), the @code{Debug} pragma, or additional checks
1867 to be checked using the @code{Check} pragma. It may appear either as
1868 a configuration pragma, or within a declarative part of package. In the
1869 latter case, it applies from the point where it appears to the end of
1870 the declarative region (like pragma @code{Suppress}).
1872 The @code{Check_Policy} pragma is similar to the
1873 predefined @code{Assertion_Policy} pragma,
1874 and if the check kind corresponds to one of the assertion kinds that
1875 are allowed by @code{Assertion_Policy}, then the effect is identical.
1877 If the first argument is Debug, then the policy applies to Debug pragmas,
1878 disabling their effect if the policy is @code{OFF}, @code{DISABLE}, or
1879 @code{IGNORE}, and allowing them to execute with normal semantics if
1880 the policy is @code{ON} or @code{CHECK}. In addition if the policy is
1881 @code{DISABLE}, then the procedure call in @code{Debug} pragmas will
1882 be totally ignored and not analyzed semantically.
1884 Finally the first argument may be some other identifier than the above
1885 possibilities, in which case it controls a set of named assertions
1886 that can be checked using pragma @code{Check}. For example, if the pragma:
1888 @smallexample @c ada
1889 pragma Check_Policy (Critical_Error, OFF);
1893 is given, then subsequent @code{Check} pragmas whose first argument is also
1894 @code{Critical_Error} will be disabled.
1896 The check policy is @code{OFF} to turn off corresponding checks, and @code{ON}
1897 to turn on corresponding checks. The default for a set of checks for which no
1898 @code{Check_Policy} is given is @code{OFF} unless the compiler switch
1899 @option{-gnata} is given, which turns on all checks by default.
1901 The check policy settings @code{CHECK} and @code{IGNORE} are recognized
1902 as synonyms for @code{ON} and @code{OFF}. These synonyms are provided for
1903 compatibility with the standard @code{Assertion_Policy} pragma. The check
1904 policy setting @code{DISABLE} causes the second argument of a corresponding
1905 @code{Check} pragma to be completely ignored and not analyzed.
1907 @node Pragma CIL_Constructor
1908 @unnumberedsec Pragma CIL_Constructor
1909 @findex CIL_Constructor
1913 @smallexample @c ada
1914 pragma CIL_Constructor ([Entity =>] function_LOCAL_NAME);
1918 This pragma is used to assert that the specified Ada function should be
1919 mapped to the .NET constructor for some Ada tagged record type.
1921 See section 4.1 of the
1922 @code{GNAT User's Guide: Supplement for the .NET Platform.}
1923 for related information.
1925 @node Pragma Comment
1926 @unnumberedsec Pragma Comment
1931 @smallexample @c ada
1932 pragma Comment (static_string_EXPRESSION);
1936 This is almost identical in effect to pragma @code{Ident}. It allows the
1937 placement of a comment into the object file and hence into the
1938 executable file if the operating system permits such usage. The
1939 difference is that @code{Comment}, unlike @code{Ident}, has
1940 no limitations on placement of the pragma (it can be placed
1941 anywhere in the main source unit), and if more than one pragma
1942 is used, all comments are retained.
1944 @node Pragma Common_Object
1945 @unnumberedsec Pragma Common_Object
1946 @findex Common_Object
1950 @smallexample @c ada
1951 pragma Common_Object (
1952 [Internal =>] LOCAL_NAME
1953 [, [External =>] EXTERNAL_SYMBOL]
1954 [, [Size =>] EXTERNAL_SYMBOL] );
1958 | static_string_EXPRESSION
1962 This pragma enables the shared use of variables stored in overlaid
1963 linker areas corresponding to the use of @code{COMMON}
1964 in Fortran. The single
1965 object @var{LOCAL_NAME} is assigned to the area designated by
1966 the @var{External} argument.
1967 You may define a record to correspond to a series
1968 of fields. The @var{Size} argument
1969 is syntax checked in GNAT, but otherwise ignored.
1971 @code{Common_Object} is not supported on all platforms. If no
1972 support is available, then the code generator will issue a message
1973 indicating that the necessary attribute for implementation of this
1974 pragma is not available.
1976 @node Pragma Compile_Time_Error
1977 @unnumberedsec Pragma Compile_Time_Error
1978 @findex Compile_Time_Error
1982 @smallexample @c ada
1983 pragma Compile_Time_Error
1984 (boolean_EXPRESSION, static_string_EXPRESSION);
1988 This pragma can be used to generate additional compile time
1990 is particularly useful in generics, where errors can be issued for
1991 specific problematic instantiations. The first parameter is a boolean
1992 expression. The pragma is effective only if the value of this expression
1993 is known at compile time, and has the value True. The set of expressions
1994 whose values are known at compile time includes all static boolean
1995 expressions, and also other values which the compiler can determine
1996 at compile time (e.g., the size of a record type set by an explicit
1997 size representation clause, or the value of a variable which was
1998 initialized to a constant and is known not to have been modified).
1999 If these conditions are met, an error message is generated using
2000 the value given as the second argument. This string value may contain
2001 embedded ASCII.LF characters to break the message into multiple lines.
2003 @node Pragma Compile_Time_Warning
2004 @unnumberedsec Pragma Compile_Time_Warning
2005 @findex Compile_Time_Warning
2009 @smallexample @c ada
2010 pragma Compile_Time_Warning
2011 (boolean_EXPRESSION, static_string_EXPRESSION);
2015 Same as pragma Compile_Time_Error, except a warning is issued instead
2016 of an error message. Note that if this pragma is used in a package that
2017 is with'ed by a client, the client will get the warning even though it
2018 is issued by a with'ed package (normally warnings in with'ed units are
2019 suppressed, but this is a special exception to that rule).
2021 One typical use is within a generic where compile time known characteristics
2022 of formal parameters are tested, and warnings given appropriately. Another use
2023 with a first parameter of True is to warn a client about use of a package,
2024 for example that it is not fully implemented.
2026 @node Pragma Compiler_Unit
2027 @unnumberedsec Pragma Compiler_Unit
2028 @findex Compiler_Unit
2032 @smallexample @c ada
2033 pragma Compiler_Unit;
2037 This pragma is obsolete. It is equivalent to Compiler_Unit_Warning. It is
2038 retained so that old versions of the GNAT run-time that use this pragma can
2039 be compiled with newer versions of the compiler.
2041 @node Pragma Compiler_Unit_Warning
2042 @unnumberedsec Pragma Compiler_Unit_Warning
2043 @findex Compiler_Unit_Warning
2047 @smallexample @c ada
2048 pragma Compiler_Unit_Warning;
2052 This pragma is intended only for internal use in the GNAT run-time library.
2053 It indicates that the unit is used as part of the compiler build. The effect
2054 is to generate warnings for the use of constructs (for example, conditional
2055 expressions) that would cause trouble when bootstrapping using an older
2056 version of GNAT. For the exact list of restrictions, see the compiler sources
2057 and references to Check_Compiler_Unit.
2059 @node Pragma Complete_Representation
2060 @unnumberedsec Pragma Complete_Representation
2061 @findex Complete_Representation
2065 @smallexample @c ada
2066 pragma Complete_Representation;
2070 This pragma must appear immediately within a record representation
2071 clause. Typical placements are before the first component clause
2072 or after the last component clause. The effect is to give an error
2073 message if any component is missing a component clause. This pragma
2074 may be used to ensure that a record representation clause is
2075 complete, and that this invariant is maintained if fields are
2076 added to the record in the future.
2078 @node Pragma Complex_Representation
2079 @unnumberedsec Pragma Complex_Representation
2080 @findex Complex_Representation
2084 @smallexample @c ada
2085 pragma Complex_Representation
2086 ([Entity =>] LOCAL_NAME);
2090 The @var{Entity} argument must be the name of a record type which has
2091 two fields of the same floating-point type. The effect of this pragma is
2092 to force gcc to use the special internal complex representation form for
2093 this record, which may be more efficient. Note that this may result in
2094 the code for this type not conforming to standard ABI (application
2095 binary interface) requirements for the handling of record types. For
2096 example, in some environments, there is a requirement for passing
2097 records by pointer, and the use of this pragma may result in passing
2098 this type in floating-point registers.
2100 @node Pragma Component_Alignment
2101 @unnumberedsec Pragma Component_Alignment
2102 @cindex Alignments of components
2103 @findex Component_Alignment
2107 @smallexample @c ada
2108 pragma Component_Alignment (
2109 [Form =>] ALIGNMENT_CHOICE
2110 [, [Name =>] type_LOCAL_NAME]);
2112 ALIGNMENT_CHOICE ::=
2120 Specifies the alignment of components in array or record types.
2121 The meaning of the @var{Form} argument is as follows:
2124 @findex Component_Size
2125 @item Component_Size
2126 Aligns scalar components and subcomponents of the array or record type
2127 on boundaries appropriate to their inherent size (naturally
2128 aligned). For example, 1-byte components are aligned on byte boundaries,
2129 2-byte integer components are aligned on 2-byte boundaries, 4-byte
2130 integer components are aligned on 4-byte boundaries and so on. These
2131 alignment rules correspond to the normal rules for C compilers on all
2132 machines except the VAX@.
2134 @findex Component_Size_4
2135 @item Component_Size_4
2136 Naturally aligns components with a size of four or fewer
2137 bytes. Components that are larger than 4 bytes are placed on the next
2140 @findex Storage_Unit
2142 Specifies that array or record components are byte aligned, i.e.@:
2143 aligned on boundaries determined by the value of the constant
2144 @code{System.Storage_Unit}.
2148 Specifies that array or record components are aligned on default
2149 boundaries, appropriate to the underlying hardware or operating system or
2150 both. For OpenVMS VAX systems, the @code{Default} choice is the same as
2151 the @code{Storage_Unit} choice (byte alignment). For all other systems,
2152 the @code{Default} choice is the same as @code{Component_Size} (natural
2157 If the @code{Name} parameter is present, @var{type_LOCAL_NAME} must
2158 refer to a local record or array type, and the specified alignment
2159 choice applies to the specified type. The use of
2160 @code{Component_Alignment} together with a pragma @code{Pack} causes the
2161 @code{Component_Alignment} pragma to be ignored. The use of
2162 @code{Component_Alignment} together with a record representation clause
2163 is only effective for fields not specified by the representation clause.
2165 If the @code{Name} parameter is absent, the pragma can be used as either
2166 a configuration pragma, in which case it applies to one or more units in
2167 accordance with the normal rules for configuration pragmas, or it can be
2168 used within a declarative part, in which case it applies to types that
2169 are declared within this declarative part, or within any nested scope
2170 within this declarative part. In either case it specifies the alignment
2171 to be applied to any record or array type which has otherwise standard
2174 If the alignment for a record or array type is not specified (using
2175 pragma @code{Pack}, pragma @code{Component_Alignment}, or a record rep
2176 clause), the GNAT uses the default alignment as described previously.
2178 @node Pragma Contract_Cases
2179 @unnumberedsec Pragma Contract_Cases
2180 @cindex Contract cases
2181 @findex Contract_Cases
2185 @smallexample @c ada
2186 pragma Contract_Cases (
2187 Condition => Consequence
2188 @{,Condition => Consequence@});
2192 The @code{Contract_Cases} pragma allows defining fine-grain specifications
2193 that can complement or replace the contract given by a precondition and a
2194 postcondition. Additionally, the @code{Contract_Cases} pragma can be used
2195 by testing and formal verification tools. The compiler checks its validity and,
2196 depending on the assertion policy at the point of declaration of the pragma,
2197 it may insert a check in the executable. For code generation, the contract
2200 @smallexample @c ada
2201 pragma Contract_Cases (
2209 @smallexample @c ada
2210 C1 : constant Boolean := Cond1; -- evaluated at subprogram entry
2211 C2 : constant Boolean := Cond2; -- evaluated at subprogram entry
2212 pragma Precondition ((C1 and not C2) or (C2 and not C1));
2213 pragma Postcondition (if C1 then Pred1);
2214 pragma Postcondition (if C2 then Pred2);
2218 The precondition ensures that one and only one of the conditions is
2219 satisfied on entry to the subprogram.
2220 The postcondition ensures that for the condition that was True on entry,
2221 the corrresponding consequence is True on exit. Other consequence expressions
2224 A precondition @code{P} and postcondition @code{Q} can also be
2225 expressed as contract cases:
2227 @smallexample @c ada
2228 pragma Contract_Cases (P => Q);
2231 The placement and visibility rules for @code{Contract_Cases} pragmas are
2232 identical to those described for preconditions and postconditions.
2234 The compiler checks that boolean expressions given in conditions and
2235 consequences are valid, where the rules for conditions are the same as
2236 the rule for an expression in @code{Precondition} and the rules for
2237 consequences are the same as the rule for an expression in
2238 @code{Postcondition}. In particular, attributes @code{'Old} and
2239 @code{'Result} can only be used within consequence expressions.
2240 The condition for the last contract case may be @code{others}, to denote
2241 any case not captured by the previous cases. The
2242 following is an example of use within a package spec:
2244 @smallexample @c ada
2245 package Math_Functions is
2247 function Sqrt (Arg : Float) return Float;
2248 pragma Contract_Cases ((Arg in 0 .. 99) => Sqrt'Result < 10,
2249 Arg >= 100 => Sqrt'Result >= 10,
2250 others => Sqrt'Result = 0);
2256 The meaning of contract cases is that only one case should apply at each
2257 call, as determined by the corresponding condition evaluating to True,
2258 and that the consequence for this case should hold when the subprogram
2261 @node Pragma Convention_Identifier
2262 @unnumberedsec Pragma Convention_Identifier
2263 @findex Convention_Identifier
2264 @cindex Conventions, synonyms
2268 @smallexample @c ada
2269 pragma Convention_Identifier (
2270 [Name =>] IDENTIFIER,
2271 [Convention =>] convention_IDENTIFIER);
2275 This pragma provides a mechanism for supplying synonyms for existing
2276 convention identifiers. The @code{Name} identifier can subsequently
2277 be used as a synonym for the given convention in other pragmas (including
2278 for example pragma @code{Import} or another @code{Convention_Identifier}
2279 pragma). As an example of the use of this, suppose you had legacy code
2280 which used Fortran77 as the identifier for Fortran. Then the pragma:
2282 @smallexample @c ada
2283 pragma Convention_Identifier (Fortran77, Fortran);
2287 would allow the use of the convention identifier @code{Fortran77} in
2288 subsequent code, avoiding the need to modify the sources. As another
2289 example, you could use this to parameterize convention requirements
2290 according to systems. Suppose you needed to use @code{Stdcall} on
2291 windows systems, and @code{C} on some other system, then you could
2292 define a convention identifier @code{Library} and use a single
2293 @code{Convention_Identifier} pragma to specify which convention
2294 would be used system-wide.
2296 @node Pragma CPP_Class
2297 @unnumberedsec Pragma CPP_Class
2299 @cindex Interfacing with C++
2303 @smallexample @c ada
2304 pragma CPP_Class ([Entity =>] LOCAL_NAME);
2308 The argument denotes an entity in the current declarative region that is
2309 declared as a record type. It indicates that the type corresponds to an
2310 externally declared C++ class type, and is to be laid out the same way
2311 that C++ would lay out the type. If the C++ class has virtual primitives
2312 then the record must be declared as a tagged record type.
2314 Types for which @code{CPP_Class} is specified do not have assignment or
2315 equality operators defined (such operations can be imported or declared
2316 as subprograms as required). Initialization is allowed only by constructor
2317 functions (see pragma @code{CPP_Constructor}). Such types are implicitly
2318 limited if not explicitly declared as limited or derived from a limited
2319 type, and an error is issued in that case.
2321 See @ref{Interfacing to C++} for related information.
2323 Note: Pragma @code{CPP_Class} is currently obsolete. It is supported
2324 for backward compatibility but its functionality is available
2325 using pragma @code{Import} with @code{Convention} = @code{CPP}.
2327 @node Pragma CPP_Constructor
2328 @unnumberedsec Pragma CPP_Constructor
2329 @cindex Interfacing with C++
2330 @findex CPP_Constructor
2334 @smallexample @c ada
2335 pragma CPP_Constructor ([Entity =>] LOCAL_NAME
2336 [, [External_Name =>] static_string_EXPRESSION ]
2337 [, [Link_Name =>] static_string_EXPRESSION ]);
2341 This pragma identifies an imported function (imported in the usual way
2342 with pragma @code{Import}) as corresponding to a C++ constructor. If
2343 @code{External_Name} and @code{Link_Name} are not specified then the
2344 @code{Entity} argument is a name that must have been previously mentioned
2345 in a pragma @code{Import} with @code{Convention} = @code{CPP}. Such name
2346 must be of one of the following forms:
2350 @code{function @var{Fname} return @var{T}}
2354 @code{function @var{Fname} return @var{T}'Class}
2357 @code{function @var{Fname} (@dots{}) return @var{T}}
2361 @code{function @var{Fname} (@dots{}) return @var{T}'Class}
2365 where @var{T} is a limited record type imported from C++ with pragma
2366 @code{Import} and @code{Convention} = @code{CPP}.
2368 The first two forms import the default constructor, used when an object
2369 of type @var{T} is created on the Ada side with no explicit constructor.
2370 The latter two forms cover all the non-default constructors of the type.
2371 See the @value{EDITION} User's Guide for details.
2373 If no constructors are imported, it is impossible to create any objects
2374 on the Ada side and the type is implicitly declared abstract.
2376 Pragma @code{CPP_Constructor} is intended primarily for automatic generation
2377 using an automatic binding generator tool (such as the @code{-fdump-ada-spec}
2379 See @ref{Interfacing to C++} for more related information.
2381 Note: The use of functions returning class-wide types for constructors is
2382 currently obsolete. They are supported for backward compatibility. The
2383 use of functions returning the type T leave the Ada sources more clear
2384 because the imported C++ constructors always return an object of type T;
2385 that is, they never return an object whose type is a descendant of type T.
2387 @node Pragma CPP_Virtual
2388 @unnumberedsec Pragma CPP_Virtual
2389 @cindex Interfacing to C++
2392 This pragma is now obsolete and, other than generating a warning if warnings
2393 on obsolescent features are enabled, is completely ignored.
2394 It is retained for compatibility
2395 purposes. It used to be required to ensure compoatibility with C++, but
2396 is no longer required for that purpose because GNAT generates
2397 the same object layout as the G++ compiler by default.
2399 See @ref{Interfacing to C++} for related information.
2401 @node Pragma CPP_Vtable
2402 @unnumberedsec Pragma CPP_Vtable
2403 @cindex Interfacing with C++
2406 This pragma is now obsolete and, other than generating a warning if warnings
2407 on obsolescent features are enabled, is completely ignored.
2408 It used to be required to ensure compatibility with C++, but
2409 is no longer required for that purpose because GNAT generates
2410 the same object layout than the G++ compiler by default.
2412 See @ref{Interfacing to C++} for related information.
2415 @unnumberedsec Pragma CPU
2420 @smallexample @c ada
2421 pragma CPU (EXPRESSION);
2425 This pragma is standard in Ada 2012, but is available in all earlier
2426 versions of Ada as an implementation-defined pragma.
2427 See Ada 2012 Reference Manual for details.
2430 @unnumberedsec Pragma Debug
2435 @smallexample @c ada
2436 pragma Debug ([CONDITION, ]PROCEDURE_CALL_WITHOUT_SEMICOLON);
2438 PROCEDURE_CALL_WITHOUT_SEMICOLON ::=
2440 | PROCEDURE_PREFIX ACTUAL_PARAMETER_PART
2444 The procedure call argument has the syntactic form of an expression, meeting
2445 the syntactic requirements for pragmas.
2447 If debug pragmas are not enabled or if the condition is present and evaluates
2448 to False, this pragma has no effect. If debug pragmas are enabled, the
2449 semantics of the pragma is exactly equivalent to the procedure call statement
2450 corresponding to the argument with a terminating semicolon. Pragmas are
2451 permitted in sequences of declarations, so you can use pragma @code{Debug} to
2452 intersperse calls to debug procedures in the middle of declarations. Debug
2453 pragmas can be enabled either by use of the command line switch @option{-gnata}
2454 or by use of the pragma @code{Check_Policy} with a first argument of
2457 @node Pragma Debug_Policy
2458 @unnumberedsec Pragma Debug_Policy
2459 @findex Debug_Policy
2463 @smallexample @c ada
2464 pragma Debug_Policy (CHECK | DISABLE | IGNORE | ON | OFF);
2468 This pragma is equivalent to a corresponding @code{Check_Policy} pragma
2469 with a first argument of @code{Debug}. It is retained for historical
2470 compatibility reasons.
2472 @node Pragma Default_Storage_Pool
2473 @unnumberedsec Pragma Default_Storage_Pool
2474 @findex Default_Storage_Pool
2478 @smallexample @c ada
2479 pragma Default_Storage_Pool (storage_pool_NAME | null);
2483 This pragma is standard in Ada 2012, but is available in all earlier
2484 versions of Ada as an implementation-defined pragma.
2485 See Ada 2012 Reference Manual for details.
2487 @node Pragma Depends
2488 @unnumberedsec Pragma Depends
2491 For the description of this pragma, see SPARK 2014 Reference Manual,
2494 @node Pragma Detect_Blocking
2495 @unnumberedsec Pragma Detect_Blocking
2496 @findex Detect_Blocking
2500 @smallexample @c ada
2501 pragma Detect_Blocking;
2505 This is a standard pragma in Ada 2005, that is available in all earlier
2506 versions of Ada as an implementation-defined pragma.
2508 This is a configuration pragma that forces the detection of potentially
2509 blocking operations within a protected operation, and to raise Program_Error
2512 @node Pragma Disable_Atomic_Synchronization
2513 @unnumberedsec Pragma Disable_Atomic_Synchronization
2514 @cindex Atomic Synchronization
2515 @findex Disable_Atomic_Synchronization
2519 @smallexample @c ada
2520 pragma Disable_Atomic_Synchronization [(Entity)];
2524 Ada requires that accesses (reads or writes) of an atomic variable be
2525 regarded as synchronization points in the case of multiple tasks.
2526 Particularly in the case of multi-processors this may require special
2527 handling, e.g. the generation of memory barriers. This capability may
2528 be turned off using this pragma in cases where it is known not to be
2531 The placement and scope rules for this pragma are the same as those
2532 for @code{pragma Suppress}. In particular it can be used as a
2533 configuration pragma, or in a declaration sequence where it applies
2534 till the end of the scope. If an @code{Entity} argument is present,
2535 the action applies only to that entity.
2537 @node Pragma Dispatching_Domain
2538 @unnumberedsec Pragma Dispatching_Domain
2539 @findex Dispatching_Domain
2543 @smallexample @c ada
2544 pragma Dispatching_Domain (EXPRESSION);
2548 This pragma is standard in Ada 2012, but is available in all earlier
2549 versions of Ada as an implementation-defined pragma.
2550 See Ada 2012 Reference Manual for details.
2552 @node Pragma Elaboration_Checks
2553 @unnumberedsec Pragma Elaboration_Checks
2554 @cindex Elaboration control
2555 @findex Elaboration_Checks
2559 @smallexample @c ada
2560 pragma Elaboration_Checks (Dynamic | Static);
2564 This is a configuration pragma that provides control over the
2565 elaboration model used by the compilation affected by the
2566 pragma. If the parameter is @code{Dynamic},
2567 then the dynamic elaboration
2568 model described in the Ada Reference Manual is used, as though
2569 the @option{-gnatE} switch had been specified on the command
2570 line. If the parameter is @code{Static}, then the default GNAT static
2571 model is used. This configuration pragma overrides the setting
2572 of the command line. For full details on the elaboration models
2573 used by the GNAT compiler, see @ref{Elaboration Order Handling in GNAT,,,
2574 gnat_ugn, @value{EDITION} User's Guide}.
2576 @node Pragma Eliminate
2577 @unnumberedsec Pragma Eliminate
2578 @cindex Elimination of unused subprograms
2583 @smallexample @c ada
2584 pragma Eliminate ([Entity =>] DEFINING_DESIGNATOR,
2585 [Source_Location =>] STRING_LITERAL);
2589 The string literal given for the source location is a string which
2590 specifies the line number of the occurrence of the entity, using
2591 the syntax for SOURCE_TRACE given below:
2593 @smallexample @c ada
2594 SOURCE_TRACE ::= SOURCE_REFERENCE [LBRACKET SOURCE_TRACE RBRACKET]
2599 SOURCE_REFERENCE ::= FILE_NAME : LINE_NUMBER
2601 LINE_NUMBER ::= DIGIT @{DIGIT@}
2605 Spaces around the colon in a @code{Source_Reference} are optional.
2607 The @code{DEFINING_DESIGNATOR} matches the defining designator used in an
2608 explicit subprogram declaration, where the @code{entity} name in this
2609 designator appears on the source line specified by the source location.
2611 The source trace that is given as the @code{Source_Location} shall obey the
2612 following rules. The @code{FILE_NAME} is the short name (with no directory
2613 information) of an Ada source file, given using exactly the required syntax
2614 for the underlying file system (e.g. case is important if the underlying
2615 operating system is case sensitive). @code{LINE_NUMBER} gives the line
2616 number of the occurrence of the @code{entity}
2617 as a decimal literal without an exponent or point. If an @code{entity} is not
2618 declared in a generic instantiation (this includes generic subprogram
2619 instances), the source trace includes only one source reference. If an entity
2620 is declared inside a generic instantiation, its source trace (when parsing
2621 from left to right) starts with the source location of the declaration of the
2622 entity in the generic unit and ends with the source location of the
2623 instantiation (it is given in square brackets). This approach is recursively
2624 used in case of nested instantiations: the rightmost (nested most deeply in
2625 square brackets) element of the source trace is the location of the outermost
2626 instantiation, the next to left element is the location of the next (first
2627 nested) instantiation in the code of the corresponding generic unit, and so
2628 on, and the leftmost element (that is out of any square brackets) is the
2629 location of the declaration of the entity to eliminate in a generic unit.
2631 Note that the @code{Source_Location} argument specifies which of a set of
2632 similarly named entities is being eliminated, dealing both with overloading,
2633 and also appearance of the same entity name in different scopes.
2635 This pragma indicates that the given entity is not used in the program to be
2636 compiled and built. The effect of the pragma is to allow the compiler to
2637 eliminate the code or data associated with the named entity. Any reference to
2638 an eliminated entity causes a compile-time or link-time error.
2640 The intention of pragma @code{Eliminate} is to allow a program to be compiled
2641 in a system-independent manner, with unused entities eliminated, without
2642 needing to modify the source text. Normally the required set of
2643 @code{Eliminate} pragmas is constructed automatically using the gnatelim tool.
2645 Any source file change that removes, splits, or
2646 adds lines may make the set of Eliminate pragmas invalid because their
2647 @code{Source_Location} argument values may get out of date.
2649 Pragma @code{Eliminate} may be used where the referenced entity is a dispatching
2650 operation. In this case all the subprograms to which the given operation can
2651 dispatch are considered to be unused (are never called as a result of a direct
2652 or a dispatching call).
2654 @node Pragma Enable_Atomic_Synchronization
2655 @unnumberedsec Pragma Enable_Atomic_Synchronization
2656 @cindex Atomic Synchronization
2657 @findex Enable_Atomic_Synchronization
2661 @smallexample @c ada
2662 pragma Enable_Atomic_Synchronization [(Entity)];
2666 Ada requires that accesses (reads or writes) of an atomic variable be
2667 regarded as synchronization points in the case of multiple tasks.
2668 Particularly in the case of multi-processors this may require special
2669 handling, e.g. the generation of memory barriers. This synchronization
2670 is performed by default, but can be turned off using
2671 @code{pragma Disable_Atomic_Synchronization}. The
2672 @code{Enable_Atomic_Synchronization} pragma can be used to turn
2675 The placement and scope rules for this pragma are the same as those
2676 for @code{pragma Unsuppress}. In particular it can be used as a
2677 configuration pragma, or in a declaration sequence where it applies
2678 till the end of the scope. If an @code{Entity} argument is present,
2679 the action applies only to that entity.
2681 @node Pragma Export_Exception
2682 @unnumberedsec Pragma Export_Exception
2684 @findex Export_Exception
2688 @smallexample @c ada
2689 pragma Export_Exception (
2690 [Internal =>] LOCAL_NAME
2691 [, [External =>] EXTERNAL_SYMBOL]
2692 [, [Form =>] Ada | VMS]
2693 [, [Code =>] static_integer_EXPRESSION]);
2697 | static_string_EXPRESSION
2701 This pragma is implemented only in the OpenVMS implementation of GNAT@. It
2702 causes the specified exception to be propagated outside of the Ada program,
2703 so that it can be handled by programs written in other OpenVMS languages.
2704 This pragma establishes an external name for an Ada exception and makes the
2705 name available to the OpenVMS Linker as a global symbol. For further details
2706 on this pragma, see the
2707 DEC Ada Language Reference Manual, section 13.9a3.2.
2709 @node Pragma Export_Function
2710 @unnumberedsec Pragma Export_Function
2711 @cindex Argument passing mechanisms
2712 @findex Export_Function
2717 @smallexample @c ada
2718 pragma Export_Function (
2719 [Internal =>] LOCAL_NAME
2720 [, [External =>] EXTERNAL_SYMBOL]
2721 [, [Parameter_Types =>] PARAMETER_TYPES]
2722 [, [Result_Type =>] result_SUBTYPE_MARK]
2723 [, [Mechanism =>] MECHANISM]
2724 [, [Result_Mechanism =>] MECHANISM_NAME]);
2728 | static_string_EXPRESSION
2733 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2737 | subtype_Name ' Access
2741 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2743 MECHANISM_ASSOCIATION ::=
2744 [formal_parameter_NAME =>] MECHANISM_NAME
2749 | Descriptor [([Class =>] CLASS_NAME)]
2750 | Short_Descriptor [([Class =>] CLASS_NAME)]
2752 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a
2756 Use this pragma to make a function externally callable and optionally
2757 provide information on mechanisms to be used for passing parameter and
2758 result values. We recommend, for the purposes of improving portability,
2759 this pragma always be used in conjunction with a separate pragma
2760 @code{Export}, which must precede the pragma @code{Export_Function}.
2761 GNAT does not require a separate pragma @code{Export}, but if none is
2762 present, @code{Convention Ada} is assumed, which is usually
2763 not what is wanted, so it is usually appropriate to use this
2764 pragma in conjunction with a @code{Export} or @code{Convention}
2765 pragma that specifies the desired foreign convention.
2766 Pragma @code{Export_Function}
2767 (and @code{Export}, if present) must appear in the same declarative
2768 region as the function to which they apply.
2770 @var{internal_name} must uniquely designate the function to which the
2771 pragma applies. If more than one function name exists of this name in
2772 the declarative part you must use the @code{Parameter_Types} and
2773 @code{Result_Type} parameters is mandatory to achieve the required
2774 unique designation. @var{subtype_mark}s in these parameters must
2775 exactly match the subtypes in the corresponding function specification,
2776 using positional notation to match parameters with subtype marks.
2777 The form with an @code{'Access} attribute can be used to match an
2778 anonymous access parameter.
2781 @cindex Passing by descriptor
2782 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
2783 The default behavior for Export_Function is to accept either 64bit or
2784 32bit descriptors unless short_descriptor is specified, then only 32bit
2785 descriptors are accepted.
2787 @cindex Suppressing external name
2788 Special treatment is given if the EXTERNAL is an explicit null
2789 string or a static string expressions that evaluates to the null
2790 string. In this case, no external name is generated. This form
2791 still allows the specification of parameter mechanisms.
2793 @node Pragma Export_Object
2794 @unnumberedsec Pragma Export_Object
2795 @findex Export_Object
2799 @smallexample @c ada
2800 pragma Export_Object
2801 [Internal =>] LOCAL_NAME
2802 [, [External =>] EXTERNAL_SYMBOL]
2803 [, [Size =>] EXTERNAL_SYMBOL]
2807 | static_string_EXPRESSION
2811 This pragma designates an object as exported, and apart from the
2812 extended rules for external symbols, is identical in effect to the use of
2813 the normal @code{Export} pragma applied to an object. You may use a
2814 separate Export pragma (and you probably should from the point of view
2815 of portability), but it is not required. @var{Size} is syntax checked,
2816 but otherwise ignored by GNAT@.
2818 @node Pragma Export_Procedure
2819 @unnumberedsec Pragma Export_Procedure
2820 @findex Export_Procedure
2824 @smallexample @c ada
2825 pragma Export_Procedure (
2826 [Internal =>] LOCAL_NAME
2827 [, [External =>] EXTERNAL_SYMBOL]
2828 [, [Parameter_Types =>] PARAMETER_TYPES]
2829 [, [Mechanism =>] MECHANISM]);
2833 | static_string_EXPRESSION
2838 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2842 | subtype_Name ' Access
2846 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2848 MECHANISM_ASSOCIATION ::=
2849 [formal_parameter_NAME =>] MECHANISM_NAME
2854 | Descriptor [([Class =>] CLASS_NAME)]
2855 | Short_Descriptor [([Class =>] CLASS_NAME)]
2857 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a
2861 This pragma is identical to @code{Export_Function} except that it
2862 applies to a procedure rather than a function and the parameters
2863 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
2864 GNAT does not require a separate pragma @code{Export}, but if none is
2865 present, @code{Convention Ada} is assumed, which is usually
2866 not what is wanted, so it is usually appropriate to use this
2867 pragma in conjunction with a @code{Export} or @code{Convention}
2868 pragma that specifies the desired foreign convention.
2871 @cindex Passing by descriptor
2872 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
2873 The default behavior for Export_Procedure is to accept either 64bit or
2874 32bit descriptors unless short_descriptor is specified, then only 32bit
2875 descriptors are accepted.
2877 @cindex Suppressing external name
2878 Special treatment is given if the EXTERNAL is an explicit null
2879 string or a static string expressions that evaluates to the null
2880 string. In this case, no external name is generated. This form
2881 still allows the specification of parameter mechanisms.
2883 @node Pragma Export_Value
2884 @unnumberedsec Pragma Export_Value
2885 @findex Export_Value
2889 @smallexample @c ada
2890 pragma Export_Value (
2891 [Value =>] static_integer_EXPRESSION,
2892 [Link_Name =>] static_string_EXPRESSION);
2896 This pragma serves to export a static integer value for external use.
2897 The first argument specifies the value to be exported. The Link_Name
2898 argument specifies the symbolic name to be associated with the integer
2899 value. This pragma is useful for defining a named static value in Ada
2900 that can be referenced in assembly language units to be linked with
2901 the application. This pragma is currently supported only for the
2902 AAMP target and is ignored for other targets.
2904 @node Pragma Export_Valued_Procedure
2905 @unnumberedsec Pragma Export_Valued_Procedure
2906 @findex Export_Valued_Procedure
2910 @smallexample @c ada
2911 pragma Export_Valued_Procedure (
2912 [Internal =>] LOCAL_NAME
2913 [, [External =>] EXTERNAL_SYMBOL]
2914 [, [Parameter_Types =>] PARAMETER_TYPES]
2915 [, [Mechanism =>] MECHANISM]);
2919 | static_string_EXPRESSION
2924 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
2928 | subtype_Name ' Access
2932 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
2934 MECHANISM_ASSOCIATION ::=
2935 [formal_parameter_NAME =>] MECHANISM_NAME
2940 | Descriptor [([Class =>] CLASS_NAME)]
2941 | Short_Descriptor [([Class =>] CLASS_NAME)]
2943 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a
2947 This pragma is identical to @code{Export_Procedure} except that the
2948 first parameter of @var{LOCAL_NAME}, which must be present, must be of
2949 mode @code{OUT}, and externally the subprogram is treated as a function
2950 with this parameter as the result of the function. GNAT provides for
2951 this capability to allow the use of @code{OUT} and @code{IN OUT}
2952 parameters in interfacing to external functions (which are not permitted
2954 GNAT does not require a separate pragma @code{Export}, but if none is
2955 present, @code{Convention Ada} is assumed, which is almost certainly
2956 not what is wanted since the whole point of this pragma is to interface
2957 with foreign language functions, so it is usually appropriate to use this
2958 pragma in conjunction with a @code{Export} or @code{Convention}
2959 pragma that specifies the desired foreign convention.
2962 @cindex Passing by descriptor
2963 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
2964 The default behavior for Export_Valued_Procedure is to accept either 64bit or
2965 32bit descriptors unless short_descriptor is specified, then only 32bit
2966 descriptors are accepted.
2968 @cindex Suppressing external name
2969 Special treatment is given if the EXTERNAL is an explicit null
2970 string or a static string expressions that evaluates to the null
2971 string. In this case, no external name is generated. This form
2972 still allows the specification of parameter mechanisms.
2974 @node Pragma Extend_System
2975 @unnumberedsec Pragma Extend_System
2976 @cindex @code{system}, extending
2978 @findex Extend_System
2982 @smallexample @c ada
2983 pragma Extend_System ([Name =>] IDENTIFIER);
2987 This pragma is used to provide backwards compatibility with other
2988 implementations that extend the facilities of package @code{System}. In
2989 GNAT, @code{System} contains only the definitions that are present in
2990 the Ada RM@. However, other implementations, notably the DEC Ada 83
2991 implementation, provide many extensions to package @code{System}.
2993 For each such implementation accommodated by this pragma, GNAT provides a
2994 package @code{Aux_@var{xxx}}, e.g.@: @code{Aux_DEC} for the DEC Ada 83
2995 implementation, which provides the required additional definitions. You
2996 can use this package in two ways. You can @code{with} it in the normal
2997 way and access entities either by selection or using a @code{use}
2998 clause. In this case no special processing is required.
3000 However, if existing code contains references such as
3001 @code{System.@var{xxx}} where @var{xxx} is an entity in the extended
3002 definitions provided in package @code{System}, you may use this pragma
3003 to extend visibility in @code{System} in a non-standard way that
3004 provides greater compatibility with the existing code. Pragma
3005 @code{Extend_System} is a configuration pragma whose single argument is
3006 the name of the package containing the extended definition
3007 (e.g.@: @code{Aux_DEC} for the DEC Ada case). A unit compiled under
3008 control of this pragma will be processed using special visibility
3009 processing that looks in package @code{System.Aux_@var{xxx}} where
3010 @code{Aux_@var{xxx}} is the pragma argument for any entity referenced in
3011 package @code{System}, but not found in package @code{System}.
3013 You can use this pragma either to access a predefined @code{System}
3014 extension supplied with the compiler, for example @code{Aux_DEC} or
3015 you can construct your own extension unit following the above
3016 definition. Note that such a package is a child of @code{System}
3017 and thus is considered part of the implementation.
3018 To compile it you will have to use the @option{-gnatg} switch,
3019 or the @option{/GNAT_INTERNAL} qualifier on OpenVMS,
3020 for compiling System units, as explained in the
3021 @value{EDITION} User's Guide.
3023 @node Pragma Extensions_Allowed
3024 @unnumberedsec Pragma Extensions_Allowed
3025 @cindex Ada Extensions
3026 @cindex GNAT Extensions
3027 @findex Extensions_Allowed
3031 @smallexample @c ada
3032 pragma Extensions_Allowed (On | Off);
3036 This configuration pragma enables or disables the implementation
3037 extension mode (the use of Off as a parameter cancels the effect
3038 of the @option{-gnatX} command switch).
3040 In extension mode, the latest version of the Ada language is
3041 implemented (currently Ada 2012), and in addition a small number
3042 of GNAT specific extensions are recognized as follows:
3045 @item Constrained attribute for generic objects
3046 The @code{Constrained} attribute is permitted for objects of
3047 generic types. The result indicates if the corresponding actual
3052 @node Pragma External
3053 @unnumberedsec Pragma External
3058 @smallexample @c ada
3060 [ Convention =>] convention_IDENTIFIER,
3061 [ Entity =>] LOCAL_NAME
3062 [, [External_Name =>] static_string_EXPRESSION ]
3063 [, [Link_Name =>] static_string_EXPRESSION ]);
3067 This pragma is identical in syntax and semantics to pragma
3068 @code{Export} as defined in the Ada Reference Manual. It is
3069 provided for compatibility with some Ada 83 compilers that
3070 used this pragma for exactly the same purposes as pragma
3071 @code{Export} before the latter was standardized.
3073 @node Pragma External_Name_Casing
3074 @unnumberedsec Pragma External_Name_Casing
3075 @cindex Dec Ada 83 casing compatibility
3076 @cindex External Names, casing
3077 @cindex Casing of External names
3078 @findex External_Name_Casing
3082 @smallexample @c ada
3083 pragma External_Name_Casing (
3084 Uppercase | Lowercase
3085 [, Uppercase | Lowercase | As_Is]);
3089 This pragma provides control over the casing of external names associated
3090 with Import and Export pragmas. There are two cases to consider:
3093 @item Implicit external names
3094 Implicit external names are derived from identifiers. The most common case
3095 arises when a standard Ada Import or Export pragma is used with only two
3098 @smallexample @c ada
3099 pragma Import (C, C_Routine);
3103 Since Ada is a case-insensitive language, the spelling of the identifier in
3104 the Ada source program does not provide any information on the desired
3105 casing of the external name, and so a convention is needed. In GNAT the
3106 default treatment is that such names are converted to all lower case
3107 letters. This corresponds to the normal C style in many environments.
3108 The first argument of pragma @code{External_Name_Casing} can be used to
3109 control this treatment. If @code{Uppercase} is specified, then the name
3110 will be forced to all uppercase letters. If @code{Lowercase} is specified,
3111 then the normal default of all lower case letters will be used.
3113 This same implicit treatment is also used in the case of extended DEC Ada 83
3114 compatible Import and Export pragmas where an external name is explicitly
3115 specified using an identifier rather than a string.
3117 @item Explicit external names
3118 Explicit external names are given as string literals. The most common case
3119 arises when a standard Ada Import or Export pragma is used with three
3122 @smallexample @c ada
3123 pragma Import (C, C_Routine, "C_routine");
3127 In this case, the string literal normally provides the exact casing required
3128 for the external name. The second argument of pragma
3129 @code{External_Name_Casing} may be used to modify this behavior.
3130 If @code{Uppercase} is specified, then the name
3131 will be forced to all uppercase letters. If @code{Lowercase} is specified,
3132 then the name will be forced to all lowercase letters. A specification of
3133 @code{As_Is} provides the normal default behavior in which the casing is
3134 taken from the string provided.
3138 This pragma may appear anywhere that a pragma is valid. In particular, it
3139 can be used as a configuration pragma in the @file{gnat.adc} file, in which
3140 case it applies to all subsequent compilations, or it can be used as a program
3141 unit pragma, in which case it only applies to the current unit, or it can
3142 be used more locally to control individual Import/Export pragmas.
3144 It is primarily intended for use with OpenVMS systems, where many
3145 compilers convert all symbols to upper case by default. For interfacing to
3146 such compilers (e.g.@: the DEC C compiler), it may be convenient to use
3149 @smallexample @c ada
3150 pragma External_Name_Casing (Uppercase, Uppercase);
3154 to enforce the upper casing of all external symbols.
3156 @node Pragma Fast_Math
3157 @unnumberedsec Pragma Fast_Math
3162 @smallexample @c ada
3167 This is a configuration pragma which activates a mode in which speed is
3168 considered more important for floating-point operations than absolutely
3169 accurate adherence to the requirements of the standard. Currently the
3170 following operations are affected:
3173 @item Complex Multiplication
3174 The normal simple formula for complex multiplication can result in intermediate
3175 overflows for numbers near the end of the range. The Ada standard requires that
3176 this situation be detected and corrected by scaling, but in Fast_Math mode such
3177 cases will simply result in overflow. Note that to take advantage of this you
3178 must instantiate your own version of @code{Ada.Numerics.Generic_Complex_Types}
3179 under control of the pragma, rather than use the preinstantiated versions.
3182 @node Pragma Favor_Top_Level
3183 @unnumberedsec Pragma Favor_Top_Level
3184 @findex Favor_Top_Level
3188 @smallexample @c ada
3189 pragma Favor_Top_Level (type_NAME);
3193 The named type must be an access-to-subprogram type. This pragma is an
3194 efficiency hint to the compiler, regarding the use of 'Access or
3195 'Unrestricted_Access on nested (non-library-level) subprograms. The
3196 pragma means that nested subprograms are not used with this type, or
3197 are rare, so that the generated code should be efficient in the
3198 top-level case. When this pragma is used, dynamically generated
3199 trampolines may be used on some targets for nested subprograms.
3200 See also the No_Implicit_Dynamic_Code restriction.
3202 @node Pragma Finalize_Storage_Only
3203 @unnumberedsec Pragma Finalize_Storage_Only
3204 @findex Finalize_Storage_Only
3208 @smallexample @c ada
3209 pragma Finalize_Storage_Only (first_subtype_LOCAL_NAME);
3213 This pragma allows the compiler not to emit a Finalize call for objects
3214 defined at the library level. This is mostly useful for types where
3215 finalization is only used to deal with storage reclamation since in most
3216 environments it is not necessary to reclaim memory just before terminating
3217 execution, hence the name.
3219 @node Pragma Float_Representation
3220 @unnumberedsec Pragma Float_Representation
3222 @findex Float_Representation
3226 @smallexample @c ada
3227 pragma Float_Representation (FLOAT_REP[, float_type_LOCAL_NAME]);
3229 FLOAT_REP ::= VAX_Float | IEEE_Float
3233 In the one argument form, this pragma is a configuration pragma which
3234 allows control over the internal representation chosen for the predefined
3235 floating point types declared in the packages @code{Standard} and
3236 @code{System}. On all systems other than OpenVMS, the argument must
3237 be @code{IEEE_Float} and the pragma has no effect. On OpenVMS, the
3238 argument may be @code{VAX_Float} to specify the use of the VAX float
3239 format for the floating-point types in Standard. This requires that
3240 the standard runtime libraries be recompiled.
3242 The two argument form specifies the representation to be used for
3243 the specified floating-point type. On all systems other than OpenVMS,
3245 be @code{IEEE_Float} to specify the use of IEEE format, as follows:
3249 For a digits value of 6, 32-bit IEEE short format will be used.
3251 For a digits value of 15, 64-bit IEEE long format will be used.
3253 No other value of digits is permitted.
3257 argument may be @code{VAX_Float} to specify the use of the VAX float
3262 For digits values up to 6, F float format will be used.
3264 For digits values from 7 to 9, D float format will be used.
3266 For digits values from 10 to 15, G float format will be used.
3268 Digits values above 15 are not allowed.
3272 @unnumberedsec Pragma Global
3275 For the description of this pragma, see SPARK 2014 Reference Manual,
3279 @unnumberedsec Pragma Ident
3284 @smallexample @c ada
3285 pragma Ident (static_string_EXPRESSION);
3289 This pragma provides a string identification in the generated object file,
3290 if the system supports the concept of this kind of identification string.
3291 This pragma is allowed only in the outermost declarative part or
3292 declarative items of a compilation unit. If more than one @code{Ident}
3293 pragma is given, only the last one processed is effective.
3295 On OpenVMS systems, the effect of the pragma is identical to the effect of
3296 the DEC Ada 83 pragma of the same name. Note that in DEC Ada 83, the
3297 maximum allowed length is 31 characters, so if it is important to
3298 maintain compatibility with this compiler, you should obey this length
3301 @node Pragma Implementation_Defined
3302 @unnumberedsec Pragma Implementation_Defined
3303 @findex Implementation_Defined
3307 @smallexample @c ada
3308 pragma Implementation_Defined (local_NAME);
3312 This pragma marks a previously declared entioty as implementation-defined.
3313 For an overloaded entity, applies to the most recent homonym.
3315 @smallexample @c ada
3316 pragma Implementation_Defined;
3320 The form with no arguments appears anywhere within a scope, most
3321 typically a package spec, and indicates that all entities that are
3322 defined within the package spec are Implementation_Defined.
3324 This pragma is used within the GNAT runtime library to identify
3325 implementation-defined entities introduced in language-defined units,
3326 for the purpose of implementing the No_Implementation_Identifiers
3329 @node Pragma Implemented
3330 @unnumberedsec Pragma Implemented
3335 @smallexample @c ada
3336 pragma Implemented (procedure_LOCAL_NAME, implementation_kind);
3338 implementation_kind ::= By_Entry | By_Protected_Procedure | By_Any
3342 This is an Ada 2012 representation pragma which applies to protected, task
3343 and synchronized interface primitives. The use of pragma Implemented provides
3344 a way to impose a static requirement on the overriding operation by adhering
3345 to one of the three implementation kinds: entry, protected procedure or any of
3346 the above. This pragma is available in all earlier versions of Ada as an
3347 implementation-defined pragma.
3349 @smallexample @c ada
3350 type Synch_Iface is synchronized interface;
3351 procedure Prim_Op (Obj : in out Iface) is abstract;
3352 pragma Implemented (Prim_Op, By_Protected_Procedure);
3354 protected type Prot_1 is new Synch_Iface with
3355 procedure Prim_Op; -- Legal
3358 protected type Prot_2 is new Synch_Iface with
3359 entry Prim_Op; -- Illegal
3362 task type Task_Typ is new Synch_Iface with
3363 entry Prim_Op; -- Illegal
3368 When applied to the procedure_or_entry_NAME of a requeue statement, pragma
3369 Implemented determines the runtime behavior of the requeue. Implementation kind
3370 By_Entry guarantees that the action of requeueing will proceed from an entry to
3371 another entry. Implementation kind By_Protected_Procedure transforms the
3372 requeue into a dispatching call, thus eliminating the chance of blocking. Kind
3373 By_Any shares the behavior of By_Entry and By_Protected_Procedure depending on
3374 the target's overriding subprogram kind.
3376 @node Pragma Implicit_Packing
3377 @unnumberedsec Pragma Implicit_Packing
3378 @findex Implicit_Packing
3379 @cindex Rational Profile
3383 @smallexample @c ada
3384 pragma Implicit_Packing;
3388 This is a configuration pragma that requests implicit packing for packed
3389 arrays for which a size clause is given but no explicit pragma Pack or
3390 specification of Component_Size is present. It also applies to records
3391 where no record representation clause is present. Consider this example:
3393 @smallexample @c ada
3394 type R is array (0 .. 7) of Boolean;
3399 In accordance with the recommendation in the RM (RM 13.3(53)), a Size clause
3400 does not change the layout of a composite object. So the Size clause in the
3401 above example is normally rejected, since the default layout of the array uses
3402 8-bit components, and thus the array requires a minimum of 64 bits.
3404 If this declaration is compiled in a region of code covered by an occurrence
3405 of the configuration pragma Implicit_Packing, then the Size clause in this
3406 and similar examples will cause implicit packing and thus be accepted. For
3407 this implicit packing to occur, the type in question must be an array of small
3408 components whose size is known at compile time, and the Size clause must
3409 specify the exact size that corresponds to the number of elements in the array
3410 multiplied by the size in bits of the component type (both single and
3411 multi-dimensioned arrays can be controlled with this pragma).
3413 @cindex Array packing
3415 Similarly, the following example shows the use in the record case
3417 @smallexample @c ada
3419 a, b, c, d, e, f, g, h : boolean;
3426 Without a pragma Pack, each Boolean field requires 8 bits, so the
3427 minimum size is 72 bits, but with a pragma Pack, 16 bits would be
3428 sufficient. The use of pragma Implicit_Packing allows this record
3429 declaration to compile without an explicit pragma Pack.
3430 @node Pragma Import_Exception
3431 @unnumberedsec Pragma Import_Exception
3433 @findex Import_Exception
3437 @smallexample @c ada
3438 pragma Import_Exception (
3439 [Internal =>] LOCAL_NAME
3440 [, [External =>] EXTERNAL_SYMBOL]
3441 [, [Form =>] Ada | VMS]
3442 [, [Code =>] static_integer_EXPRESSION]);
3446 | static_string_EXPRESSION
3450 This pragma is implemented only in the OpenVMS implementation of GNAT@.
3451 It allows OpenVMS conditions (for example, from OpenVMS system services or
3452 other OpenVMS languages) to be propagated to Ada programs as Ada exceptions.
3453 The pragma specifies that the exception associated with an exception
3454 declaration in an Ada program be defined externally (in non-Ada code).
3455 For further details on this pragma, see the
3456 DEC Ada Language Reference Manual, section 13.9a.3.1.
3458 @node Pragma Import_Function
3459 @unnumberedsec Pragma Import_Function
3460 @findex Import_Function
3464 @smallexample @c ada
3465 pragma Import_Function (
3466 [Internal =>] LOCAL_NAME,
3467 [, [External =>] EXTERNAL_SYMBOL]
3468 [, [Parameter_Types =>] PARAMETER_TYPES]
3469 [, [Result_Type =>] SUBTYPE_MARK]
3470 [, [Mechanism =>] MECHANISM]
3471 [, [Result_Mechanism =>] MECHANISM_NAME]
3472 [, [First_Optional_Parameter =>] IDENTIFIER]);
3476 | static_string_EXPRESSION
3480 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
3484 | subtype_Name ' Access
3488 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
3490 MECHANISM_ASSOCIATION ::=
3491 [formal_parameter_NAME =>] MECHANISM_NAME
3496 | Descriptor [([Class =>] CLASS_NAME)]
3497 | Short_Descriptor [([Class =>] CLASS_NAME)]
3499 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
3503 This pragma is used in conjunction with a pragma @code{Import} to
3504 specify additional information for an imported function. The pragma
3505 @code{Import} (or equivalent pragma @code{Interface}) must precede the
3506 @code{Import_Function} pragma and both must appear in the same
3507 declarative part as the function specification.
3509 The @var{Internal} argument must uniquely designate
3510 the function to which the
3511 pragma applies. If more than one function name exists of this name in
3512 the declarative part you must use the @code{Parameter_Types} and
3513 @var{Result_Type} parameters to achieve the required unique
3514 designation. Subtype marks in these parameters must exactly match the
3515 subtypes in the corresponding function specification, using positional
3516 notation to match parameters with subtype marks.
3517 The form with an @code{'Access} attribute can be used to match an
3518 anonymous access parameter.
3520 You may optionally use the @var{Mechanism} and @var{Result_Mechanism}
3521 parameters to specify passing mechanisms for the
3522 parameters and result. If you specify a single mechanism name, it
3523 applies to all parameters. Otherwise you may specify a mechanism on a
3524 parameter by parameter basis using either positional or named
3525 notation. If the mechanism is not specified, the default mechanism
3529 @cindex Passing by descriptor
3530 Passing by descriptor is supported only on the OpenVMS ports of GNAT@.
3531 The default behavior for Import_Function is to pass a 64bit descriptor
3532 unless short_descriptor is specified, then a 32bit descriptor is passed.
3534 @code{First_Optional_Parameter} applies only to OpenVMS ports of GNAT@.
3535 It specifies that the designated parameter and all following parameters
3536 are optional, meaning that they are not passed at the generated code
3537 level (this is distinct from the notion of optional parameters in Ada
3538 where the parameters are passed anyway with the designated optional
3539 parameters). All optional parameters must be of mode @code{IN} and have
3540 default parameter values that are either known at compile time
3541 expressions, or uses of the @code{'Null_Parameter} attribute.
3543 @node Pragma Import_Object
3544 @unnumberedsec Pragma Import_Object
3545 @findex Import_Object
3549 @smallexample @c ada
3550 pragma Import_Object
3551 [Internal =>] LOCAL_NAME
3552 [, [External =>] EXTERNAL_SYMBOL]
3553 [, [Size =>] EXTERNAL_SYMBOL]);
3557 | static_string_EXPRESSION
3561 This pragma designates an object as imported, and apart from the
3562 extended rules for external symbols, is identical in effect to the use of
3563 the normal @code{Import} pragma applied to an object. Unlike the
3564 subprogram case, you need not use a separate @code{Import} pragma,
3565 although you may do so (and probably should do so from a portability
3566 point of view). @var{size} is syntax checked, but otherwise ignored by
3569 @node Pragma Import_Procedure
3570 @unnumberedsec Pragma Import_Procedure
3571 @findex Import_Procedure
3575 @smallexample @c ada
3576 pragma Import_Procedure (
3577 [Internal =>] LOCAL_NAME
3578 [, [External =>] EXTERNAL_SYMBOL]
3579 [, [Parameter_Types =>] PARAMETER_TYPES]
3580 [, [Mechanism =>] MECHANISM]
3581 [, [First_Optional_Parameter =>] IDENTIFIER]);
3585 | static_string_EXPRESSION
3589 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
3593 | subtype_Name ' Access
3597 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
3599 MECHANISM_ASSOCIATION ::=
3600 [formal_parameter_NAME =>] MECHANISM_NAME
3605 | Descriptor [([Class =>] CLASS_NAME)]
3606 | Short_Descriptor [([Class =>] CLASS_NAME)]
3608 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
3612 This pragma is identical to @code{Import_Function} except that it
3613 applies to a procedure rather than a function and the parameters
3614 @code{Result_Type} and @code{Result_Mechanism} are not permitted.
3616 @node Pragma Import_Valued_Procedure
3617 @unnumberedsec Pragma Import_Valued_Procedure
3618 @findex Import_Valued_Procedure
3622 @smallexample @c ada
3623 pragma Import_Valued_Procedure (
3624 [Internal =>] LOCAL_NAME
3625 [, [External =>] EXTERNAL_SYMBOL]
3626 [, [Parameter_Types =>] PARAMETER_TYPES]
3627 [, [Mechanism =>] MECHANISM]
3628 [, [First_Optional_Parameter =>] IDENTIFIER]);
3632 | static_string_EXPRESSION
3636 | TYPE_DESIGNATOR @{, TYPE_DESIGNATOR@}
3640 | subtype_Name ' Access
3644 | (MECHANISM_ASSOCIATION @{, MECHANISM_ASSOCIATION@})
3646 MECHANISM_ASSOCIATION ::=
3647 [formal_parameter_NAME =>] MECHANISM_NAME
3652 | Descriptor [([Class =>] CLASS_NAME)]
3653 | Short_Descriptor [([Class =>] CLASS_NAME)]
3655 CLASS_NAME ::= ubs | ubsb | uba | s | sb | a | nca
3659 This pragma is identical to @code{Import_Procedure} except that the
3660 first parameter of @var{LOCAL_NAME}, which must be present, must be of
3661 mode @code{OUT}, and externally the subprogram is treated as a function
3662 with this parameter as the result of the function. The purpose of this
3663 capability is to allow the use of @code{OUT} and @code{IN OUT}
3664 parameters in interfacing to external functions (which are not permitted
3665 in Ada functions). You may optionally use the @code{Mechanism}
3666 parameters to specify passing mechanisms for the parameters.
3667 If you specify a single mechanism name, it applies to all parameters.
3668 Otherwise you may specify a mechanism on a parameter by parameter
3669 basis using either positional or named notation. If the mechanism is not
3670 specified, the default mechanism is used.
3672 Note that it is important to use this pragma in conjunction with a separate
3673 pragma Import that specifies the desired convention, since otherwise the
3674 default convention is Ada, which is almost certainly not what is required.
3676 @node Pragma Independent
3677 @unnumberedsec Pragma Independent
3682 @smallexample @c ada
3683 pragma Independent (Local_NAME);
3687 This pragma is standard in Ada 2012 mode (which also provides an aspect
3688 of the same name). It is also available as an implementation-defined
3689 pragma in all earlier versions. It specifies that the
3690 designated object or all objects of the designated type must be
3691 independently addressable. This means that separate tasks can safely
3692 manipulate such objects. For example, if two components of a record are
3693 independent, then two separate tasks may access these two components.
3695 constraints on the representation of the object (for instance prohibiting
3698 @node Pragma Independent_Components
3699 @unnumberedsec Pragma Independent_Components
3700 @findex Independent_Components
3704 @smallexample @c ada
3705 pragma Independent_Components (Local_NAME);
3709 This pragma is standard in Ada 2012 mode (which also provides an aspect
3710 of the same name). It is also available as an implementation-defined
3711 pragma in all earlier versions. It specifies that the components of the
3712 designated object, or the components of each object of the designated
3714 independently addressable. This means that separate tasks can safely
3715 manipulate separate components in the composite object. This may place
3716 constraints on the representation of the object (for instance prohibiting
3719 @node Pragma Initial_Condition
3720 @unnumberedsec Pragma Initial_Condition
3721 @findex Initial_Condition
3723 For the description of this pragma, see SPARK 2014 Reference Manual,
3726 @node Pragma Initialize_Scalars
3727 @unnumberedsec Pragma Initialize_Scalars
3728 @findex Initialize_Scalars
3729 @cindex debugging with Initialize_Scalars
3733 @smallexample @c ada
3734 pragma Initialize_Scalars;
3738 This pragma is similar to @code{Normalize_Scalars} conceptually but has
3739 two important differences. First, there is no requirement for the pragma
3740 to be used uniformly in all units of a partition, in particular, it is fine
3741 to use this just for some or all of the application units of a partition,
3742 without needing to recompile the run-time library.
3744 In the case where some units are compiled with the pragma, and some without,
3745 then a declaration of a variable where the type is defined in package
3746 Standard or is locally declared will always be subject to initialization,
3747 as will any declaration of a scalar variable. For composite variables,
3748 whether the variable is initialized may also depend on whether the package
3749 in which the type of the variable is declared is compiled with the pragma.
3751 The other important difference is that you can control the value used
3752 for initializing scalar objects. At bind time, you can select several
3753 options for initialization. You can
3754 initialize with invalid values (similar to Normalize_Scalars, though for
3755 Initialize_Scalars it is not always possible to determine the invalid
3756 values in complex cases like signed component fields with non-standard
3757 sizes). You can also initialize with high or
3758 low values, or with a specified bit pattern. See the @value{EDITION}
3759 User's Guide for binder options for specifying these cases.
3761 This means that you can compile a program, and then without having to
3762 recompile the program, you can run it with different values being used
3763 for initializing otherwise uninitialized values, to test if your program
3764 behavior depends on the choice. Of course the behavior should not change,
3765 and if it does, then most likely you have an incorrect reference to an
3766 uninitialized value.
3768 It is even possible to change the value at execution time eliminating even
3769 the need to rebind with a different switch using an environment variable.
3770 See the @value{EDITION} User's Guide for details.
3772 Note that pragma @code{Initialize_Scalars} is particularly useful in
3773 conjunction with the enhanced validity checking that is now provided
3774 in GNAT, which checks for invalid values under more conditions.
3775 Using this feature (see description of the @option{-gnatV} flag in the
3776 @value{EDITION} User's Guide) in conjunction with
3777 pragma @code{Initialize_Scalars}
3778 provides a powerful new tool to assist in the detection of problems
3779 caused by uninitialized variables.
3781 Note: the use of @code{Initialize_Scalars} has a fairly extensive
3782 effect on the generated code. This may cause your code to be
3783 substantially larger. It may also cause an increase in the amount
3784 of stack required, so it is probably a good idea to turn on stack
3785 checking (see description of stack checking in the @value{EDITION}
3786 User's Guide) when using this pragma.
3788 @node Pragma Initializes
3789 @unnumberedsec Pragma Initializes
3792 For the description of this pragma, see SPARK 2014 Reference Manual,
3795 @node Pragma Inline_Always
3796 @unnumberedsec Pragma Inline_Always
3797 @findex Inline_Always
3801 @smallexample @c ada
3802 pragma Inline_Always (NAME [, NAME]);
3806 Similar to pragma @code{Inline} except that inlining is not subject to
3807 the use of option @option{-gnatn} or @option{-gnatN} and the inlining
3808 happens regardless of whether these options are used.
3810 @node Pragma Inline_Generic
3811 @unnumberedsec Pragma Inline_Generic
3812 @findex Inline_Generic
3816 @smallexample @c ada
3817 pragma Inline_Generic (GNAME @{, GNAME@});
3819 GNAME ::= generic_unit_NAME | generic_instance_NAME
3823 This pragma is provided for compatibility with Dec Ada 83. It has
3824 no effect in @code{GNAT} (which always inlines generics), other
3825 than to check that the given names are all names of generic units or
3828 @node Pragma Interface
3829 @unnumberedsec Pragma Interface
3834 @smallexample @c ada
3836 [Convention =>] convention_identifier,
3837 [Entity =>] local_NAME
3838 [, [External_Name =>] static_string_expression]
3839 [, [Link_Name =>] static_string_expression]);
3843 This pragma is identical in syntax and semantics to
3844 the standard Ada pragma @code{Import}. It is provided for compatibility
3845 with Ada 83. The definition is upwards compatible both with pragma
3846 @code{Interface} as defined in the Ada 83 Reference Manual, and also
3847 with some extended implementations of this pragma in certain Ada 83
3848 implementations. The only difference between pragma @code{Interface}
3849 and pragma @code{Import} is that there is special circuitry to allow
3850 both pragmas to appear for the same subprogram entity (normally it
3851 is illegal to have multiple @code{Import} pragmas. This is useful in
3852 maintaining Ada 83/Ada 95 compatibility and is compatible with other
3855 @node Pragma Interface_Name
3856 @unnumberedsec Pragma Interface_Name
3857 @findex Interface_Name
3861 @smallexample @c ada
3862 pragma Interface_Name (
3863 [Entity =>] LOCAL_NAME
3864 [, [External_Name =>] static_string_EXPRESSION]
3865 [, [Link_Name =>] static_string_EXPRESSION]);
3869 This pragma provides an alternative way of specifying the interface name
3870 for an interfaced subprogram, and is provided for compatibility with Ada
3871 83 compilers that use the pragma for this purpose. You must provide at
3872 least one of @var{External_Name} or @var{Link_Name}.
3874 @node Pragma Interrupt_Handler
3875 @unnumberedsec Pragma Interrupt_Handler
3876 @findex Interrupt_Handler
3880 @smallexample @c ada
3881 pragma Interrupt_Handler (procedure_LOCAL_NAME);
3885 This program unit pragma is supported for parameterless protected procedures
3886 as described in Annex C of the Ada Reference Manual. On the AAMP target
3887 the pragma can also be specified for nonprotected parameterless procedures
3888 that are declared at the library level (which includes procedures
3889 declared at the top level of a library package). In the case of AAMP,
3890 when this pragma is applied to a nonprotected procedure, the instruction
3891 @code{IERET} is generated for returns from the procedure, enabling
3892 maskable interrupts, in place of the normal return instruction.
3894 @node Pragma Interrupt_State
3895 @unnumberedsec Pragma Interrupt_State
3896 @findex Interrupt_State
3900 @smallexample @c ada
3901 pragma Interrupt_State
3903 [State =>] SYSTEM | RUNTIME | USER);
3907 Normally certain interrupts are reserved to the implementation. Any attempt
3908 to attach an interrupt causes Program_Error to be raised, as described in
3909 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
3910 many systems for an @kbd{Ctrl-C} interrupt. Normally this interrupt is
3911 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
3912 interrupt execution. Additionally, signals such as @code{SIGSEGV},
3913 @code{SIGABRT}, @code{SIGFPE} and @code{SIGILL} are often mapped to specific
3914 Ada exceptions, or used to implement run-time functions such as the
3915 @code{abort} statement and stack overflow checking.
3917 Pragma @code{Interrupt_State} provides a general mechanism for overriding
3918 such uses of interrupts. It subsumes the functionality of pragma
3919 @code{Unreserve_All_Interrupts}. Pragma @code{Interrupt_State} is not
3920 available on Windows or VMS. On all other platforms than VxWorks,
3921 it applies to signals; on VxWorks, it applies to vectored hardware interrupts
3922 and may be used to mark interrupts required by the board support package
3925 Interrupts can be in one of three states:
3929 The interrupt is reserved (no Ada handler can be installed), and the
3930 Ada run-time may not install a handler. As a result you are guaranteed
3931 standard system default action if this interrupt is raised.
3935 The interrupt is reserved (no Ada handler can be installed). The run time
3936 is allowed to install a handler for internal control purposes, but is
3937 not required to do so.
3941 The interrupt is unreserved. The user may install a handler to provide
3946 These states are the allowed values of the @code{State} parameter of the
3947 pragma. The @code{Name} parameter is a value of the type
3948 @code{Ada.Interrupts.Interrupt_ID}. Typically, it is a name declared in
3949 @code{Ada.Interrupts.Names}.
3951 This is a configuration pragma, and the binder will check that there
3952 are no inconsistencies between different units in a partition in how a
3953 given interrupt is specified. It may appear anywhere a pragma is legal.
3955 The effect is to move the interrupt to the specified state.
3957 By declaring interrupts to be SYSTEM, you guarantee the standard system
3958 action, such as a core dump.
3960 By declaring interrupts to be USER, you guarantee that you can install
3963 Note that certain signals on many operating systems cannot be caught and
3964 handled by applications. In such cases, the pragma is ignored. See the
3965 operating system documentation, or the value of the array @code{Reserved}
3966 declared in the spec of package @code{System.OS_Interface}.
3968 Overriding the default state of signals used by the Ada runtime may interfere
3969 with an application's runtime behavior in the cases of the synchronous signals,
3970 and in the case of the signal used to implement the @code{abort} statement.
3972 @node Pragma Invariant
3973 @unnumberedsec Pragma Invariant
3978 @smallexample @c ada
3980 ([Entity =>] private_type_LOCAL_NAME,
3981 [Check =>] EXPRESSION
3982 [,[Message =>] String_Expression]);
3986 This pragma provides exactly the same capabilities as the Type_Invariant aspect
3987 defined in AI05-0146-1, and in the Ada 2012 Reference Manual. The
3988 Type_Invariant aspect is fully implemented in Ada 2012 mode, but since it
3989 requires the use of the aspect syntax, which is not available except in 2012
3990 mode, it is not possible to use the Type_Invariant aspect in earlier versions
3991 of Ada. However the Invariant pragma may be used in any version of Ada. Also
3992 note that the aspect Invariant is a synonym in GNAT for the aspect
3993 Type_Invariant, but there is no pragma Type_Invariant.
3995 The pragma must appear within the visible part of the package specification,
3996 after the type to which its Entity argument appears. As with the Invariant
3997 aspect, the Check expression is not analyzed until the end of the visible
3998 part of the package, so it may contain forward references. The Message
3999 argument, if present, provides the exception message used if the invariant
4000 is violated. If no Message parameter is provided, a default message that
4001 identifies the line on which the pragma appears is used.
4003 It is permissible to have multiple Invariants for the same type entity, in
4004 which case they are and'ed together. It is permissible to use this pragma
4005 in Ada 2012 mode, but you cannot have both an invariant aspect and an
4006 invariant pragma for the same entity.
4008 For further details on the use of this pragma, see the Ada 2012 documentation
4009 of the Type_Invariant aspect.
4011 @node Pragma Java_Constructor
4012 @unnumberedsec Pragma Java_Constructor
4013 @findex Java_Constructor
4017 @smallexample @c ada
4018 pragma Java_Constructor ([Entity =>] function_LOCAL_NAME);
4022 This pragma is used to assert that the specified Ada function should be
4023 mapped to the Java constructor for some Ada tagged record type.
4025 See section 7.3.2 of the
4026 @code{GNAT User's Guide: Supplement for the JVM Platform.}
4027 for related information.
4029 @node Pragma Java_Interface
4030 @unnumberedsec Pragma Java_Interface
4031 @findex Java_Interface
4035 @smallexample @c ada
4036 pragma Java_Interface ([Entity =>] abstract_tagged_type_LOCAL_NAME);
4040 This pragma is used to assert that the specified Ada abstract tagged type
4041 is to be mapped to a Java interface name.
4043 See sections 7.1 and 7.2 of the
4044 @code{GNAT User's Guide: Supplement for the JVM Platform.}
4045 for related information.
4047 @node Pragma Keep_Names
4048 @unnumberedsec Pragma Keep_Names
4053 @smallexample @c ada
4054 pragma Keep_Names ([On =>] enumeration_first_subtype_LOCAL_NAME);
4058 The @var{LOCAL_NAME} argument
4059 must refer to an enumeration first subtype
4060 in the current declarative part. The effect is to retain the enumeration
4061 literal names for use by @code{Image} and @code{Value} even if a global
4062 @code{Discard_Names} pragma applies. This is useful when you want to
4063 generally suppress enumeration literal names and for example you therefore
4064 use a @code{Discard_Names} pragma in the @file{gnat.adc} file, but you
4065 want to retain the names for specific enumeration types.
4067 @node Pragma License
4068 @unnumberedsec Pragma License
4070 @cindex License checking
4074 @smallexample @c ada
4075 pragma License (Unrestricted | GPL | Modified_GPL | Restricted);
4079 This pragma is provided to allow automated checking for appropriate license
4080 conditions with respect to the standard and modified GPL@. A pragma
4081 @code{License}, which is a configuration pragma that typically appears at
4082 the start of a source file or in a separate @file{gnat.adc} file, specifies
4083 the licensing conditions of a unit as follows:
4087 This is used for a unit that can be freely used with no license restrictions.
4088 Examples of such units are public domain units, and units from the Ada
4092 This is used for a unit that is licensed under the unmodified GPL, and which
4093 therefore cannot be @code{with}'ed by a restricted unit.
4096 This is used for a unit licensed under the GNAT modified GPL that includes
4097 a special exception paragraph that specifically permits the inclusion of
4098 the unit in programs without requiring the entire program to be released
4102 This is used for a unit that is restricted in that it is not permitted to
4103 depend on units that are licensed under the GPL@. Typical examples are
4104 proprietary code that is to be released under more restrictive license
4105 conditions. Note that restricted units are permitted to @code{with} units
4106 which are licensed under the modified GPL (this is the whole point of the
4112 Normally a unit with no @code{License} pragma is considered to have an
4113 unknown license, and no checking is done. However, standard GNAT headers
4114 are recognized, and license information is derived from them as follows.
4118 A GNAT license header starts with a line containing 78 hyphens. The following
4119 comment text is searched for the appearance of any of the following strings.
4121 If the string ``GNU General Public License'' is found, then the unit is assumed
4122 to have GPL license, unless the string ``As a special exception'' follows, in
4123 which case the license is assumed to be modified GPL@.
4125 If one of the strings
4126 ``This specification is adapted from the Ada Semantic Interface'' or
4127 ``This specification is derived from the Ada Reference Manual'' is found
4128 then the unit is assumed to be unrestricted.
4132 These default actions means that a program with a restricted license pragma
4133 will automatically get warnings if a GPL unit is inappropriately
4134 @code{with}'ed. For example, the program:
4136 @smallexample @c ada
4139 procedure Secret_Stuff is
4145 if compiled with pragma @code{License} (@code{Restricted}) in a
4146 @file{gnat.adc} file will generate the warning:
4151 >>> license of withed unit "Sem_Ch3" is incompatible
4153 2. with GNAT.Sockets;
4154 3. procedure Secret_Stuff is
4158 Here we get a warning on @code{Sem_Ch3} since it is part of the GNAT
4159 compiler and is licensed under the
4160 GPL, but no warning for @code{GNAT.Sockets} which is part of the GNAT
4161 run time, and is therefore licensed under the modified GPL@.
4163 @node Pragma Link_With
4164 @unnumberedsec Pragma Link_With
4169 @smallexample @c ada
4170 pragma Link_With (static_string_EXPRESSION @{,static_string_EXPRESSION@});
4174 This pragma is provided for compatibility with certain Ada 83 compilers.
4175 It has exactly the same effect as pragma @code{Linker_Options} except
4176 that spaces occurring within one of the string expressions are treated
4177 as separators. For example, in the following case:
4179 @smallexample @c ada
4180 pragma Link_With ("-labc -ldef");
4184 results in passing the strings @code{-labc} and @code{-ldef} as two
4185 separate arguments to the linker. In addition pragma Link_With allows
4186 multiple arguments, with the same effect as successive pragmas.
4188 @node Pragma Linker_Alias
4189 @unnumberedsec Pragma Linker_Alias
4190 @findex Linker_Alias
4194 @smallexample @c ada
4195 pragma Linker_Alias (
4196 [Entity =>] LOCAL_NAME,
4197 [Target =>] static_string_EXPRESSION);
4201 @var{LOCAL_NAME} must refer to an object that is declared at the library
4202 level. This pragma establishes the given entity as a linker alias for the
4203 given target. It is equivalent to @code{__attribute__((alias))} in GNU C
4204 and causes @var{LOCAL_NAME} to be emitted as an alias for the symbol
4205 @var{static_string_EXPRESSION} in the object file, that is to say no space
4206 is reserved for @var{LOCAL_NAME} by the assembler and it will be resolved
4207 to the same address as @var{static_string_EXPRESSION} by the linker.
4209 The actual linker name for the target must be used (e.g.@: the fully
4210 encoded name with qualification in Ada, or the mangled name in C++),
4211 or it must be declared using the C convention with @code{pragma Import}
4212 or @code{pragma Export}.
4214 Not all target machines support this pragma. On some of them it is accepted
4215 only if @code{pragma Weak_External} has been applied to @var{LOCAL_NAME}.
4217 @smallexample @c ada
4218 -- Example of the use of pragma Linker_Alias
4222 pragma Export (C, i);
4224 new_name_for_i : Integer;
4225 pragma Linker_Alias (new_name_for_i, "i");
4229 @node Pragma Linker_Constructor
4230 @unnumberedsec Pragma Linker_Constructor
4231 @findex Linker_Constructor
4235 @smallexample @c ada
4236 pragma Linker_Constructor (procedure_LOCAL_NAME);
4240 @var{procedure_LOCAL_NAME} must refer to a parameterless procedure that
4241 is declared at the library level. A procedure to which this pragma is
4242 applied will be treated as an initialization routine by the linker.
4243 It is equivalent to @code{__attribute__((constructor))} in GNU C and
4244 causes @var{procedure_LOCAL_NAME} to be invoked before the entry point
4245 of the executable is called (or immediately after the shared library is
4246 loaded if the procedure is linked in a shared library), in particular
4247 before the Ada run-time environment is set up.
4249 Because of these specific contexts, the set of operations such a procedure
4250 can perform is very limited and the type of objects it can manipulate is
4251 essentially restricted to the elementary types. In particular, it must only
4252 contain code to which pragma Restrictions (No_Elaboration_Code) applies.
4254 This pragma is used by GNAT to implement auto-initialization of shared Stand
4255 Alone Libraries, which provides a related capability without the restrictions
4256 listed above. Where possible, the use of Stand Alone Libraries is preferable
4257 to the use of this pragma.
4259 @node Pragma Linker_Destructor
4260 @unnumberedsec Pragma Linker_Destructor
4261 @findex Linker_Destructor
4265 @smallexample @c ada
4266 pragma Linker_Destructor (procedure_LOCAL_NAME);
4270 @var{procedure_LOCAL_NAME} must refer to a parameterless procedure that
4271 is declared at the library level. A procedure to which this pragma is
4272 applied will be treated as a finalization routine by the linker.
4273 It is equivalent to @code{__attribute__((destructor))} in GNU C and
4274 causes @var{procedure_LOCAL_NAME} to be invoked after the entry point
4275 of the executable has exited (or immediately before the shared library
4276 is unloaded if the procedure is linked in a shared library), in particular
4277 after the Ada run-time environment is shut down.
4279 See @code{pragma Linker_Constructor} for the set of restrictions that apply
4280 because of these specific contexts.
4282 @node Pragma Linker_Section
4283 @unnumberedsec Pragma Linker_Section
4284 @findex Linker_Section
4288 @smallexample @c ada
4289 pragma Linker_Section (
4290 [Entity =>] LOCAL_NAME,
4291 [Section =>] static_string_EXPRESSION);
4295 @var{LOCAL_NAME} must refer to an object, type, or subprogram that is
4296 declared at the library level. This pragma specifies the name of the
4297 linker section for the given entity. It is equivalent to
4298 @code{__attribute__((section))} in GNU C and causes @var{LOCAL_NAME} to
4299 be placed in the @var{static_string_EXPRESSION} section of the
4300 executable (assuming the linker doesn't rename the section).
4301 GNAT also provides an implementation defined aspect of the same name.
4303 In the case of specifying this aspect for a type, the effect is to
4304 specify the corresponding for all library level objects of the type which
4305 do not have an explicit linker section set. Note that this only applies to
4306 whole objects, not to components of composite objects.
4308 In the case of a subprogram, the linker section applies to all previously
4309 declared matching overloaded subprograms in the current declarative part
4310 which do not already have a linker section assigned. The linker section
4311 aspect is useful in this case for specifying different linker sections
4312 for different elements of such an overloaded set.
4314 Note that an empty string specifies that no linker section is specified.
4315 This is not quite the same as omitting the pragma or aspect, since it
4316 can be used to specify that one element of an overloaded set of subprograms
4317 has the default linker section, or that one object of a type for which a
4318 linker section is specified should has the default linker section.
4320 The compiler normally places library-level entities in standard sections
4321 depending on the class: procedures and functions generally go in the
4322 @code{.text} section, initialized variables in the @code{.data} section
4323 and uninitialized variables in the @code{.bss} section.
4325 Other, special sections may exist on given target machines to map special
4326 hardware, for example I/O ports or flash memory. This pragma is a means to
4327 defer the final layout of the executable to the linker, thus fully working
4328 at the symbolic level with the compiler.
4330 Some file formats do not support arbitrary sections so not all target
4331 machines support this pragma. The use of this pragma may cause a program
4332 execution to be erroneous if it is used to place an entity into an
4333 inappropriate section (e.g.@: a modified variable into the @code{.text}
4334 section). See also @code{pragma Persistent_BSS}.
4336 @smallexample @c ada
4337 -- Example of the use of pragma Linker_Section
4341 pragma Volatile (Port_A);
4342 pragma Linker_Section (Port_A, ".bss.port_a");
4345 pragma Volatile (Port_B);
4346 pragma Linker_Section (Port_B, ".bss.port_b");
4348 type Port_Type is new Integer with Linker_Section => ".bss";
4349 PA : Port_Type with Linker_Section => ".bss.PA";
4350 PB : Port_Type; -- ends up in linker section ".bss"
4352 procedure Q with Linker_Section => "Qsection";
4356 @node Pragma Long_Float
4357 @unnumberedsec Pragma Long_Float
4363 @smallexample @c ada
4364 pragma Long_Float (FLOAT_FORMAT);
4366 FLOAT_FORMAT ::= D_Float | G_Float
4370 This pragma is implemented only in the OpenVMS implementation of GNAT@.
4371 It allows control over the internal representation chosen for the predefined
4372 type @code{Long_Float} and for floating point type representations with
4373 @code{digits} specified in the range 7 through 15.
4374 For further details on this pragma, see the
4375 @cite{DEC Ada Language Reference Manual}, section 3.5.7b. Note that to use
4376 this pragma, the standard runtime libraries must be recompiled.
4378 @node Pragma Loop_Invariant
4379 @unnumberedsec Pragma Loop_Invariant
4380 @findex Loop_Invariant
4384 @smallexample @c ada
4385 pragma Loop_Invariant ( boolean_EXPRESSION );
4389 The effect of this pragma is similar to that of pragma @code{Assert},
4390 except that in an @code{Assertion_Policy} pragma, the identifier
4391 @code{Loop_Invariant} is used to control whether it is ignored or checked
4394 @code{Loop_Invariant} can only appear as one of the items in the sequence
4395 of statements of a loop body, or nested inside block statements that
4396 appear in the sequence of statements of a loop body.
4397 The intention is that it be used to
4398 represent a "loop invariant" assertion, i.e. something that is true each
4399 time through the loop, and which can be used to show that the loop is
4400 achieving its purpose.
4402 Multiple @code{Loop_Invariant} and @code{Loop_Variant} pragmas that
4403 apply to the same loop should be grouped in the same sequence of
4406 To aid in writing such invariants, the special attribute @code{Loop_Entry}
4407 may be used to refer to the value of an expression on entry to the loop. This
4408 attribute can only be used within the expression of a @code{Loop_Invariant}
4409 pragma. For full details, see documentation of attribute @code{Loop_Entry}.
4411 @node Pragma Loop_Optimize
4412 @unnumberedsec Pragma Loop_Optimize
4413 @findex Loop_Optimize
4417 @smallexample @c ada
4418 pragma Loop_Optimize (OPTIMIZATION_HINT @{, OPTIMIZATION_HINT@});
4420 OPTIMIZATION_HINT ::= No_Unroll | Unroll | No_Vector | Vector
4424 This pragma must appear immediately within a loop statement. It allows the
4425 programmer to specify optimization hints for the enclosing loop. The hints
4426 are not mutually exclusive and can be freely mixed, but not all combinations
4427 will yield a sensible outcome.
4429 There are four supported optimization hints for a loop:
4433 The loop must not be unrolled. This is a strong hint: the compiler will not
4434 unroll a loop marked with this hint.
4438 The loop should be unrolled. This is a weak hint: the compiler will try to
4439 apply unrolling to this loop preferably to other optimizations, notably
4440 vectorization, but there is no guarantee that the loop will be unrolled.
4444 The loop must not be vectorized. This is a strong hint: the compiler will not
4445 vectorize a loop marked with this hint.
4449 The loop should be vectorized. This is a weak hint: the compiler will try to
4450 apply vectorization to this loop preferably to other optimizations, notably
4451 unrolling, but there is no guarantee that the loop will be vectorized.
4455 These hints do not void the need to pass the appropriate switches to the
4456 compiler in order to enable the relevant optimizations, that is to say
4457 @option{-funroll-loops} for unrolling and @option{-ftree-vectorize} for
4460 @node Pragma Loop_Variant
4461 @unnumberedsec Pragma Loop_Variant
4462 @findex Loop_Variant
4466 @smallexample @c ada
4467 pragma Loop_Variant ( LOOP_VARIANT_ITEM @{, LOOP_VARIANT_ITEM @} );
4468 LOOP_VARIANT_ITEM ::= CHANGE_DIRECTION => discrete_EXPRESSION
4469 CHANGE_DIRECTION ::= Increases | Decreases
4473 @code{Loop_Variant} can only appear as one of the items in the sequence
4474 of statements of a loop body, or nested inside block statements that
4475 appear in the sequence of statements of a loop body.
4476 It allows the specification of quantities which must always
4477 decrease or increase in successive iterations of the loop. In its simplest
4478 form, just one expression is specified, whose value must increase or decrease
4479 on each iteration of the loop.
4481 In a more complex form, multiple arguments can be given which are intepreted
4482 in a nesting lexicographic manner. For example:
4484 @smallexample @c ada
4485 pragma Loop_Variant (Increases => X, Decreases => Y);
4489 specifies that each time through the loop either X increases, or X stays
4490 the same and Y decreases. A @code{Loop_Variant} pragma ensures that the
4491 loop is making progress. It can be useful in helping to show informally
4492 or prove formally that the loop always terminates.
4494 @code{Loop_Variant} is an assertion whose effect can be controlled using
4495 an @code{Assertion_Policy} with a check name of @code{Loop_Variant}. The
4496 policy can be @code{Check} to enable the loop variant check, @code{Ignore}
4497 to ignore the check (in which case the pragma has no effect on the program),
4498 or @code{Disable} in which case the pragma is not even checked for correct
4501 Multiple @code{Loop_Invariant} and @code{Loop_Variant} pragmas that
4502 apply to the same loop should be grouped in the same sequence of
4505 The @code{Loop_Entry} attribute may be used within the expressions of the
4506 @code{Loop_Variant} pragma to refer to values on entry to the loop.
4508 @node Pragma Machine_Attribute
4509 @unnumberedsec Pragma Machine_Attribute
4510 @findex Machine_Attribute
4514 @smallexample @c ada
4515 pragma Machine_Attribute (
4516 [Entity =>] LOCAL_NAME,
4517 [Attribute_Name =>] static_string_EXPRESSION
4518 [, [Info =>] static_EXPRESSION] );
4522 Machine-dependent attributes can be specified for types and/or
4523 declarations. This pragma is semantically equivalent to
4524 @code{__attribute__((@var{attribute_name}))} (if @var{info} is not
4525 specified) or @code{__attribute__((@var{attribute_name}(@var{info})))}
4526 in GNU C, where @code{@var{attribute_name}} is recognized by the
4527 compiler middle-end or the @code{TARGET_ATTRIBUTE_TABLE} machine
4528 specific macro. A string literal for the optional parameter @var{info}
4529 is transformed into an identifier, which may make this pragma unusable
4530 for some attributes. @xref{Target Attributes,, Defining target-specific
4531 uses of @code{__attribute__}, gccint, GNU Compiler Collection (GCC)
4532 Internals}, further information.
4535 @unnumberedsec Pragma Main
4541 @smallexample @c ada
4543 (MAIN_OPTION [, MAIN_OPTION]);
4546 [Stack_Size =>] static_integer_EXPRESSION
4547 | [Task_Stack_Size_Default =>] static_integer_EXPRESSION
4548 | [Time_Slicing_Enabled =>] static_boolean_EXPRESSION
4552 This pragma is provided for compatibility with OpenVMS VAX Systems. It has
4553 no effect in GNAT, other than being syntax checked.
4555 @node Pragma Main_Storage
4556 @unnumberedsec Pragma Main_Storage
4558 @findex Main_Storage
4562 @smallexample @c ada
4564 (MAIN_STORAGE_OPTION [, MAIN_STORAGE_OPTION]);
4566 MAIN_STORAGE_OPTION ::=
4567 [WORKING_STORAGE =>] static_SIMPLE_EXPRESSION
4568 | [TOP_GUARD =>] static_SIMPLE_EXPRESSION
4572 This pragma is provided for compatibility with OpenVMS VAX Systems. It has
4573 no effect in GNAT, other than being syntax checked. Note that the pragma
4574 also has no effect in DEC Ada 83 for OpenVMS Alpha Systems.
4576 @node Pragma No_Body
4577 @unnumberedsec Pragma No_Body
4582 @smallexample @c ada
4587 There are a number of cases in which a package spec does not require a body,
4588 and in fact a body is not permitted. GNAT will not permit the spec to be
4589 compiled if there is a body around. The pragma No_Body allows you to provide
4590 a body file, even in a case where no body is allowed. The body file must
4591 contain only comments and a single No_Body pragma. This is recognized by
4592 the compiler as indicating that no body is logically present.
4594 This is particularly useful during maintenance when a package is modified in
4595 such a way that a body needed before is no longer needed. The provision of a
4596 dummy body with a No_Body pragma ensures that there is no interference from
4597 earlier versions of the package body.
4599 @node Pragma No_Inline
4600 @unnumberedsec Pragma No_Inline
4605 @smallexample @c ada
4606 pragma No_Inline (NAME @{, NAME@});
4610 This pragma suppresses inlining for the callable entity or the instances of
4611 the generic subprogram designated by @var{NAME}, including inlining that
4612 results from the use of pragma @code{Inline}. This pragma is always active,
4613 in particular it is not subject to the use of option @option{-gnatn} or
4614 @option{-gnatN}. It is illegal to specify both pragma @code{No_Inline} and
4615 pragma @code{Inline_Always} for the same @var{NAME}.
4617 @node Pragma No_Return
4618 @unnumberedsec Pragma No_Return
4623 @smallexample @c ada
4624 pragma No_Return (procedure_LOCAL_NAME @{, procedure_LOCAL_NAME@});
4628 Each @var{procedure_LOCAL_NAME} argument must refer to one or more procedure
4629 declarations in the current declarative part. A procedure to which this
4630 pragma is applied may not contain any explicit @code{return} statements.
4631 In addition, if the procedure contains any implicit returns from falling
4632 off the end of a statement sequence, then execution of that implicit
4633 return will cause Program_Error to be raised.
4635 One use of this pragma is to identify procedures whose only purpose is to raise
4636 an exception. Another use of this pragma is to suppress incorrect warnings
4637 about missing returns in functions, where the last statement of a function
4638 statement sequence is a call to such a procedure.
4640 Note that in Ada 2005 mode, this pragma is part of the language. It is
4641 available in all earlier versions of Ada as an implementation-defined
4644 @node Pragma No_Run_Time
4645 @unnumberedsec Pragma No_Run_Time
4650 @smallexample @c ada
4655 This is an obsolete configuration pragma that historically was used to
4656 setup what is now called the "zero footprint" library. It causes any
4657 library units outside this basic library to be ignored. The use of
4658 this pragma has been superseded by the general configurable run-time
4659 capability of @code{GNAT} where the compiler takes into account whatever
4660 units happen to be accessible in the library.
4662 @node Pragma No_Strict_Aliasing
4663 @unnumberedsec Pragma No_Strict_Aliasing
4664 @findex No_Strict_Aliasing
4668 @smallexample @c ada
4669 pragma No_Strict_Aliasing [([Entity =>] type_LOCAL_NAME)];
4673 @var{type_LOCAL_NAME} must refer to an access type
4674 declaration in the current declarative part. The effect is to inhibit
4675 strict aliasing optimization for the given type. The form with no
4676 arguments is a configuration pragma which applies to all access types
4677 declared in units to which the pragma applies. For a detailed
4678 description of the strict aliasing optimization, and the situations
4679 in which it must be suppressed, see @ref{Optimization and Strict
4680 Aliasing,,, gnat_ugn, @value{EDITION} User's Guide}.
4682 This pragma currently has no effects on access to unconstrained array types.
4684 @node Pragma Normalize_Scalars
4685 @unnumberedsec Pragma Normalize_Scalars
4686 @findex Normalize_Scalars
4690 @smallexample @c ada
4691 pragma Normalize_Scalars;
4695 This is a language defined pragma which is fully implemented in GNAT@. The
4696 effect is to cause all scalar objects that are not otherwise initialized
4697 to be initialized. The initial values are implementation dependent and
4701 @item Standard.Character
4703 Objects whose root type is Standard.Character are initialized to
4704 Character'Last unless the subtype range excludes NUL (in which case
4705 NUL is used). This choice will always generate an invalid value if
4708 @item Standard.Wide_Character
4710 Objects whose root type is Standard.Wide_Character are initialized to
4711 Wide_Character'Last unless the subtype range excludes NUL (in which case
4712 NUL is used). This choice will always generate an invalid value if
4715 @item Standard.Wide_Wide_Character
4717 Objects whose root type is Standard.Wide_Wide_Character are initialized to
4718 the invalid value 16#FFFF_FFFF# unless the subtype range excludes NUL (in
4719 which case NUL is used). This choice will always generate an invalid value if
4724 Objects of an integer type are treated differently depending on whether
4725 negative values are present in the subtype. If no negative values are
4726 present, then all one bits is used as the initial value except in the
4727 special case where zero is excluded from the subtype, in which case
4728 all zero bits are used. This choice will always generate an invalid
4729 value if one exists.
4731 For subtypes with negative values present, the largest negative number
4732 is used, except in the unusual case where this largest negative number
4733 is in the subtype, and the largest positive number is not, in which case
4734 the largest positive value is used. This choice will always generate
4735 an invalid value if one exists.
4737 @item Floating-Point Types
4738 Objects of all floating-point types are initialized to all 1-bits. For
4739 standard IEEE format, this corresponds to a NaN (not a number) which is
4740 indeed an invalid value.
4742 @item Fixed-Point Types
4743 Objects of all fixed-point types are treated as described above for integers,
4744 with the rules applying to the underlying integer value used to represent
4745 the fixed-point value.
4748 Objects of a modular type are initialized to all one bits, except in
4749 the special case where zero is excluded from the subtype, in which
4750 case all zero bits are used. This choice will always generate an
4751 invalid value if one exists.
4753 @item Enumeration types
4754 Objects of an enumeration type are initialized to all one-bits, i.e.@: to
4755 the value @code{2 ** typ'Size - 1} unless the subtype excludes the literal
4756 whose Pos value is zero, in which case a code of zero is used. This choice
4757 will always generate an invalid value if one exists.
4761 @node Pragma Obsolescent
4762 @unnumberedsec Pragma Obsolescent
4767 @smallexample @c ada
4770 pragma Obsolescent (
4771 [Message =>] static_string_EXPRESSION
4772 [,[Version =>] Ada_05]]);
4774 pragma Obsolescent (
4776 [,[Message =>] static_string_EXPRESSION
4777 [,[Version =>] Ada_05]] );
4781 This pragma can occur immediately following a declaration of an entity,
4782 including the case of a record component. If no Entity argument is present,
4783 then this declaration is the one to which the pragma applies. If an Entity
4784 parameter is present, it must either match the name of the entity in this
4785 declaration, or alternatively, the pragma can immediately follow an enumeration
4786 type declaration, where the Entity argument names one of the enumeration
4789 This pragma is used to indicate that the named entity
4790 is considered obsolescent and should not be used. Typically this is
4791 used when an API must be modified by eventually removing or modifying
4792 existing subprograms or other entities. The pragma can be used at an
4793 intermediate stage when the entity is still present, but will be
4796 The effect of this pragma is to output a warning message on a reference to
4797 an entity thus marked that the subprogram is obsolescent if the appropriate
4798 warning option in the compiler is activated. If the Message parameter is
4799 present, then a second warning message is given containing this text. In
4800 addition, a reference to the entity is considered to be a violation of pragma
4801 Restrictions (No_Obsolescent_Features).
4803 This pragma can also be used as a program unit pragma for a package,
4804 in which case the entity name is the name of the package, and the
4805 pragma indicates that the entire package is considered
4806 obsolescent. In this case a client @code{with}'ing such a package
4807 violates the restriction, and the @code{with} statement is
4808 flagged with warnings if the warning option is set.
4810 If the Version parameter is present (which must be exactly
4811 the identifier Ada_05, no other argument is allowed), then the
4812 indication of obsolescence applies only when compiling in Ada 2005
4813 mode. This is primarily intended for dealing with the situations
4814 in the predefined library where subprograms or packages
4815 have become defined as obsolescent in Ada 2005
4816 (e.g.@: in Ada.Characters.Handling), but may be used anywhere.
4818 The following examples show typical uses of this pragma:
4820 @smallexample @c ada
4822 pragma Obsolescent (p, Message => "use pp instead of p");
4827 pragma Obsolescent ("use q2new instead");
4829 type R is new integer;
4832 Message => "use RR in Ada 2005",
4842 type E is (a, bc, 'd', quack);
4843 pragma Obsolescent (Entity => bc)
4844 pragma Obsolescent (Entity => 'd')
4847 (a, b : character) return character;
4848 pragma Obsolescent (Entity => "+");
4853 Note that, as for all pragmas, if you use a pragma argument identifier,
4854 then all subsequent parameters must also use a pragma argument identifier.
4855 So if you specify "Entity =>" for the Entity argument, and a Message
4856 argument is present, it must be preceded by "Message =>".
4858 @node Pragma Optimize_Alignment
4859 @unnumberedsec Pragma Optimize_Alignment
4860 @findex Optimize_Alignment
4861 @cindex Alignment, default settings
4865 @smallexample @c ada
4866 pragma Optimize_Alignment (TIME | SPACE | OFF);
4870 This is a configuration pragma which affects the choice of default alignments
4871 for types and objects where no alignment is explicitly specified. There is a
4872 time/space trade-off in the selection of these values. Large alignments result
4873 in more efficient code, at the expense of larger data space, since sizes have
4874 to be increased to match these alignments. Smaller alignments save space, but
4875 the access code is slower. The normal choice of default alignments for types
4876 and individual alignment promotions for objects (which is what you get if you
4877 do not use this pragma, or if you use an argument of OFF), tries to balance
4878 these two requirements.
4880 Specifying SPACE causes smaller default alignments to be chosen in two cases.
4881 First any packed record is given an alignment of 1. Second, if a size is given
4882 for the type, then the alignment is chosen to avoid increasing this size. For
4885 @smallexample @c ada
4895 In the default mode, this type gets an alignment of 4, so that access to the
4896 Integer field X are efficient. But this means that objects of the type end up
4897 with a size of 8 bytes. This is a valid choice, since sizes of objects are
4898 allowed to be bigger than the size of the type, but it can waste space if for
4899 example fields of type R appear in an enclosing record. If the above type is
4900 compiled in @code{Optimize_Alignment (Space)} mode, the alignment is set to 1.
4902 However, there is one case in which SPACE is ignored. If a variable length
4903 record (that is a discriminated record with a component which is an array
4904 whose length depends on a discriminant), has a pragma Pack, then it is not
4905 in general possible to set the alignment of such a record to one, so the
4906 pragma is ignored in this case (with a warning).
4908 Specifying SPACE also disables alignment promotions for standalone objects,
4909 which occur when the compiler increases the alignment of a specific object
4910 without changing the alignment of its type.
4912 Specifying TIME causes larger default alignments to be chosen in the case of
4913 small types with sizes that are not a power of 2. For example, consider:
4915 @smallexample @c ada
4927 The default alignment for this record is normally 1, but if this type is
4928 compiled in @code{Optimize_Alignment (Time)} mode, then the alignment is set
4929 to 4, which wastes space for objects of the type, since they are now 4 bytes
4930 long, but results in more efficient access when the whole record is referenced.
4932 As noted above, this is a configuration pragma, and there is a requirement
4933 that all units in a partition be compiled with a consistent setting of the
4934 optimization setting. This would normally be achieved by use of a configuration
4935 pragma file containing the appropriate setting. The exception to this rule is
4936 that units with an explicit configuration pragma in the same file as the source
4937 unit are excluded from the consistency check, as are all predefined units. The
4938 latter are compiled by default in pragma Optimize_Alignment (Off) mode if no
4939 pragma appears at the start of the file.
4941 @node Pragma Ordered
4942 @unnumberedsec Pragma Ordered
4944 @findex pragma @code{Ordered}
4948 @smallexample @c ada
4949 pragma Ordered (enumeration_first_subtype_LOCAL_NAME);
4953 Most enumeration types are from a conceptual point of view unordered.
4954 For example, consider:
4956 @smallexample @c ada
4957 type Color is (Red, Blue, Green, Yellow);
4961 By Ada semantics @code{Blue > Red} and @code{Green > Blue},
4962 but really these relations make no sense; the enumeration type merely
4963 specifies a set of possible colors, and the order is unimportant.
4965 For unordered enumeration types, it is generally a good idea if
4966 clients avoid comparisons (other than equality or inequality) and
4967 explicit ranges. (A @emph{client} is a unit where the type is referenced,
4968 other than the unit where the type is declared, its body, and its subunits.)
4969 For example, if code buried in some client says:
4971 @smallexample @c ada
4972 if Current_Color < Yellow then ...
4973 if Current_Color in Blue .. Green then ...
4977 then the client code is relying on the order, which is undesirable.
4978 It makes the code hard to read and creates maintenance difficulties if
4979 entries have to be added to the enumeration type. Instead,
4980 the code in the client should list the possibilities, or an
4981 appropriate subtype should be declared in the unit that declares
4982 the original enumeration type. E.g., the following subtype could
4983 be declared along with the type @code{Color}:
4985 @smallexample @c ada
4986 subtype RBG is Color range Red .. Green;
4990 and then the client could write:
4992 @smallexample @c ada
4993 if Current_Color in RBG then ...
4994 if Current_Color = Blue or Current_Color = Green then ...
4998 However, some enumeration types are legitimately ordered from a conceptual
4999 point of view. For example, if you declare:
5001 @smallexample @c ada
5002 type Day is (Mon, Tue, Wed, Thu, Fri, Sat, Sun);
5006 then the ordering imposed by the language is reasonable, and
5007 clients can depend on it, writing for example:
5009 @smallexample @c ada
5010 if D in Mon .. Fri then ...
5015 The pragma @option{Ordered} is provided to mark enumeration types that
5016 are conceptually ordered, alerting the reader that clients may depend
5017 on the ordering. GNAT provides a pragma to mark enumerations as ordered
5018 rather than one to mark them as unordered, since in our experience,
5019 the great majority of enumeration types are conceptually unordered.
5021 The types @code{Boolean}, @code{Character}, @code{Wide_Character},
5022 and @code{Wide_Wide_Character}
5023 are considered to be ordered types, so each is declared with a
5024 pragma @code{Ordered} in package @code{Standard}.
5026 Normally pragma @code{Ordered} serves only as documentation and a guide for
5027 coding standards, but GNAT provides a warning switch @option{-gnatw.u} that
5028 requests warnings for inappropriate uses (comparisons and explicit
5029 subranges) for unordered types. If this switch is used, then any
5030 enumeration type not marked with pragma @code{Ordered} will be considered
5031 as unordered, and will generate warnings for inappropriate uses.
5033 For additional information please refer to the description of the
5034 @option{-gnatw.u} switch in the @value{EDITION} User's Guide.
5036 @node Pragma Overflow_Mode
5037 @unnumberedsec Pragma Overflow_Mode
5038 @findex Overflow checks
5039 @findex Overflow mode
5040 @findex pragma @code{Overflow_Mode}
5044 @smallexample @c ada
5045 pragma Overflow_Mode
5047 [,[Assertions =>] MODE]);
5049 MODE ::= STRICT | MINIMIZED | ELIMINATED
5053 This pragma sets the current overflow mode to the given setting. For details
5054 of the meaning of these modes, please refer to the
5055 ``Overflow Check Handling in GNAT'' appendix in the
5056 @value{EDITION} User's Guide. If only the @code{General} parameter is present,
5057 the given mode applies to all expressions. If both parameters are present,
5058 the @code{General} mode applies to expressions outside assertions, and
5059 the @code{Eliminated} mode applies to expressions within assertions.
5061 The case of the @code{MODE} parameter is ignored,
5062 so @code{MINIMIZED}, @code{Minimized} and
5063 @code{minimized} all have the same effect.
5065 The @code{Overflow_Mode} pragma has the same scoping and placement
5066 rules as pragma @code{Suppress}, so it can occur either as a
5067 configuration pragma, specifying a default for the whole
5068 program, or in a declarative scope, where it applies to the
5069 remaining declarations and statements in that scope.
5071 The pragma @code{Suppress (Overflow_Check)} suppresses
5072 overflow checking, but does not affect the overflow mode.
5074 The pragma @code{Unsuppress (Overflow_Check)} unsuppresses (enables)
5075 overflow checking, but does not affect the overflow mode.
5077 @node Pragma Overriding_Renamings
5078 @unnumberedsec Pragma Overriding_Renamings
5079 @findex Overriding_Renamings
5080 @cindex Rational profile
5081 @cindex Rational compatibility
5085 @smallexample @c ada
5086 pragma Overriding_Renamings;
5090 This is a GNAT configuration pragma to simplify porting
5091 legacy code accepted by the Rational
5092 Ada compiler. In the presence of this pragma, a renaming declaration that
5093 renames an inherited operation declared in the same scope is legal if selected
5094 notation is used as in:
5096 @smallexample @c ada
5097 pragma Overriding_Renamings;
5102 function F (..) renames R.F;
5107 RM 8.3 (15) stipulates that an overridden operation is not visible within the
5108 declaration of the overriding operation.
5110 @node Pragma Partition_Elaboration_Policy
5111 @unnumberedsec Pragma Partition_Elaboration_Policy
5112 @findex Partition_Elaboration_Policy
5116 @smallexample @c ada
5117 pragma Partition_Elaboration_Policy (POLICY_IDENTIFIER);
5119 POLICY_IDENTIFIER ::= Concurrent | Sequential
5123 This pragma is standard in Ada 2005, but is available in all earlier
5124 versions of Ada as an implementation-defined pragma.
5125 See Ada 2012 Reference Manual for details.
5127 @node Pragma Passive
5128 @unnumberedsec Pragma Passive
5133 @smallexample @c ada
5134 pragma Passive [(Semaphore | No)];
5138 Syntax checked, but otherwise ignored by GNAT@. This is recognized for
5139 compatibility with DEC Ada 83 implementations, where it is used within a
5140 task definition to request that a task be made passive. If the argument
5141 @code{Semaphore} is present, or the argument is omitted, then DEC Ada 83
5142 treats the pragma as an assertion that the containing task is passive
5143 and that optimization of context switch with this task is permitted and
5144 desired. If the argument @code{No} is present, the task must not be
5145 optimized. GNAT does not attempt to optimize any tasks in this manner
5146 (since protected objects are available in place of passive tasks).
5148 For more information on the subject of passive tasks, see the section
5149 ``Passive Task Optimization'' in the GNAT Users Guide.
5151 @node Pragma Persistent_BSS
5152 @unnumberedsec Pragma Persistent_BSS
5153 @findex Persistent_BSS
5157 @smallexample @c ada
5158 pragma Persistent_BSS [(LOCAL_NAME)]
5162 This pragma allows selected objects to be placed in the @code{.persistent_bss}
5163 section. On some targets the linker and loader provide for special
5164 treatment of this section, allowing a program to be reloaded without
5165 affecting the contents of this data (hence the name persistent).
5167 There are two forms of usage. If an argument is given, it must be the
5168 local name of a library level object, with no explicit initialization
5169 and whose type is potentially persistent. If no argument is given, then
5170 the pragma is a configuration pragma, and applies to all library level
5171 objects with no explicit initialization of potentially persistent types.
5173 A potentially persistent type is a scalar type, or a non-tagged,
5174 non-discriminated record, all of whose components have no explicit
5175 initialization and are themselves of a potentially persistent type,
5176 or an array, all of whose constraints are static, and whose component
5177 type is potentially persistent.
5179 If this pragma is used on a target where this feature is not supported,
5180 then the pragma will be ignored. See also @code{pragma Linker_Section}.
5182 @node Pragma Polling
5183 @unnumberedsec Pragma Polling
5188 @smallexample @c ada
5189 pragma Polling (ON | OFF);
5193 This pragma controls the generation of polling code. This is normally off.
5194 If @code{pragma Polling (ON)} is used then periodic calls are generated to
5195 the routine @code{Ada.Exceptions.Poll}. This routine is a separate unit in the
5196 runtime library, and can be found in file @file{a-excpol.adb}.
5198 Pragma @code{Polling} can appear as a configuration pragma (for example it
5199 can be placed in the @file{gnat.adc} file) to enable polling globally, or it
5200 can be used in the statement or declaration sequence to control polling
5203 A call to the polling routine is generated at the start of every loop and
5204 at the start of every subprogram call. This guarantees that the @code{Poll}
5205 routine is called frequently, and places an upper bound (determined by
5206 the complexity of the code) on the period between two @code{Poll} calls.
5208 The primary purpose of the polling interface is to enable asynchronous
5209 aborts on targets that cannot otherwise support it (for example Windows
5210 NT), but it may be used for any other purpose requiring periodic polling.
5211 The standard version is null, and can be replaced by a user program. This
5212 will require re-compilation of the @code{Ada.Exceptions} package that can
5213 be found in files @file{a-except.ads} and @file{a-except.adb}.
5215 A standard alternative unit (in file @file{4wexcpol.adb} in the standard GNAT
5216 distribution) is used to enable the asynchronous abort capability on
5217 targets that do not normally support the capability. The version of
5218 @code{Poll} in this file makes a call to the appropriate runtime routine
5219 to test for an abort condition.
5221 Note that polling can also be enabled by use of the @option{-gnatP} switch.
5222 @xref{Switches for gcc,,, gnat_ugn, @value{EDITION} User's Guide}, for
5226 @unnumberedsec Pragma Post
5228 @cindex Checks, postconditions
5229 @findex Postconditions
5233 @smallexample @c ada
5234 pragma Post (Boolean_Expression);
5238 The @code{Post} pragma is intended to be an exact replacement for
5239 the language-defined
5240 @code{Post} aspect, and shares its restrictions and semantics.
5241 It must appear either immediately following the corresponding
5242 subprogram declaration (only other pragmas may intervene), or
5243 if there is no separate subprogram declaration, then it can
5244 appear at the start of the declarations in a subprogram body
5245 (preceded only by other pragmas).
5247 @node Pragma Postcondition
5248 @unnumberedsec Pragma Postcondition
5249 @cindex Postcondition
5250 @cindex Checks, postconditions
5251 @findex Postconditions
5255 @smallexample @c ada
5256 pragma Postcondition (
5257 [Check =>] Boolean_Expression
5258 [,[Message =>] String_Expression]);
5262 The @code{Postcondition} pragma allows specification of automatic
5263 postcondition checks for subprograms. These checks are similar to
5264 assertions, but are automatically inserted just prior to the return
5265 statements of the subprogram with which they are associated (including
5266 implicit returns at the end of procedure bodies and associated
5267 exception handlers).
5269 In addition, the boolean expression which is the condition which
5270 must be true may contain references to function'Result in the case
5271 of a function to refer to the returned value.
5273 @code{Postcondition} pragmas may appear either immediately following the
5274 (separate) declaration of a subprogram, or at the start of the
5275 declarations of a subprogram body. Only other pragmas may intervene
5276 (that is appear between the subprogram declaration and its
5277 postconditions, or appear before the postcondition in the
5278 declaration sequence in a subprogram body). In the case of a
5279 postcondition appearing after a subprogram declaration, the
5280 formal arguments of the subprogram are visible, and can be
5281 referenced in the postcondition expressions.
5283 The postconditions are collected and automatically tested just
5284 before any return (implicit or explicit) in the subprogram body.
5285 A postcondition is only recognized if postconditions are active
5286 at the time the pragma is encountered. The compiler switch @option{gnata}
5287 turns on all postconditions by default, and pragma @code{Check_Policy}
5288 with an identifier of @code{Postcondition} can also be used to
5289 control whether postconditions are active.
5291 The general approach is that postconditions are placed in the spec
5292 if they represent functional aspects which make sense to the client.
5293 For example we might have:
5295 @smallexample @c ada
5296 function Direction return Integer;
5297 pragma Postcondition
5298 (Direction'Result = +1
5300 Direction'Result = -1);
5304 which serves to document that the result must be +1 or -1, and
5305 will test that this is the case at run time if postcondition
5308 Postconditions within the subprogram body can be used to
5309 check that some internal aspect of the implementation,
5310 not visible to the client, is operating as expected.
5311 For instance if a square root routine keeps an internal
5312 counter of the number of times it is called, then we
5313 might have the following postcondition:
5315 @smallexample @c ada
5316 Sqrt_Calls : Natural := 0;
5318 function Sqrt (Arg : Float) return Float is
5319 pragma Postcondition
5320 (Sqrt_Calls = Sqrt_Calls'Old + 1);
5326 As this example, shows, the use of the @code{Old} attribute
5327 is often useful in postconditions to refer to the state on
5328 entry to the subprogram.
5330 Note that postconditions are only checked on normal returns
5331 from the subprogram. If an abnormal return results from
5332 raising an exception, then the postconditions are not checked.
5334 If a postcondition fails, then the exception
5335 @code{System.Assertions.Assert_Failure} is raised. If
5336 a message argument was supplied, then the given string
5337 will be used as the exception message. If no message
5338 argument was supplied, then the default message has
5339 the form "Postcondition failed at file:line". The
5340 exception is raised in the context of the subprogram
5341 body, so it is possible to catch postcondition failures
5342 within the subprogram body itself.
5344 Within a package spec, normal visibility rules
5345 in Ada would prevent forward references within a
5346 postcondition pragma to functions defined later in
5347 the same package. This would introduce undesirable
5348 ordering constraints. To avoid this problem, all
5349 postcondition pragmas are analyzed at the end of
5350 the package spec, allowing forward references.
5352 The following example shows that this even allows
5353 mutually recursive postconditions as in:
5355 @smallexample @c ada
5356 package Parity_Functions is
5357 function Odd (X : Natural) return Boolean;
5358 pragma Postcondition
5362 (x /= 0 and then Even (X - 1))));
5364 function Even (X : Natural) return Boolean;
5365 pragma Postcondition
5369 (x /= 1 and then Odd (X - 1))));
5371 end Parity_Functions;
5375 There are no restrictions on the complexity or form of
5376 conditions used within @code{Postcondition} pragmas.
5377 The following example shows that it is even possible
5378 to verify performance behavior.
5380 @smallexample @c ada
5383 Performance : constant Float;
5384 -- Performance constant set by implementation
5385 -- to match target architecture behavior.
5387 procedure Treesort (Arg : String);
5388 -- Sorts characters of argument using N*logN sort
5389 pragma Postcondition
5390 (Float (Clock - Clock'Old) <=
5391 Float (Arg'Length) *
5392 log (Float (Arg'Length)) *
5398 Note: postcondition pragmas associated with subprograms that are
5399 marked as Inline_Always, or those marked as Inline with front-end
5400 inlining (-gnatN option set) are accepted and legality-checked
5401 by the compiler, but are ignored at run-time even if postcondition
5402 checking is enabled.
5404 Note that pragma @code{Postcondition} differs from the language-defined
5405 @code{Post} aspect (and corresponding @code{Post} pragma) in allowing
5406 multiple occurrences, allowing occurences in the body even if there
5407 is a separate spec, and allowing a second string parameter, and the
5408 use of the pragma identifier @code{Check}. Historically, pragma
5409 @code{Postcondition} was implemented prior to the development of
5410 Ada 2012, and has been retained in its original form for
5411 compatibility purposes.
5413 @node Pragma Post_Class
5414 @unnumberedsec Pragma Post_Class
5416 @cindex Checks, postconditions
5417 @findex Postconditions
5421 @smallexample @c ada
5422 pragma Post_Class (Boolean_Expression);
5426 The @code{Post_Class} pragma is intended to be an exact replacement for
5427 the language-defined
5428 @code{Post'Class} aspect, and shares its restrictions and semantics.
5429 It must appear either immediately following the corresponding
5430 subprogram declaration (only other pragmas may intervene), or
5431 if there is no separate subprogram declaration, then it can
5432 appear at the start of the declarations in a subprogram body
5433 (preceded only by other pragmas).
5435 Note: This pragma is called @code{Post_Class} rather than
5436 @code{Post'Class} because the latter would not be strictly
5437 conforming to the allowed syntax for pragmas. The motivation
5438 for provinding pragmas equivalent to the aspects is to allow a program
5439 to be written using the pragmas, and then compiled if necessary
5440 using an Ada compiler that does not recognize the pragmas or
5441 aspects, but is prepared to ignore the pragmas. The assertion
5442 policy that controls this pragma is @code{Post'Class}, not
5446 @unnumberedsec Pragma Pre
5448 @cindex Checks, preconditions
5449 @findex Preconditions
5453 @smallexample @c ada
5454 pragma Pre (Boolean_Expression);
5458 The @code{Pre} pragma is intended to be an exact replacement for
5459 the language-defined
5460 @code{Pre} aspect, and shares its restrictions and semantics.
5461 It must appear either immediately following the corresponding
5462 subprogram declaration (only other pragmas may intervene), or
5463 if there is no separate subprogram declaration, then it can
5464 appear at the start of the declarations in a subprogram body
5465 (preceded only by other pragmas).
5467 @node Pragma Precondition
5468 @unnumberedsec Pragma Precondition
5469 @cindex Preconditions
5470 @cindex Checks, preconditions
5471 @findex Preconditions
5475 @smallexample @c ada
5476 pragma Precondition (
5477 [Check =>] Boolean_Expression
5478 [,[Message =>] String_Expression]);
5482 The @code{Precondition} pragma is similar to @code{Postcondition}
5483 except that the corresponding checks take place immediately upon
5484 entry to the subprogram, and if a precondition fails, the exception
5485 is raised in the context of the caller, and the attribute 'Result
5486 cannot be used within the precondition expression.
5488 Otherwise, the placement and visibility rules are identical to those
5489 described for postconditions. The following is an example of use
5490 within a package spec:
5492 @smallexample @c ada
5493 package Math_Functions is
5495 function Sqrt (Arg : Float) return Float;
5496 pragma Precondition (Arg >= 0.0)
5502 @code{Precondition} pragmas may appear either immediately following the
5503 (separate) declaration of a subprogram, or at the start of the
5504 declarations of a subprogram body. Only other pragmas may intervene
5505 (that is appear between the subprogram declaration and its
5506 postconditions, or appear before the postcondition in the
5507 declaration sequence in a subprogram body).
5509 Note: precondition pragmas associated with subprograms that are
5510 marked as Inline_Always, or those marked as Inline with front-end
5511 inlining (-gnatN option set) are accepted and legality-checked
5512 by the compiler, but are ignored at run-time even if precondition
5513 checking is enabled.
5515 Note that pragma @code{Precondition} differs from the language-defined
5516 @code{Pre} aspect (and corresponding @code{Pre} pragma) in allowing
5517 multiple occurrences, allowing occurences in the body even if there
5518 is a separate spec, and allowing a second string parameter, and the
5519 use of the pragma identifier @code{Check}. Historically, pragma
5520 @code{Precondition} was implemented prior to the development of
5521 Ada 2012, and has been retained in its original form for
5522 compatibility purposes.
5524 @node Pragma Predicate
5525 @unnumberedsec Pragma Predicate
5527 @findex Predicate pragma
5531 @smallexample @c ada
5533 ([Entity =>] type_LOCAL_NAME,
5534 [Check =>] EXPRESSION);
5538 This pragma (available in all versions of Ada in GNAT) encompasses both
5539 the @code{Static_Predicate} and @code{Dynamic_Predicate} aspects in
5540 Ada 2012. A predicate is regarded as static if it has an allowed form
5541 for @code{Static_Predicate} and is otherwise treated as a
5542 @code{Dynamic_Predicate}. Otherwise, predicates specified by this
5543 pragma behave exactly as described in the Ada 2012 reference manual.
5544 For example, if we have
5546 @smallexample @c ada
5547 type R is range 1 .. 10;
5549 pragma Predicate (Entity => S, Check => S not in 4 .. 6);
5551 pragma Predicate (Entity => Q, Check => F(Q) or G(Q));
5555 the effect is identical to the following Ada 2012 code:
5557 @smallexample @c ada
5558 type R is range 1 .. 10;
5560 Static_Predicate => S not in 4 .. 6;
5562 Dynamic_Predicate => F(Q) or G(Q);
5565 Note that there is are no pragmas @code{Dynamic_Predicate}
5566 or @code{Static_Predicate}. That is
5567 because these pragmas would affect legality and semantics of
5568 the program and thus do not have a neutral effect if ignored.
5569 The motivation behind providing pragmas equivalent to
5570 corresponding aspects is to allow a program to be written
5571 using the pragmas, and then compiled with a compiler that
5572 will ignore the pragmas. That doesn't work in the case of
5573 static and dynamic predicates, since if the corresponding
5574 pragmas are ignored, then the behavior of the program is
5575 fundamentally changed (for example a membership test
5576 @code{A in B} would not take into account a predicate
5577 defined for subtype B). When following this approach, the
5578 use of predicates should be avoided.
5580 @node Pragma Preelaborable_Initialization
5581 @unnumberedsec Pragma Preelaborable_Initialization
5582 @findex Preelaborable_Initialization
5586 @smallexample @c ada
5587 pragma Preelaborable_Initialization (DIRECT_NAME);
5591 This pragma is standard in Ada 2005, but is available in all earlier
5592 versions of Ada as an implementation-defined pragma.
5593 See Ada 2012 Reference Manual for details.
5595 @node Pragma Preelaborate_05
5596 @unnumberedsec Pragma Preelaborate_05
5597 @findex Preelaborate_05
5601 @smallexample @c ada
5602 pragma Preelaborate_05 [(library_unit_NAME)];
5606 This pragma is only available in GNAT mode (@option{-gnatg} switch set)
5607 and is intended for use in the standard run-time library only. It has
5608 no effect in Ada 83 or Ada 95 mode, but is
5609 equivalent to @code{pragma Prelaborate} when operating in later
5610 Ada versions. This is used to handle some cases where packages
5611 not previously preelaborable became so in Ada 2005.
5613 @node Pragma Pre_Class
5614 @unnumberedsec Pragma Pre_Class
5616 @cindex Checks, preconditions
5617 @findex Preconditions
5621 @smallexample @c ada
5622 pragma Pre_Class (Boolean_Expression);
5626 The @code{Pre_Class} pragma is intended to be an exact replacement for
5627 the language-defined
5628 @code{Pre'Class} aspect, and shares its restrictions and semantics.
5629 It must appear either immediately following the corresponding
5630 subprogram declaration (only other pragmas may intervene), or
5631 if there is no separate subprogram declaration, then it can
5632 appear at the start of the declarations in a subprogram body
5633 (preceded only by other pragmas).
5635 Note: This pragma is called @code{Pre_Class} rather than
5636 @code{Pre'Class} because the latter would not be strictly
5637 conforming to the allowed syntax for pragmas. The motivation
5638 for providing pragmas equivalent to the aspects is to allow a program
5639 to be written using the pragmas, and then compiled if necessary
5640 using an Ada compiler that does not recognize the pragmas or
5641 aspects, but is prepared to ignore the pragmas. The assertion
5642 policy that controls this pragma is @code{Pre'Class}, not
5645 @node Pragma Priority_Specific_Dispatching
5646 @unnumberedsec Pragma Priority_Specific_Dispatching
5647 @findex Priority_Specific_Dispatching
5651 @smallexample @c ada
5652 pragma Priority_Specific_Dispatching (
5654 first_priority_EXPRESSION,
5655 last_priority_EXPRESSION)
5657 POLICY_IDENTIFIER ::=
5658 EDF_Across_Priorities |
5659 FIFO_Within_Priorities |
5660 Non_Preemptive_Within_Priorities |
5661 Round_Robin_Within_Priorities
5665 This pragma is standard in Ada 2005, but is available in all earlier
5666 versions of Ada as an implementation-defined pragma.
5667 See Ada 2012 Reference Manual for details.
5669 @node Pragma Profile
5670 @unnumberedsec Pragma Profile
5675 @smallexample @c ada
5676 pragma Profile (Ravenscar | Restricted | Rational);
5680 This pragma is standard in Ada 2005, but is available in all earlier
5681 versions of Ada as an implementation-defined pragma. This is a
5682 configuration pragma that establishes a set of configiuration pragmas
5683 that depend on the argument. @code{Ravenscar} is standard in Ada 2005.
5684 The other two possibilities (@code{Restricted} or @code{Rational})
5685 are implementation-defined. The set of configuration pragmas
5686 is defined in the following sections.
5690 @item Pragma Profile (Ravenscar)
5694 The @code{Ravenscar} profile is standard in Ada 2005,
5695 but is available in all earlier
5696 versions of Ada as an implementation-defined pragma. This profile
5697 establishes the following set of configuration pragmas:
5700 @item Task_Dispatching_Policy (FIFO_Within_Priorities)
5701 [RM D.2.2] Tasks are dispatched following a preemptive
5702 priority-ordered scheduling policy.
5704 @item Locking_Policy (Ceiling_Locking)
5705 [RM D.3] While tasks and interrupts execute a protected action, they inherit
5706 the ceiling priority of the corresponding protected object.
5708 @item Detect_Blocking
5709 This pragma forces the detection of potentially blocking operations within a
5710 protected operation, and to raise Program_Error if that happens.
5714 plus the following set of restrictions:
5717 @item Max_Entry_Queue_Length => 1
5718 No task can be queued on a protected entry.
5719 @item Max_Protected_Entries => 1
5720 @item Max_Task_Entries => 0
5721 No rendezvous statements are allowed.
5722 @item No_Abort_Statements
5723 @item No_Dynamic_Attachment
5724 @item No_Dynamic_Priorities
5725 @item No_Implicit_Heap_Allocations
5726 @item No_Local_Protected_Objects
5727 @item No_Local_Timing_Events
5728 @item No_Protected_Type_Allocators
5729 @item No_Relative_Delay
5730 @item No_Requeue_Statements
5731 @item No_Select_Statements
5732 @item No_Specific_Termination_Handlers
5733 @item No_Task_Allocators
5734 @item No_Task_Hierarchy
5735 @item No_Task_Termination
5736 @item Simple_Barriers
5740 The Ravenscar profile also includes the following restrictions that specify
5741 that there are no semantic dependences on the corresponding predefined
5745 @item No_Dependence => Ada.Asynchronous_Task_Control
5746 @item No_Dependence => Ada.Calendar
5747 @item No_Dependence => Ada.Execution_Time.Group_Budget
5748 @item No_Dependence => Ada.Execution_Time.Timers
5749 @item No_Dependence => Ada.Task_Attributes
5750 @item No_Dependence => System.Multiprocessors.Dispatching_Domains
5755 This set of configuration pragmas and restrictions correspond to the
5756 definition of the ``Ravenscar Profile'' for limited tasking, devised and
5757 published by the @cite{International Real-Time Ada Workshop}, 1997,
5758 and whose most recent description is available at
5759 @url{http://www-users.cs.york.ac.uk/~burns/ravenscar.ps}.
5761 The original definition of the profile was revised at subsequent IRTAW
5762 meetings. It has been included in the ISO
5763 @cite{Guide for the Use of the Ada Programming Language in High
5764 Integrity Systems}, and has been approved by ISO/IEC/SC22/WG9 for inclusion in
5765 the next revision of the standard. The formal definition given by
5766 the Ada Rapporteur Group (ARG) can be found in two Ada Issues (AI-249 and
5767 AI-305) available at
5768 @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/ais/ai-00249.txt} and
5769 @url{http://www.ada-auth.org/cgi-bin/cvsweb.cgi/ais/ai-00305.txt}.
5771 The above set is a superset of the restrictions provided by pragma
5772 @code{Profile (Restricted)}, it includes six additional restrictions
5773 (@code{Simple_Barriers}, @code{No_Select_Statements},
5774 @code{No_Calendar}, @code{No_Implicit_Heap_Allocations},
5775 @code{No_Relative_Delay} and @code{No_Task_Termination}). This means
5776 that pragma @code{Profile (Ravenscar)}, like the pragma
5777 @code{Profile (Restricted)},
5778 automatically causes the use of a simplified,
5779 more efficient version of the tasking run-time system.
5781 @item Pragma Profile (Restricted)
5782 @findex Restricted Run Time
5784 This profile corresponds to the GNAT restricted run time. It
5785 establishes the following set of restrictions:
5788 @item No_Abort_Statements
5789 @item No_Entry_Queue
5790 @item No_Task_Hierarchy
5791 @item No_Task_Allocators
5792 @item No_Dynamic_Priorities
5793 @item No_Terminate_Alternatives
5794 @item No_Dynamic_Attachment
5795 @item No_Protected_Type_Allocators
5796 @item No_Local_Protected_Objects
5797 @item No_Requeue_Statements
5798 @item No_Task_Attributes_Package
5799 @item Max_Asynchronous_Select_Nesting = 0
5800 @item Max_Task_Entries = 0
5801 @item Max_Protected_Entries = 1
5802 @item Max_Select_Alternatives = 0
5806 This set of restrictions causes the automatic selection of a simplified
5807 version of the run time that provides improved performance for the
5808 limited set of tasking functionality permitted by this set of restrictions.
5810 @item Pragma Profile (Rational)
5811 @findex Rational compatibility mode
5813 The Rational profile is intended to facilitate porting legacy code that
5814 compiles with the Rational APEX compiler, even when the code includes non-
5815 conforming Ada constructs. The profile enables the following three pragmas:
5818 @item pragma Implicit_Packing
5819 @item pragma Overriding_Renamings
5820 @item pragma Use_VADS_Size
5825 @node Pragma Profile_Warnings
5826 @unnumberedsec Pragma Profile_Warnings
5827 @findex Profile_Warnings
5831 @smallexample @c ada
5832 pragma Profile_Warnings (Ravenscar | Restricted | Rational);
5836 This is an implementation-defined pragma that is similar in
5837 effect to @code{pragma Profile} except that instead of
5838 generating @code{Restrictions} pragmas, it generates
5839 @code{Restriction_Warnings} pragmas. The result is that
5840 violations of the profile generate warning messages instead
5843 @node Pragma Propagate_Exceptions
5844 @unnumberedsec Pragma Propagate_Exceptions
5845 @cindex Interfacing to C++
5846 @findex Propagate_Exceptions
5850 @smallexample @c ada
5851 pragma Propagate_Exceptions;
5855 This pragma is now obsolete and, other than generating a warning if warnings
5856 on obsolescent features are enabled, is ignored.
5857 It is retained for compatibility
5858 purposes. It used to be used in connection with optimization of
5859 a now-obsolete mechanism for implementation of exceptions.
5861 @node Pragma Provide_Shift_Operators
5862 @unnumberedsec Pragma Provide_Shift_Operators
5863 @cindex Shift operators
5864 @findex Provide_Shift_Operators
5868 @smallexample @c ada
5869 pragma Provide_Shift_Operators (integer_first_subtype_LOCAL_NAME);
5873 This pragma can be applied to a first subtype local name that specifies
5874 either an unsigned or signed type. It has the effect of providing the
5875 five shift operators (Shift_Left, Shift_Right, Shift_Right_Arithmetic,
5876 Rotate_Left and Rotate_Right) for the given type. It is equivalent to
5877 including the function declarations for these five operators, together
5878 with the pragma Import (Intrinsic, ...) statements.
5880 @node Pragma Psect_Object
5881 @unnumberedsec Pragma Psect_Object
5882 @findex Psect_Object
5886 @smallexample @c ada
5887 pragma Psect_Object (
5888 [Internal =>] LOCAL_NAME,
5889 [, [External =>] EXTERNAL_SYMBOL]
5890 [, [Size =>] EXTERNAL_SYMBOL]);
5894 | static_string_EXPRESSION
5898 This pragma is identical in effect to pragma @code{Common_Object}.
5900 @node Pragma Pure_05
5901 @unnumberedsec Pragma Pure_05
5906 @smallexample @c ada
5907 pragma Pure_05 [(library_unit_NAME)];
5911 This pragma is only available in GNAT mode (@option{-gnatg} switch set)
5912 and is intended for use in the standard run-time library only. It has
5913 no effect in Ada 83 or Ada 95 mode, but is
5914 equivalent to @code{pragma Pure} when operating in later
5915 Ada versions. This is used to handle some cases where packages
5916 not previously pure became so in Ada 2005.
5918 @node Pragma Pure_12
5919 @unnumberedsec Pragma Pure_12
5924 @smallexample @c ada
5925 pragma Pure_12 [(library_unit_NAME)];
5929 This pragma is only available in GNAT mode (@option{-gnatg} switch set)
5930 and is intended for use in the standard run-time library only. It has
5931 no effect in Ada 83, Ada 95, or Ada 2005 modes, but is
5932 equivalent to @code{pragma Pure} when operating in later
5933 Ada versions. This is used to handle some cases where packages
5934 not previously pure became so in Ada 2012.
5936 @node Pragma Pure_Function
5937 @unnumberedsec Pragma Pure_Function
5938 @findex Pure_Function
5942 @smallexample @c ada
5943 pragma Pure_Function ([Entity =>] function_LOCAL_NAME);
5947 This pragma appears in the same declarative part as a function
5948 declaration (or a set of function declarations if more than one
5949 overloaded declaration exists, in which case the pragma applies
5950 to all entities). It specifies that the function @code{Entity} is
5951 to be considered pure for the purposes of code generation. This means
5952 that the compiler can assume that there are no side effects, and
5953 in particular that two calls with identical arguments produce the
5954 same result. It also means that the function can be used in an
5957 Note that, quite deliberately, there are no static checks to try
5958 to ensure that this promise is met, so @code{Pure_Function} can be used
5959 with functions that are conceptually pure, even if they do modify
5960 global variables. For example, a square root function that is
5961 instrumented to count the number of times it is called is still
5962 conceptually pure, and can still be optimized, even though it
5963 modifies a global variable (the count). Memo functions are another
5964 example (where a table of previous calls is kept and consulted to
5965 avoid re-computation).
5967 Note also that the normal rules excluding optimization of subprograms
5968 in pure units (when parameter types are descended from System.Address,
5969 or when the full view of a parameter type is limited), do not apply
5970 for the Pure_Function case. If you explicitly specify Pure_Function,
5971 the compiler may optimize away calls with identical arguments, and
5972 if that results in unexpected behavior, the proper action is not to
5973 use the pragma for subprograms that are not (conceptually) pure.
5976 Note: Most functions in a @code{Pure} package are automatically pure, and
5977 there is no need to use pragma @code{Pure_Function} for such functions. One
5978 exception is any function that has at least one formal of type
5979 @code{System.Address} or a type derived from it. Such functions are not
5980 considered pure by default, since the compiler assumes that the
5981 @code{Address} parameter may be functioning as a pointer and that the
5982 referenced data may change even if the address value does not.
5983 Similarly, imported functions are not considered to be pure by default,
5984 since there is no way of checking that they are in fact pure. The use
5985 of pragma @code{Pure_Function} for such a function will override these default
5986 assumption, and cause the compiler to treat a designated subprogram as pure
5989 Note: If pragma @code{Pure_Function} is applied to a renamed function, it
5990 applies to the underlying renamed function. This can be used to
5991 disambiguate cases of overloading where some but not all functions
5992 in a set of overloaded functions are to be designated as pure.
5994 If pragma @code{Pure_Function} is applied to a library level function, the
5995 function is also considered pure from an optimization point of view, but the
5996 unit is not a Pure unit in the categorization sense. So for example, a function
5997 thus marked is free to @code{with} non-pure units.
5999 @node Pragma Ravenscar
6000 @unnumberedsec Pragma Ravenscar
6001 @findex Pragma Ravenscar
6005 @smallexample @c ada
6010 This pragma is considered obsolescent, but is retained for
6011 compatibility purposes. It is equivalent to:
6013 @smallexample @c ada
6014 pragma Profile (Ravenscar);
6018 which is the preferred method of setting the @code{Ravenscar} profile.
6020 @node Pragma Refined_State
6021 @unnumberedsec Pragma Refined_State
6022 @findex Refined_State
6024 For the description of this pragma, see SPARK 2014 Reference Manual,
6027 @node Pragma Relative_Deadline
6028 @unnumberedsec Pragma Relative_Deadline
6029 @findex Relative_Deadline
6033 @smallexample @c ada
6034 pragma Relative_Deadline (time_span_EXPRESSION);
6038 This pragma is standard in Ada 2005, but is available in all earlier
6039 versions of Ada as an implementation-defined pragma.
6040 See Ada 2012 Reference Manual for details.
6042 @node Pragma Remote_Access_Type
6043 @unnumberedsec Pragma Remote_Access_Type
6044 @findex Remote_Access_Type
6048 @smallexample @c ada
6049 pragma Remote_Access_Type ([Entity =>] formal_access_type_LOCAL_NAME);
6053 This pragma appears in the formal part of a generic declaration.
6054 It specifies an exception to the RM rule from E.2.2(17/2), which forbids
6055 the use of a remote access to class-wide type as actual for a formal
6058 When this pragma applies to a formal access type @code{Entity}, that
6059 type is treated as a remote access to class-wide type in the generic.
6060 It must be a formal general access type, and its designated type must
6061 be the class-wide type of a formal tagged limited private type from the
6062 same generic declaration.
6064 In the generic unit, the formal type is subject to all restrictions
6065 pertaining to remote access to class-wide types. At instantiation, the
6066 actual type must be a remote access to class-wide type.
6068 @node Pragma Restricted_Run_Time
6069 @unnumberedsec Pragma Restricted_Run_Time
6070 @findex Pragma Restricted_Run_Time
6074 @smallexample @c ada
6075 pragma Restricted_Run_Time;
6079 This pragma is considered obsolescent, but is retained for
6080 compatibility purposes. It is equivalent to:
6082 @smallexample @c ada
6083 pragma Profile (Restricted);
6087 which is the preferred method of setting the restricted run time
6090 @node Pragma Restriction_Warnings
6091 @unnumberedsec Pragma Restriction_Warnings
6092 @findex Restriction_Warnings
6096 @smallexample @c ada
6097 pragma Restriction_Warnings
6098 (restriction_IDENTIFIER @{, restriction_IDENTIFIER@});
6102 This pragma allows a series of restriction identifiers to be
6103 specified (the list of allowed identifiers is the same as for
6104 pragma @code{Restrictions}). For each of these identifiers
6105 the compiler checks for violations of the restriction, but
6106 generates a warning message rather than an error message
6107 if the restriction is violated.
6109 One use of this is in situations where you want to know
6110 about violations of a restriction, but you want to ignore some of
6111 these violations. Consider this example, where you want to set
6112 Ada_95 mode and enable style checks, but you want to know about
6113 any other use of implementation pragmas:
6115 @smallexample @c ada
6116 pragma Restriction_Warnings (No_Implementation_Pragmas);
6117 pragma Warnings (Off, "violation of*No_Implementation_Pragmas*");
6119 pragma Style_Checks ("2bfhkM160");
6120 pragma Warnings (On, "violation of*No_Implementation_Pragmas*");
6124 By including the above lines in a configuration pragmas file,
6125 the Ada_95 and Style_Checks pragmas are accepted without
6126 generating a warning, but any other use of implementation
6127 defined pragmas will cause a warning to be generated.
6129 @node Pragma Reviewable
6130 @unnumberedsec Pragma Reviewable
6135 @smallexample @c ada
6140 This pragma is an RM-defined standard pragma, but has no effect on the
6141 program being compiled, or on the code generated for the program.
6143 To obtain the required output specified in RM H.3.1, the compiler must be
6144 run with various special switches as follows:
6148 @item Where compiler-generated run-time checks remain
6150 The switch @option{-gnatGL}
6151 @findex @option{-gnatGL}
6152 may be used to list the expanded code in pseudo-Ada form.
6153 Runtime checks show up in the listing either as explicit
6154 checks or operators marked with @{@} to indicate a check is present.
6156 @item An identification of known exceptions at compile time
6158 If the program is compiled with @option{-gnatwa},
6159 @findex @option{-gnatwa}
6160 the compiler warning messages will indicate all cases where the compiler
6161 detects that an exception is certain to occur at run time.
6163 @item Possible reads of uninitialized variables
6165 The compiler warns of many such cases, but its output is incomplete.
6167 The CodePeer analysis tool
6168 @findex CodePeer static analysis tool
6171 A supplemental static analysis tool
6173 may be used to obtain a comprehensive list of all
6174 possible points at which uninitialized data may be read.
6176 @item Where run-time support routines are implicitly invoked
6178 In the output from @option{-gnatGL},
6179 @findex @option{-gnatGL}
6180 run-time calls are explicitly listed as calls to the relevant
6183 @item Object code listing
6185 This may be obtained either by using the @option{-S} switch,
6187 or the objdump utility.
6190 @item Constructs known to be erroneous at compile time
6192 These are identified by warnings issued by the compiler (use @option{-gnatwa}).
6193 @findex @option{-gnatwa}
6195 @item Stack usage information
6197 Static stack usage data (maximum per-subprogram) can be obtained via the
6198 @option{-fstack-usage} switch to the compiler.
6199 @findex @option{-fstack-usage}
6200 Dynamic stack usage data (per task) can be obtained via the @option{-u} switch
6204 The gnatstack utility
6206 can be used to provide additional information on stack usage.
6209 @item Object code listing of entire partition
6211 This can be obtained by compiling the partition with @option{-S},
6213 or by applying objdump
6215 to all the object files that are part of the partition.
6217 @item A description of the run-time model
6219 The full sources of the run-time are available, and the documentation of
6220 these routines describes how these run-time routines interface to the
6221 underlying operating system facilities.
6223 @item Control and data-flow information
6227 @findex CodePeer static analysis tool
6230 A supplemental static analysis tool
6232 may be used to obtain complete control and data-flow information, as well as
6233 comprehensive messages identifying possible problems based on this
6237 @node Pragma Share_Generic
6238 @unnumberedsec Pragma Share_Generic
6239 @findex Share_Generic
6243 @smallexample @c ada
6244 pragma Share_Generic (GNAME @{, GNAME@});
6246 GNAME ::= generic_unit_NAME | generic_instance_NAME
6250 This pragma is provided for compatibility with Dec Ada 83. It has
6251 no effect in @code{GNAT} (which does not implement shared generics), other
6252 than to check that the given names are all names of generic units or
6256 @unnumberedsec Pragma Shared
6260 This pragma is provided for compatibility with Ada 83. The syntax and
6261 semantics are identical to pragma Atomic.
6263 @node Pragma Short_Circuit_And_Or
6264 @unnumberedsec Pragma Short_Circuit_And_Or
6265 @findex Short_Circuit_And_Or
6269 @smallexample @c ada
6270 pragma Short_Circuit_And_Or;
6274 This configuration pragma causes any occurrence of the AND operator applied to
6275 operands of type Standard.Boolean to be short-circuited (i.e. the AND operator
6276 is treated as if it were AND THEN). Or is similarly treated as OR ELSE. This
6277 may be useful in the context of certification protocols requiring the use of
6278 short-circuited logical operators. If this configuration pragma occurs locally
6279 within the file being compiled, it applies only to the file being compiled.
6280 There is no requirement that all units in a partition use this option.
6282 @node Pragma Short_Descriptors
6283 @unnumberedsec Pragma Short_Descriptors
6284 @findex Short_Descriptors
6288 @smallexample @c ada
6289 pragma Short_Descriptors
6293 In VMS versions of the compiler, this configuration pragma causes all
6294 occurrences of the mechanism types Descriptor[_xxx] to be treated as
6295 Short_Descriptor[_xxx]. This is helpful in porting legacy applications from a
6296 32-bit environment to a 64-bit environment. This pragma is ignored for non-VMS
6299 @node Pragma Simple_Storage_Pool_Type
6300 @unnumberedsec Pragma Simple_Storage_Pool_Type
6301 @findex Simple_Storage_Pool_Type
6302 @cindex Storage pool, simple
6303 @cindex Simple storage pool
6307 @smallexample @c ada
6308 pragma Simple_Storage_Pool_Type (type_LOCAL_NAME);
6312 A type can be established as a ``simple storage pool type'' by applying
6313 the representation pragma @code{Simple_Storage_Pool_Type} to the type.
6314 A type named in the pragma must be a library-level immutably limited record
6315 type or limited tagged type declared immediately within a package declaration.
6316 The type can also be a limited private type whose full type is allowed as
6317 a simple storage pool type.
6319 For a simple storage pool type @var{SSP}, nonabstract primitive subprograms
6320 @code{Allocate}, @code{Deallocate}, and @code{Storage_Size} can be declared that
6321 are subtype conformant with the following subprogram declarations:
6323 @smallexample @c ada
6326 Storage_Address : out System.Address;
6327 Size_In_Storage_Elements : System.Storage_Elements.Storage_Count;
6328 Alignment : System.Storage_Elements.Storage_Count);
6330 procedure Deallocate
6332 Storage_Address : System.Address;
6333 Size_In_Storage_Elements : System.Storage_Elements.Storage_Count;
6334 Alignment : System.Storage_Elements.Storage_Count);
6336 function Storage_Size (Pool : SSP)
6337 return System.Storage_Elements.Storage_Count;
6341 Procedure @code{Allocate} must be declared, whereas @code{Deallocate} and
6342 @code{Storage_Size} are optional. If @code{Deallocate} is not declared, then
6343 applying an unchecked deallocation has no effect other than to set its actual
6344 parameter to null. If @code{Storage_Size} is not declared, then the
6345 @code{Storage_Size} attribute applied to an access type associated with
6346 a pool object of type SSP returns zero. Additional operations can be declared
6347 for a simple storage pool type (such as for supporting a mark/release
6348 storage-management discipline).
6350 An object of a simple storage pool type can be associated with an access
6351 type by specifying the attribute @code{Simple_Storage_Pool}. For example:
6353 @smallexample @c ada
6355 My_Pool : My_Simple_Storage_Pool_Type;
6357 type Acc is access My_Data_Type;
6359 for Acc'Simple_Storage_Pool use My_Pool;
6364 See attribute @code{Simple_Storage_Pool} for further details.
6366 @node Pragma Source_File_Name
6367 @unnumberedsec Pragma Source_File_Name
6368 @findex Source_File_Name
6372 @smallexample @c ada
6373 pragma Source_File_Name (
6374 [Unit_Name =>] unit_NAME,
6375 Spec_File_Name => STRING_LITERAL,
6376 [Index => INTEGER_LITERAL]);
6378 pragma Source_File_Name (
6379 [Unit_Name =>] unit_NAME,
6380 Body_File_Name => STRING_LITERAL,
6381 [Index => INTEGER_LITERAL]);
6385 Use this to override the normal naming convention. It is a configuration
6386 pragma, and so has the usual applicability of configuration pragmas
6387 (i.e.@: it applies to either an entire partition, or to all units in a
6388 compilation, or to a single unit, depending on how it is used.
6389 @var{unit_name} is mapped to @var{file_name_literal}. The identifier for
6390 the second argument is required, and indicates whether this is the file
6391 name for the spec or for the body.
6393 The optional Index argument should be used when a file contains multiple
6394 units, and when you do not want to use @code{gnatchop} to separate then
6395 into multiple files (which is the recommended procedure to limit the
6396 number of recompilations that are needed when some sources change).
6397 For instance, if the source file @file{source.ada} contains
6399 @smallexample @c ada
6411 you could use the following configuration pragmas:
6413 @smallexample @c ada
6414 pragma Source_File_Name
6415 (B, Spec_File_Name => "source.ada", Index => 1);
6416 pragma Source_File_Name
6417 (A, Body_File_Name => "source.ada", Index => 2);
6420 Note that the @code{gnatname} utility can also be used to generate those
6421 configuration pragmas.
6423 Another form of the @code{Source_File_Name} pragma allows
6424 the specification of patterns defining alternative file naming schemes
6425 to apply to all files.
6427 @smallexample @c ada
6428 pragma Source_File_Name
6429 ( [Spec_File_Name =>] STRING_LITERAL
6430 [,[Casing =>] CASING_SPEC]
6431 [,[Dot_Replacement =>] STRING_LITERAL]);
6433 pragma Source_File_Name
6434 ( [Body_File_Name =>] STRING_LITERAL
6435 [,[Casing =>] CASING_SPEC]
6436 [,[Dot_Replacement =>] STRING_LITERAL]);
6438 pragma Source_File_Name
6439 ( [Subunit_File_Name =>] STRING_LITERAL
6440 [,[Casing =>] CASING_SPEC]
6441 [,[Dot_Replacement =>] STRING_LITERAL]);
6443 CASING_SPEC ::= Lowercase | Uppercase | Mixedcase
6447 The first argument is a pattern that contains a single asterisk indicating
6448 the point at which the unit name is to be inserted in the pattern string
6449 to form the file name. The second argument is optional. If present it
6450 specifies the casing of the unit name in the resulting file name string.
6451 The default is lower case. Finally the third argument allows for systematic
6452 replacement of any dots in the unit name by the specified string literal.
6454 Note that Source_File_Name pragmas should not be used if you are using
6455 project files. The reason for this rule is that the project manager is not
6456 aware of these pragmas, and so other tools that use the projet file would not
6457 be aware of the intended naming conventions. If you are using project files,
6458 file naming is controlled by Source_File_Name_Project pragmas, which are
6459 usually supplied automatically by the project manager. A pragma
6460 Source_File_Name cannot appear after a @ref{Pragma Source_File_Name_Project}.
6462 For more details on the use of the @code{Source_File_Name} pragma,
6463 @xref{Using Other File Names,,, gnat_ugn, @value{EDITION} User's Guide},
6464 and @ref{Alternative File Naming Schemes,,, gnat_ugn, @value{EDITION}
6467 @node Pragma Source_File_Name_Project
6468 @unnumberedsec Pragma Source_File_Name_Project
6469 @findex Source_File_Name_Project
6472 This pragma has the same syntax and semantics as pragma Source_File_Name.
6473 It is only allowed as a stand alone configuration pragma.
6474 It cannot appear after a @ref{Pragma Source_File_Name}, and
6475 most importantly, once pragma Source_File_Name_Project appears,
6476 no further Source_File_Name pragmas are allowed.
6478 The intention is that Source_File_Name_Project pragmas are always
6479 generated by the Project Manager in a manner consistent with the naming
6480 specified in a project file, and when naming is controlled in this manner,
6481 it is not permissible to attempt to modify this naming scheme using
6482 Source_File_Name or Source_File_Name_Project pragmas (which would not be
6483 known to the project manager).
6485 @node Pragma Source_Reference
6486 @unnumberedsec Pragma Source_Reference
6487 @findex Source_Reference
6491 @smallexample @c ada
6492 pragma Source_Reference (INTEGER_LITERAL, STRING_LITERAL);
6496 This pragma must appear as the first line of a source file.
6497 @var{integer_literal} is the logical line number of the line following
6498 the pragma line (for use in error messages and debugging
6499 information). @var{string_literal} is a static string constant that
6500 specifies the file name to be used in error messages and debugging
6501 information. This is most notably used for the output of @code{gnatchop}
6502 with the @option{-r} switch, to make sure that the original unchopped
6503 source file is the one referred to.
6505 The second argument must be a string literal, it cannot be a static
6506 string expression other than a string literal. This is because its value
6507 is needed for error messages issued by all phases of the compiler.
6509 @node Pragma SPARK_Mode
6510 @unnumberedsec Pragma SPARK_Mode
6515 @smallexample @c ada
6516 pragma SPARK_Mode [(On | Off)] ;
6520 In general a program can have some parts that are in SPARK 2014 (and
6521 follow all the rules in the SPARK Reference Manual), and some parts
6522 that are full Ada 2012.
6524 The SPARK_Mode pragma is used to identify which parts are in SPARK
6525 2014 (by default programs are in full Ada). The SPARK_Mode pragma can
6526 be used in the following places:
6531 As a configuration pragma, in which case it sets the default mode for
6532 all units compiled with this pragma.
6535 Immediately following a library-level subprogram spec
6538 Immediately within a library-level package body
6541 Immediately following the @code{private} keyword of a library-level
6545 Immediately following the @code{begin} keyword of a library-level
6549 Immediately within a library-level subprogram body
6554 Normally a subprogram or package spec/body inherits the current mode
6555 that is active at the point it is declared. But this can be overridden
6556 by pragma within the spec or body as above.
6558 The basic consistency rule is that you can't turn SPARK_Mode back
6559 @code{On}, once you have explicitly (with a pragma) turned if
6560 @code{Off}. So the following rules apply:
6563 If a subprogram spec has SPARK_Mode @code{Off}, then the body must
6564 also have SPARK_Mode @code{Off}.
6567 For a package, we have four parts:
6571 the package public declarations
6573 the package private part
6575 the body of the package
6577 the elaboration code after @code{begin}
6581 For a package, the rule is that if you explicitly turn SPARK_Mode
6582 @code{Off} for any part, then all the following parts must have
6583 SPARK_Mode @code{Off}. Note that this may require repeating a pragma
6584 SPARK_Mode (@code{Off}) in the body. For example, if we have a
6585 configuration pragma SPARK_Mode (@code{On}) that turns the mode on by
6586 default everywhere, and one particular package spec has pragma
6587 SPARK_Mode (@code{Off}), then that pragma will need to be repeated in
6590 @node Pragma Static_Elaboration_Desired
6591 @unnumberedsec Pragma Static_Elaboration_Desired
6592 @findex Static_Elaboration_Desired
6596 @smallexample @c ada
6597 pragma Static_Elaboration_Desired;
6601 This pragma is used to indicate that the compiler should attempt to initialize
6602 statically the objects declared in the library unit to which the pragma applies,
6603 when these objects are initialized (explicitly or implicitly) by an aggregate.
6604 In the absence of this pragma, aggregates in object declarations are expanded
6605 into assignments and loops, even when the aggregate components are static
6606 constants. When the aggregate is present the compiler builds a static expression
6607 that requires no run-time code, so that the initialized object can be placed in
6608 read-only data space. If the components are not static, or the aggregate has
6609 more that 100 components, the compiler emits a warning that the pragma cannot
6610 be obeyed. (See also the restriction No_Implicit_Loops, which supports static
6611 construction of larger aggregates with static components that include an others
6614 @node Pragma Stream_Convert
6615 @unnumberedsec Pragma Stream_Convert
6616 @findex Stream_Convert
6620 @smallexample @c ada
6621 pragma Stream_Convert (
6622 [Entity =>] type_LOCAL_NAME,
6623 [Read =>] function_NAME,
6624 [Write =>] function_NAME);
6628 This pragma provides an efficient way of providing user-defined stream
6629 attributes. Not only is it simpler to use than specifying the attributes
6630 directly, but more importantly, it allows the specification to be made in such
6631 a way that the predefined unit Ada.Streams is not loaded unless it is actually
6632 needed (i.e. unless the stream attributes are actually used); the use of
6633 the Stream_Convert pragma adds no overhead at all, unless the stream
6634 attributes are actually used on the designated type.
6636 The first argument specifies the type for which stream functions are
6637 provided. The second parameter provides a function used to read values
6638 of this type. It must name a function whose argument type may be any
6639 subtype, and whose returned type must be the type given as the first
6640 argument to the pragma.
6642 The meaning of the @var{Read} parameter is that if a stream attribute directly
6643 or indirectly specifies reading of the type given as the first parameter,
6644 then a value of the type given as the argument to the Read function is
6645 read from the stream, and then the Read function is used to convert this
6646 to the required target type.
6648 Similarly the @var{Write} parameter specifies how to treat write attributes
6649 that directly or indirectly apply to the type given as the first parameter.
6650 It must have an input parameter of the type specified by the first parameter,
6651 and the return type must be the same as the input type of the Read function.
6652 The effect is to first call the Write function to convert to the given stream
6653 type, and then write the result type to the stream.
6655 The Read and Write functions must not be overloaded subprograms. If necessary
6656 renamings can be supplied to meet this requirement.
6657 The usage of this attribute is best illustrated by a simple example, taken
6658 from the GNAT implementation of package Ada.Strings.Unbounded:
6660 @smallexample @c ada
6661 function To_Unbounded (S : String)
6662 return Unbounded_String
6663 renames To_Unbounded_String;
6665 pragma Stream_Convert
6666 (Unbounded_String, To_Unbounded, To_String);
6670 The specifications of the referenced functions, as given in the Ada
6671 Reference Manual are:
6673 @smallexample @c ada
6674 function To_Unbounded_String (Source : String)
6675 return Unbounded_String;
6677 function To_String (Source : Unbounded_String)
6682 The effect is that if the value of an unbounded string is written to a stream,
6683 then the representation of the item in the stream is in the same format that
6684 would be used for @code{Standard.String'Output}, and this same representation
6685 is expected when a value of this type is read from the stream. Note that the
6686 value written always includes the bounds, even for Unbounded_String'Write,
6687 since Unbounded_String is not an array type.
6689 Note that the @code{Stream_Convert} pragma is not effective in the case of
6690 a derived type of a non-limited tagged type. If such a type is specified then
6691 the pragma is silently ignored, and the default implementation of the stream
6692 attributes is used instead.
6694 @node Pragma Style_Checks
6695 @unnumberedsec Pragma Style_Checks
6696 @findex Style_Checks
6700 @smallexample @c ada
6701 pragma Style_Checks (string_LITERAL | ALL_CHECKS |
6702 On | Off [, LOCAL_NAME]);
6706 This pragma is used in conjunction with compiler switches to control the
6707 built in style checking provided by GNAT@. The compiler switches, if set,
6708 provide an initial setting for the switches, and this pragma may be used
6709 to modify these settings, or the settings may be provided entirely by
6710 the use of the pragma. This pragma can be used anywhere that a pragma
6711 is legal, including use as a configuration pragma (including use in
6712 the @file{gnat.adc} file).
6714 The form with a string literal specifies which style options are to be
6715 activated. These are additive, so they apply in addition to any previously
6716 set style check options. The codes for the options are the same as those
6717 used in the @option{-gnaty} switch to @command{gcc} or @command{gnatmake}.
6718 For example the following two methods can be used to enable
6723 @smallexample @c ada
6724 pragma Style_Checks ("l");
6729 gcc -c -gnatyl @dots{}
6734 The form ALL_CHECKS activates all standard checks (its use is equivalent
6735 to the use of the @code{gnaty} switch with no options. @xref{Top,
6736 @value{EDITION} User's Guide, About This Guide, gnat_ugn,
6737 @value{EDITION} User's Guide}, for details.)
6739 Note: the behavior is slightly different in GNAT mode (@option{-gnatg} used).
6740 In this case, ALL_CHECKS implies the standard set of GNAT mode style check
6741 options (i.e. equivalent to -gnatyg).
6743 The forms with @code{Off} and @code{On}
6744 can be used to temporarily disable style checks
6745 as shown in the following example:
6747 @smallexample @c ada
6751 pragma Style_Checks ("k"); -- requires keywords in lower case
6752 pragma Style_Checks (Off); -- turn off style checks
6753 NULL; -- this will not generate an error message
6754 pragma Style_Checks (On); -- turn style checks back on
6755 NULL; -- this will generate an error message
6759 Finally the two argument form is allowed only if the first argument is
6760 @code{On} or @code{Off}. The effect is to turn of semantic style checks
6761 for the specified entity, as shown in the following example:
6763 @smallexample @c ada
6767 pragma Style_Checks ("r"); -- require consistency of identifier casing
6769 Rf1 : Integer := ARG; -- incorrect, wrong case
6770 pragma Style_Checks (Off, Arg);
6771 Rf2 : Integer := ARG; -- OK, no error
6774 @node Pragma Subtitle
6775 @unnumberedsec Pragma Subtitle
6780 @smallexample @c ada
6781 pragma Subtitle ([Subtitle =>] STRING_LITERAL);
6785 This pragma is recognized for compatibility with other Ada compilers
6786 but is ignored by GNAT@.
6788 @node Pragma Suppress
6789 @unnumberedsec Pragma Suppress
6794 @smallexample @c ada
6795 pragma Suppress (Identifier [, [On =>] Name]);
6799 This is a standard pragma, and supports all the check names required in
6800 the RM. It is included here because GNAT recognizes some additional check
6801 names that are implementation defined (as permitted by the RM):
6806 @code{Alignment_Check} can be used to suppress alignment checks
6807 on addresses used in address clauses. Such checks can also be suppressed
6808 by suppressing range checks, but the specific use of @code{Alignment_Check}
6809 allows suppression of alignment checks without suppressing other range checks.
6812 @code{Predicate_Check} can be used to control whether predicate checks are
6813 active. It is applicable only to predicates for which the policy is
6814 @code{Check}. Unlike @code{Assertion_Policy}, which determines if a given
6815 predicate is ignored or checked for the whole program, the use of
6816 @code{Suppress} and @code{Unsuppress} with this check name allows a given
6817 predicate to be turned on and off at specific points in the program.
6820 @code{Validity_Check} can be used specifically to control validity checks.
6821 If @code{Suppress} is used to suppress validity checks, then no validity
6822 checks are performed, including those specified by the appropriate compiler
6823 switch or the @code{Validity_Checks} pragma.
6826 Additional check names previously introduced by use of the @code{Check_Name}
6827 pragma are also allowed.
6832 Note that pragma Suppress gives the compiler permission to omit
6833 checks, but does not require the compiler to omit checks. The compiler
6834 will generate checks if they are essentially free, even when they are
6835 suppressed. In particular, if the compiler can prove that a certain
6836 check will necessarily fail, it will generate code to do an
6837 unconditional ``raise'', even if checks are suppressed. The compiler
6840 Of course, run-time checks are omitted whenever the compiler can prove
6841 that they will not fail, whether or not checks are suppressed.
6843 @node Pragma Suppress_All
6844 @unnumberedsec Pragma Suppress_All
6845 @findex Suppress_All
6849 @smallexample @c ada
6850 pragma Suppress_All;
6854 This pragma can appear anywhere within a unit.
6855 The effect is to apply @code{Suppress (All_Checks)} to the unit
6856 in which it appears. This pragma is implemented for compatibility with DEC
6857 Ada 83 usage where it appears at the end of a unit, and for compatibility
6858 with Rational Ada, where it appears as a program unit pragma.
6859 The use of the standard Ada pragma @code{Suppress (All_Checks)}
6860 as a normal configuration pragma is the preferred usage in GNAT@.
6862 @node Pragma Suppress_Debug_Info
6863 @unnumberedsec Pragma Suppress_Debug_Info
6864 @findex Suppress_Debug_Info
6868 @smallexample @c ada
6869 Suppress_Debug_Info ([Entity =>] LOCAL_NAME);
6873 This pragma can be used to suppress generation of debug information
6874 for the specified entity. It is intended primarily for use in debugging
6875 the debugger, and navigating around debugger problems.
6877 @node Pragma Suppress_Exception_Locations
6878 @unnumberedsec Pragma Suppress_Exception_Locations
6879 @findex Suppress_Exception_Locations
6883 @smallexample @c ada
6884 pragma Suppress_Exception_Locations;
6888 In normal mode, a raise statement for an exception by default generates
6889 an exception message giving the file name and line number for the location
6890 of the raise. This is useful for debugging and logging purposes, but this
6891 entails extra space for the strings for the messages. The configuration
6892 pragma @code{Suppress_Exception_Locations} can be used to suppress the
6893 generation of these strings, with the result that space is saved, but the
6894 exception message for such raises is null. This configuration pragma may
6895 appear in a global configuration pragma file, or in a specific unit as
6896 usual. It is not required that this pragma be used consistently within
6897 a partition, so it is fine to have some units within a partition compiled
6898 with this pragma and others compiled in normal mode without it.
6900 @node Pragma Suppress_Initialization
6901 @unnumberedsec Pragma Suppress_Initialization
6902 @findex Suppress_Initialization
6903 @cindex Suppressing initialization
6904 @cindex Initialization, suppression of
6908 @smallexample @c ada
6909 pragma Suppress_Initialization ([Entity =>] subtype_Name);
6913 Here subtype_Name is the name introduced by a type declaration
6914 or subtype declaration.
6915 This pragma suppresses any implicit or explicit initialization
6916 for all variables of the given type or subtype,
6917 including initialization resulting from the use of pragmas
6918 Normalize_Scalars or Initialize_Scalars.
6920 This is considered a representation item, so it cannot be given after
6921 the type is frozen. It applies to all subsequent object declarations,
6922 and also any allocator that creates objects of the type.
6924 If the pragma is given for the first subtype, then it is considered
6925 to apply to the base type and all its subtypes. If the pragma is given
6926 for other than a first subtype, then it applies only to the given subtype.
6927 The pragma may not be given after the type is frozen.
6929 @node Pragma Task_Info
6930 @unnumberedsec Pragma Task_Info
6935 @smallexample @c ada
6936 pragma Task_Info (EXPRESSION);
6940 This pragma appears within a task definition (like pragma
6941 @code{Priority}) and applies to the task in which it appears. The
6942 argument must be of type @code{System.Task_Info.Task_Info_Type}.
6943 The @code{Task_Info} pragma provides system dependent control over
6944 aspects of tasking implementation, for example, the ability to map
6945 tasks to specific processors. For details on the facilities available
6946 for the version of GNAT that you are using, see the documentation
6947 in the spec of package System.Task_Info in the runtime
6950 @node Pragma Task_Name
6951 @unnumberedsec Pragma Task_Name
6956 @smallexample @c ada
6957 pragma Task_Name (string_EXPRESSION);
6961 This pragma appears within a task definition (like pragma
6962 @code{Priority}) and applies to the task in which it appears. The
6963 argument must be of type String, and provides a name to be used for
6964 the task instance when the task is created. Note that this expression
6965 is not required to be static, and in particular, it can contain
6966 references to task discriminants. This facility can be used to
6967 provide different names for different tasks as they are created,
6968 as illustrated in the example below.
6970 The task name is recorded internally in the run-time structures
6971 and is accessible to tools like the debugger. In addition the
6972 routine @code{Ada.Task_Identification.Image} will return this
6973 string, with a unique task address appended.
6975 @smallexample @c ada
6976 -- Example of the use of pragma Task_Name
6978 with Ada.Task_Identification;
6979 use Ada.Task_Identification;
6980 with Text_IO; use Text_IO;
6983 type Astring is access String;
6985 task type Task_Typ (Name : access String) is
6986 pragma Task_Name (Name.all);
6989 task body Task_Typ is
6990 Nam : constant String := Image (Current_Task);
6992 Put_Line ("-->" & Nam (1 .. 14) & "<--");
6995 type Ptr_Task is access Task_Typ;
6996 Task_Var : Ptr_Task;
7000 new Task_Typ (new String'("This is task 1"));
7002 new Task_Typ (new String'("This is task 2"));
7006 @node Pragma Task_Storage
7007 @unnumberedsec Pragma Task_Storage
7008 @findex Task_Storage
7011 @smallexample @c ada
7012 pragma Task_Storage (
7013 [Task_Type =>] LOCAL_NAME,
7014 [Top_Guard =>] static_integer_EXPRESSION);
7018 This pragma specifies the length of the guard area for tasks. The guard
7019 area is an additional storage area allocated to a task. A value of zero
7020 means that either no guard area is created or a minimal guard area is
7021 created, depending on the target. This pragma can appear anywhere a
7022 @code{Storage_Size} attribute definition clause is allowed for a task
7025 @node Pragma Test_Case
7026 @unnumberedsec Pragma Test_Case
7032 @smallexample @c ada
7034 [Name =>] static_string_Expression
7035 ,[Mode =>] (Nominal | Robustness)
7036 [, Requires => Boolean_Expression]
7037 [, Ensures => Boolean_Expression]);
7041 The @code{Test_Case} pragma allows defining fine-grain specifications
7042 for use by testing tools.
7043 The compiler checks the validity of the @code{Test_Case} pragma, but its
7044 presence does not lead to any modification of the code generated by the
7047 @code{Test_Case} pragmas may only appear immediately following the
7048 (separate) declaration of a subprogram in a package declaration, inside
7049 a package spec unit. Only other pragmas may intervene (that is appear
7050 between the subprogram declaration and a test case).
7052 The compiler checks that boolean expressions given in @code{Requires} and
7053 @code{Ensures} are valid, where the rules for @code{Requires} are the
7054 same as the rule for an expression in @code{Precondition} and the rules
7055 for @code{Ensures} are the same as the rule for an expression in
7056 @code{Postcondition}. In particular, attributes @code{'Old} and
7057 @code{'Result} can only be used within the @code{Ensures}
7058 expression. The following is an example of use within a package spec:
7060 @smallexample @c ada
7061 package Math_Functions is
7063 function Sqrt (Arg : Float) return Float;
7064 pragma Test_Case (Name => "Test 1",
7066 Requires => Arg < 10000,
7067 Ensures => Sqrt'Result < 10);
7073 The meaning of a test case is that there is at least one context where
7074 @code{Requires} holds such that, if the associated subprogram is executed in
7075 that context, then @code{Ensures} holds when the subprogram returns.
7076 Mode @code{Nominal} indicates that the input context should also satisfy the
7077 precondition of the subprogram, and the output context should also satisfy its
7078 postcondition. More @code{Robustness} indicates that the precondition and
7079 postcondition of the subprogram should be ignored for this test case.
7081 @node Pragma Thread_Local_Storage
7082 @unnumberedsec Pragma Thread_Local_Storage
7083 @findex Thread_Local_Storage
7084 @cindex Task specific storage
7085 @cindex TLS (Thread Local Storage)
7086 @cindex Task_Attributes
7089 @smallexample @c ada
7090 pragma Thread_Local_Storage ([Entity =>] LOCAL_NAME);
7094 This pragma specifies that the specified entity, which must be
7095 a variable declared in a library level package, is to be marked as
7096 "Thread Local Storage" (@code{TLS}). On systems supporting this (which
7097 include Solaris, GNU/Linux and VxWorks 6), this causes each thread
7098 (and hence each Ada task) to see a distinct copy of the variable.
7100 The variable may not have default initialization, and if there is
7101 an explicit initialization, it must be either @code{null} for an
7102 access variable, or a static expression for a scalar variable.
7103 This provides a low level mechanism similar to that provided by
7104 the @code{Ada.Task_Attributes} package, but much more efficient
7105 and is also useful in writing interface code that will interact
7106 with foreign threads.
7108 If this pragma is used on a system where @code{TLS} is not supported,
7109 then an error message will be generated and the program will be rejected.
7111 @node Pragma Time_Slice
7112 @unnumberedsec Pragma Time_Slice
7117 @smallexample @c ada
7118 pragma Time_Slice (static_duration_EXPRESSION);
7122 For implementations of GNAT on operating systems where it is possible
7123 to supply a time slice value, this pragma may be used for this purpose.
7124 It is ignored if it is used in a system that does not allow this control,
7125 or if it appears in other than the main program unit.
7127 Note that the effect of this pragma is identical to the effect of the
7128 DEC Ada 83 pragma of the same name when operating under OpenVMS systems.
7131 @unnumberedsec Pragma Title
7136 @smallexample @c ada
7137 pragma Title (TITLING_OPTION [, TITLING OPTION]);
7140 [Title =>] STRING_LITERAL,
7141 | [Subtitle =>] STRING_LITERAL
7145 Syntax checked but otherwise ignored by GNAT@. This is a listing control
7146 pragma used in DEC Ada 83 implementations to provide a title and/or
7147 subtitle for the program listing. The program listing generated by GNAT
7148 does not have titles or subtitles.
7150 Unlike other pragmas, the full flexibility of named notation is allowed
7151 for this pragma, i.e.@: the parameters may be given in any order if named
7152 notation is used, and named and positional notation can be mixed
7153 following the normal rules for procedure calls in Ada.
7155 @node Pragma Type_Invariant
7156 @unnumberedsec Pragma Type_Invariant
7158 @findex Type_Invariant pragma
7162 @smallexample @c ada
7163 pragma Type_Invariant
7164 ([Entity =>] type_LOCAL_NAME,
7165 [Check =>] EXPRESSION);
7169 The @code{Type_Invariant} pragma is intended to be an exact
7170 replacement for the language-defined @code{Type_Invariant}
7171 aspect, and shares its restrictions and semantics. It differs
7172 from the language defined @code{Invariant} pragma in that it
7173 does not permit a string parameter, and it is
7174 controlled by the assertion identifier @code{Type_Invariant}
7175 rather than @code{Invariant}.
7177 @node Pragma Type_Invariant_Class
7178 @unnumberedsec Pragma Type_Invariant_Class
7180 @findex Type_Invariant_Class pragma
7184 @smallexample @c ada
7185 pragma Type_Invariant_Class
7186 ([Entity =>] type_LOCAL_NAME,
7187 [Check =>] EXPRESSION);
7191 The @code{Type_Invariant_Class} pragma is intended to be an exact
7192 replacement for the language-defined @code{Type_Invariant'Class}
7193 aspect, and shares its restrictions and semantics.
7195 Note: This pragma is called @code{Type_Invariant_Class} rather than
7196 @code{Type_Invariant'Class} because the latter would not be strictly
7197 conforming to the allowed syntax for pragmas. The motivation
7198 for providing pragmas equivalent to the aspects is to allow a program
7199 to be written using the pragmas, and then compiled if necessary
7200 using an Ada compiler that does not recognize the pragmas or
7201 aspects, but is prepared to ignore the pragmas. The assertion
7202 policy that controls this pragma is @code{Type_Invariant'Class},
7203 not @code{Type_Invariant_Class}.
7205 @node Pragma Unchecked_Union
7206 @unnumberedsec Pragma Unchecked_Union
7208 @findex Unchecked_Union
7212 @smallexample @c ada
7213 pragma Unchecked_Union (first_subtype_LOCAL_NAME);
7217 This pragma is used to specify a representation of a record type that is
7218 equivalent to a C union. It was introduced as a GNAT implementation defined
7219 pragma in the GNAT Ada 95 mode. Ada 2005 includes an extended version of this
7220 pragma, making it language defined, and GNAT fully implements this extended
7221 version in all language modes (Ada 83, Ada 95, and Ada 2005). For full
7222 details, consult the Ada 2012 Reference Manual, section B.3.3.
7224 @node Pragma Unimplemented_Unit
7225 @unnumberedsec Pragma Unimplemented_Unit
7226 @findex Unimplemented_Unit
7230 @smallexample @c ada
7231 pragma Unimplemented_Unit;
7235 If this pragma occurs in a unit that is processed by the compiler, GNAT
7236 aborts with the message @samp{@var{xxx} not implemented}, where
7237 @var{xxx} is the name of the current compilation unit. This pragma is
7238 intended to allow the compiler to handle unimplemented library units in
7241 The abort only happens if code is being generated. Thus you can use
7242 specs of unimplemented packages in syntax or semantic checking mode.
7244 @node Pragma Universal_Aliasing
7245 @unnumberedsec Pragma Universal_Aliasing
7246 @findex Universal_Aliasing
7250 @smallexample @c ada
7251 pragma Universal_Aliasing [([Entity =>] type_LOCAL_NAME)];
7255 @var{type_LOCAL_NAME} must refer to a type declaration in the current
7256 declarative part. The effect is to inhibit strict type-based aliasing
7257 optimization for the given type. In other words, the effect is as though
7258 access types designating this type were subject to pragma No_Strict_Aliasing.
7259 For a detailed description of the strict aliasing optimization, and the
7260 situations in which it must be suppressed, @xref{Optimization and Strict
7261 Aliasing,,, gnat_ugn, @value{EDITION} User's Guide}.
7263 @node Pragma Universal_Data
7264 @unnumberedsec Pragma Universal_Data
7265 @findex Universal_Data
7269 @smallexample @c ada
7270 pragma Universal_Data [(library_unit_Name)];
7274 This pragma is supported only for the AAMP target and is ignored for
7275 other targets. The pragma specifies that all library-level objects
7276 (Counter 0 data) associated with the library unit are to be accessed
7277 and updated using universal addressing (24-bit addresses for AAMP5)
7278 rather than the default of 16-bit Data Environment (DENV) addressing.
7279 Use of this pragma will generally result in less efficient code for
7280 references to global data associated with the library unit, but
7281 allows such data to be located anywhere in memory. This pragma is
7282 a library unit pragma, but can also be used as a configuration pragma
7283 (including use in the @file{gnat.adc} file). The functionality
7284 of this pragma is also available by applying the -univ switch on the
7285 compilations of units where universal addressing of the data is desired.
7287 @node Pragma Unmodified
7288 @unnumberedsec Pragma Unmodified
7290 @cindex Warnings, unmodified
7294 @smallexample @c ada
7295 pragma Unmodified (LOCAL_NAME @{, LOCAL_NAME@});
7299 This pragma signals that the assignable entities (variables,
7300 @code{out} parameters, @code{in out} parameters) whose names are listed are
7301 deliberately not assigned in the current source unit. This
7302 suppresses warnings about the
7303 entities being referenced but not assigned, and in addition a warning will be
7304 generated if one of these entities is in fact assigned in the
7305 same unit as the pragma (or in the corresponding body, or one
7308 This is particularly useful for clearly signaling that a particular
7309 parameter is not modified, even though the spec suggests that it might
7312 For the variable case, warnings are never given for unreferenced variables
7313 whose name contains one of the substrings
7314 @code{DISCARD, DUMMY, IGNORE, JUNK, UNUSED} in any casing. Such names
7315 are typically to be used in cases where such warnings are expected.
7316 Thus it is never necessary to use @code{pragma Unmodified} for such
7317 variables, though it is harmless to do so.
7319 @node Pragma Unreferenced
7320 @unnumberedsec Pragma Unreferenced
7321 @findex Unreferenced
7322 @cindex Warnings, unreferenced
7326 @smallexample @c ada
7327 pragma Unreferenced (LOCAL_NAME @{, LOCAL_NAME@});
7328 pragma Unreferenced (library_unit_NAME @{, library_unit_NAME@});
7332 This pragma signals that the entities whose names are listed are
7333 deliberately not referenced in the current source unit. This
7334 suppresses warnings about the
7335 entities being unreferenced, and in addition a warning will be
7336 generated if one of these entities is in fact subsequently referenced in the
7337 same unit as the pragma (or in the corresponding body, or one
7340 This is particularly useful for clearly signaling that a particular
7341 parameter is not referenced in some particular subprogram implementation
7342 and that this is deliberate. It can also be useful in the case of
7343 objects declared only for their initialization or finalization side
7346 If @code{LOCAL_NAME} identifies more than one matching homonym in the
7347 current scope, then the entity most recently declared is the one to which
7348 the pragma applies. Note that in the case of accept formals, the pragma
7349 Unreferenced may appear immediately after the keyword @code{do} which
7350 allows the indication of whether or not accept formals are referenced
7351 or not to be given individually for each accept statement.
7353 The left hand side of an assignment does not count as a reference for the
7354 purpose of this pragma. Thus it is fine to assign to an entity for which
7355 pragma Unreferenced is given.
7357 Note that if a warning is desired for all calls to a given subprogram,
7358 regardless of whether they occur in the same unit as the subprogram
7359 declaration, then this pragma should not be used (calls from another
7360 unit would not be flagged); pragma Obsolescent can be used instead
7361 for this purpose, see @xref{Pragma Obsolescent}.
7363 The second form of pragma @code{Unreferenced} is used within a context
7364 clause. In this case the arguments must be unit names of units previously
7365 mentioned in @code{with} clauses (similar to the usage of pragma
7366 @code{Elaborate_All}. The effect is to suppress warnings about unreferenced
7367 units and unreferenced entities within these units.
7369 For the variable case, warnings are never given for unreferenced variables
7370 whose name contains one of the substrings
7371 @code{DISCARD, DUMMY, IGNORE, JUNK, UNUSED} in any casing. Such names
7372 are typically to be used in cases where such warnings are expected.
7373 Thus it is never necessary to use @code{pragma Unreferenced} for such
7374 variables, though it is harmless to do so.
7376 @node Pragma Unreferenced_Objects
7377 @unnumberedsec Pragma Unreferenced_Objects
7378 @findex Unreferenced_Objects
7379 @cindex Warnings, unreferenced
7383 @smallexample @c ada
7384 pragma Unreferenced_Objects (local_subtype_NAME @{, local_subtype_NAME@});
7388 This pragma signals that for the types or subtypes whose names are
7389 listed, objects which are declared with one of these types or subtypes may
7390 not be referenced, and if no references appear, no warnings are given.
7392 This is particularly useful for objects which are declared solely for their
7393 initialization and finalization effect. Such variables are sometimes referred
7394 to as RAII variables (Resource Acquisition Is Initialization). Using this
7395 pragma on the relevant type (most typically a limited controlled type), the
7396 compiler will automatically suppress unwanted warnings about these variables
7397 not being referenced.
7399 @node Pragma Unreserve_All_Interrupts
7400 @unnumberedsec Pragma Unreserve_All_Interrupts
7401 @findex Unreserve_All_Interrupts
7405 @smallexample @c ada
7406 pragma Unreserve_All_Interrupts;
7410 Normally certain interrupts are reserved to the implementation. Any attempt
7411 to attach an interrupt causes Program_Error to be raised, as described in
7412 RM C.3.2(22). A typical example is the @code{SIGINT} interrupt used in
7413 many systems for a @kbd{Ctrl-C} interrupt. Normally this interrupt is
7414 reserved to the implementation, so that @kbd{Ctrl-C} can be used to
7415 interrupt execution.
7417 If the pragma @code{Unreserve_All_Interrupts} appears anywhere in any unit in
7418 a program, then all such interrupts are unreserved. This allows the
7419 program to handle these interrupts, but disables their standard
7420 functions. For example, if this pragma is used, then pressing
7421 @kbd{Ctrl-C} will not automatically interrupt execution. However,
7422 a program can then handle the @code{SIGINT} interrupt as it chooses.
7424 For a full list of the interrupts handled in a specific implementation,
7425 see the source code for the spec of @code{Ada.Interrupts.Names} in
7426 file @file{a-intnam.ads}. This is a target dependent file that contains the
7427 list of interrupts recognized for a given target. The documentation in
7428 this file also specifies what interrupts are affected by the use of
7429 the @code{Unreserve_All_Interrupts} pragma.
7431 For a more general facility for controlling what interrupts can be
7432 handled, see pragma @code{Interrupt_State}, which subsumes the functionality
7433 of the @code{Unreserve_All_Interrupts} pragma.
7435 @node Pragma Unsuppress
7436 @unnumberedsec Pragma Unsuppress
7441 @smallexample @c ada
7442 pragma Unsuppress (IDENTIFIER [, [On =>] NAME]);
7446 This pragma undoes the effect of a previous pragma @code{Suppress}. If
7447 there is no corresponding pragma @code{Suppress} in effect, it has no
7448 effect. The range of the effect is the same as for pragma
7449 @code{Suppress}. The meaning of the arguments is identical to that used
7450 in pragma @code{Suppress}.
7452 One important application is to ensure that checks are on in cases where
7453 code depends on the checks for its correct functioning, so that the code
7454 will compile correctly even if the compiler switches are set to suppress
7457 This pragma is standard in Ada 2005. It is available in all earlier versions
7458 of Ada as an implementation-defined pragma.
7460 Note that in addition to the checks defined in the Ada RM, GNAT recogizes
7461 a number of implementation-defined check names. See description of pragma
7462 @code{Suppress} for full details.
7464 @node Pragma Use_VADS_Size
7465 @unnumberedsec Pragma Use_VADS_Size
7466 @cindex @code{Size}, VADS compatibility
7467 @cindex Rational profile
7468 @findex Use_VADS_Size
7472 @smallexample @c ada
7473 pragma Use_VADS_Size;
7477 This is a configuration pragma. In a unit to which it applies, any use
7478 of the 'Size attribute is automatically interpreted as a use of the
7479 'VADS_Size attribute. Note that this may result in incorrect semantic
7480 processing of valid Ada 95 or Ada 2005 programs. This is intended to aid in
7481 the handling of existing code which depends on the interpretation of Size
7482 as implemented in the VADS compiler. See description of the VADS_Size
7483 attribute for further details.
7485 @node Pragma Validity_Checks
7486 @unnumberedsec Pragma Validity_Checks
7487 @findex Validity_Checks
7491 @smallexample @c ada
7492 pragma Validity_Checks (string_LITERAL | ALL_CHECKS | On | Off);
7496 This pragma is used in conjunction with compiler switches to control the
7497 built-in validity checking provided by GNAT@. The compiler switches, if set
7498 provide an initial setting for the switches, and this pragma may be used
7499 to modify these settings, or the settings may be provided entirely by
7500 the use of the pragma. This pragma can be used anywhere that a pragma
7501 is legal, including use as a configuration pragma (including use in
7502 the @file{gnat.adc} file).
7504 The form with a string literal specifies which validity options are to be
7505 activated. The validity checks are first set to include only the default
7506 reference manual settings, and then a string of letters in the string
7507 specifies the exact set of options required. The form of this string
7508 is exactly as described for the @option{-gnatVx} compiler switch (see the
7509 @value{EDITION} User's Guide for details). For example the following two
7510 methods can be used to enable validity checking for mode @code{in} and
7511 @code{in out} subprogram parameters:
7515 @smallexample @c ada
7516 pragma Validity_Checks ("im");
7521 gcc -c -gnatVim @dots{}
7526 The form ALL_CHECKS activates all standard checks (its use is equivalent
7527 to the use of the @code{gnatva} switch.
7529 The forms with @code{Off} and @code{On}
7530 can be used to temporarily disable validity checks
7531 as shown in the following example:
7533 @smallexample @c ada
7537 pragma Validity_Checks ("c"); -- validity checks for copies
7538 pragma Validity_Checks (Off); -- turn off validity checks
7539 A := B; -- B will not be validity checked
7540 pragma Validity_Checks (On); -- turn validity checks back on
7541 A := C; -- C will be validity checked
7544 @node Pragma Volatile
7545 @unnumberedsec Pragma Volatile
7550 @smallexample @c ada
7551 pragma Volatile (LOCAL_NAME);
7555 This pragma is defined by the Ada Reference Manual, and the GNAT
7556 implementation is fully conformant with this definition. The reason it
7557 is mentioned in this section is that a pragma of the same name was supplied
7558 in some Ada 83 compilers, including DEC Ada 83. The Ada 95 / Ada 2005
7559 implementation of pragma Volatile is upwards compatible with the
7560 implementation in DEC Ada 83.
7562 @node Pragma Warning_As_Error
7563 @unnumberedsec Pragma Warning_As_Error
7564 @findex Warning_As_Error
7568 @smallexample @c ada
7569 pragma Warning_As_Error (static_string_EXPRESSION);
7573 This configuration pragma allows the programmer to specify a set
7574 of warnings that will be treated as errors. Any warning which
7575 matches the pattern given by the pragma argument will be treated
7576 as an error. This gives much more precise control that -gnatwe
7577 which treats all warnings as errors.
7579 The pattern may contain asterisks, which match zero or more characters in
7580 the message. For example, you can use
7581 @code{pragma Warning_As_Error ("*bits of*unused")} to treat the warning
7582 message @code{warning: 960 bits of "a" unused} as an error. No other regular
7583 expression notations are permitted. All characters other than asterisk in
7584 these three specific cases are treated as literal characters in the match.
7585 The match is case insensitive, for example XYZ matches xyz.
7587 Another possibility for the static_string_EXPRESSION which works whether
7588 or not error tags are enabled (@option{-gnatw.d}) is to use the
7589 @option{-gnatw} tag string, enclosed in brackets,
7590 as shown in the example below, to treat a class of warnings as errors.
7592 The above use of patterns to match the message applies only to warning
7593 messages generated by the front end. This pragma can also be applied to
7594 warnings provided by the back end and mentioned in @ref{Pragma Warnings}.
7595 By using a single full @option{-Wxxx} switch in the pragma, such warnings
7596 can also be treated as errors.
7598 The pragma can appear either in a global configuration pragma file
7599 (e.g. @file{gnat.adc}), or at the start of a file. Given a global
7600 configuration pragma file containing:
7602 @smallexample @c ada
7603 pragma Warning_As_Error ("[-gnatwj]");
7607 which will treat all obsolescent feature warnings as errors, the
7608 following program compiles as shown (compile options here are
7609 @option{-gnatwa.d -gnatl -gnatj55}).
7611 @smallexample @c ada
7612 1. pragma Warning_As_Error ("*never assigned*");
7613 2. function Warnerr return String is
7616 >>> error: variable "X" is never read and
7617 never assigned [-gnatwv] [warning-as-error]
7621 >>> warning: variable "Y" is assigned but
7622 never read [-gnatwu]
7628 >>> error: use of "%" is an obsolescent
7629 feature (RM J.2(4)), use """ instead
7630 [-gnatwj] [warning-as-error]
7634 8 lines: No errors, 3 warnings (2 treated as errors)
7638 Note that this pragma does not affect the set of warnings issued in
7639 any way, it merely changes the effect of a matching warning if one
7640 is produced as a result of other warnings options. As shown in this
7641 example, if the pragma results in a warning being treated as an error,
7642 the tag is changed from "warning:" to "error:" and the string
7643 "[warning-as-error]" is appended to the end of the message.
7645 @node Pragma Warnings
7646 @unnumberedsec Pragma Warnings
7651 @smallexample @c ada
7652 pragma Warnings (On | Off [,REASON]);
7653 pragma Warnings (On | Off, LOCAL_NAME [,REASON]);
7654 pragma Warnings (static_string_EXPRESSION [,REASON]);
7655 pragma Warnings (On | Off, static_string_EXPRESSION [,REASON]);
7657 REASON ::= Reason => STRING_LITERAL @{& STRING_LITERAL@}
7661 Normally warnings are enabled, with the output being controlled by
7662 the command line switch. Warnings (@code{Off}) turns off generation of
7663 warnings until a Warnings (@code{On}) is encountered or the end of the
7664 current unit. If generation of warnings is turned off using this
7665 pragma, then some or all of the warning messages are suppressed,
7666 regardless of the setting of the command line switches.
7668 The @code{Reason} parameter may optionally appear as the last argument
7669 in any of the forms of this pragma. It is intended purely for the
7670 purposes of documenting the reason for the @code{Warnings} pragma.
7671 The compiler will check that the argument is a static string but
7672 otherwise ignore this argument. Other tools may provide specialized
7673 processing for this string.
7675 The form with a single argument (or two arguments if Reason present),
7676 where the first argument is @code{ON} or @code{OFF}
7677 may be used as a configuration pragma.
7679 If the @var{LOCAL_NAME} parameter is present, warnings are suppressed for
7680 the specified entity. This suppression is effective from the point where
7681 it occurs till the end of the extended scope of the variable (similar to
7682 the scope of @code{Suppress}). This form cannot be used as a configuration
7685 The form with a single static_string_EXPRESSION argument (and possible
7686 reason) provides more precise
7687 control over which warnings are active. The string is a list of letters
7688 specifying which warnings are to be activated and which deactivated. The
7689 code for these letters is the same as the string used in the command
7690 line switch controlling warnings. For a brief summary, use the gnatmake
7691 command with no arguments, which will generate usage information containing
7692 the list of warnings switches supported. For
7693 full details see @ref{Warning Message Control,,, gnat_ugn, @value{EDITION}
7694 User's Guide}. This form can also be used as a configuration pragma.
7697 The warnings controlled by the @option{-gnatw} switch are generated by the
7698 front end of the compiler. The GCC back end can provide additional warnings
7699 and they are controlled by the @option{-W} switch. Such warnings can be
7700 identified by the appearance of a string of the form @code{[-Wxxx]} in the
7701 message which designates the @option{-Wxxx} switch that controls the message.
7702 The form with a single static_string_EXPRESSION argument also works for these
7703 warnings, but the string must be a single full @option{-Wxxx} switch in this
7704 case. The above reference lists a few examples of these additional warnings.
7707 The specified warnings will be in effect until the end of the program
7708 or another pragma Warnings is encountered. The effect of the pragma is
7709 cumulative. Initially the set of warnings is the standard default set
7710 as possibly modified by compiler switches. Then each pragma Warning
7711 modifies this set of warnings as specified. This form of the pragma may
7712 also be used as a configuration pragma.
7714 The fourth form, with an @code{On|Off} parameter and a string, is used to
7715 control individual messages, based on their text. The string argument
7716 is a pattern that is used to match against the text of individual
7717 warning messages (not including the initial "warning: " tag).
7719 The pattern may contain asterisks, which match zero or more characters in
7720 the message. For example, you can use
7721 @code{pragma Warnings (Off, "*bits of*unused")} to suppress the warning
7722 message @code{warning: 960 bits of "a" unused}. No other regular
7723 expression notations are permitted. All characters other than asterisk in
7724 these three specific cases are treated as literal characters in the match.
7725 The match is case insensitive, for example XYZ matches xyz.
7727 The above use of patterns to match the message applies only to warning
7728 messages generated by the front end. This form of the pragma with a string
7729 argument can also be used to control warnings provided by the back end and
7730 mentioned above. By using a single full @option{-Wxxx} switch in the pragma,
7731 such warnings can be turned on and off.
7733 There are two ways to use the pragma in this form. The OFF form can be used as a
7734 configuration pragma. The effect is to suppress all warnings (if any)
7735 that match the pattern string throughout the compilation (or match the
7736 -W switch in the back end case).
7738 The second usage is to suppress a warning locally, and in this case, two
7739 pragmas must appear in sequence:
7741 @smallexample @c ada
7742 pragma Warnings (Off, Pattern);
7743 @dots{} code where given warning is to be suppressed
7744 pragma Warnings (On, Pattern);
7748 In this usage, the pattern string must match in the Off and On pragmas,
7749 and at least one matching warning must be suppressed.
7751 Note: to write a string that will match any warning, use the string
7752 @code{"***"}. It will not work to use a single asterisk or two asterisks
7753 since this looks like an operator name. This form with three asterisks
7754 is similar in effect to specifying @code{pragma Warnings (Off)} except that a
7755 matching @code{pragma Warnings (On, "***")} will be required. This can be
7756 helpful in avoiding forgetting to turn warnings back on.
7758 Note: the debug flag -gnatd.i (@code{/NOWARNINGS_PRAGMAS} in VMS) can be
7759 used to cause the compiler to entirely ignore all WARNINGS pragmas. This can
7760 be useful in checking whether obsolete pragmas in existing programs are hiding
7763 Note: pragma Warnings does not affect the processing of style messages. See
7764 separate entry for pragma Style_Checks for control of style messages.
7766 @node Pragma Weak_External
7767 @unnumberedsec Pragma Weak_External
7768 @findex Weak_External
7772 @smallexample @c ada
7773 pragma Weak_External ([Entity =>] LOCAL_NAME);
7777 @var{LOCAL_NAME} must refer to an object that is declared at the library
7778 level. This pragma specifies that the given entity should be marked as a
7779 weak symbol for the linker. It is equivalent to @code{__attribute__((weak))}
7780 in GNU C and causes @var{LOCAL_NAME} to be emitted as a weak symbol instead
7781 of a regular symbol, that is to say a symbol that does not have to be
7782 resolved by the linker if used in conjunction with a pragma Import.
7784 When a weak symbol is not resolved by the linker, its address is set to
7785 zero. This is useful in writing interfaces to external modules that may
7786 or may not be linked in the final executable, for example depending on
7787 configuration settings.
7789 If a program references at run time an entity to which this pragma has been
7790 applied, and the corresponding symbol was not resolved at link time, then
7791 the execution of the program is erroneous. It is not erroneous to take the
7792 Address of such an entity, for example to guard potential references,
7793 as shown in the example below.
7795 Some file formats do not support weak symbols so not all target machines
7796 support this pragma.
7798 @smallexample @c ada
7799 -- Example of the use of pragma Weak_External
7801 package External_Module is
7803 pragma Import (C, key);
7804 pragma Weak_External (key);
7805 function Present return boolean;
7806 end External_Module;
7808 with System; use System;
7809 package body External_Module is
7810 function Present return boolean is
7812 return key'Address /= System.Null_Address;
7814 end External_Module;
7817 @node Pragma Wide_Character_Encoding
7818 @unnumberedsec Pragma Wide_Character_Encoding
7819 @findex Wide_Character_Encoding
7823 @smallexample @c ada
7824 pragma Wide_Character_Encoding (IDENTIFIER | CHARACTER_LITERAL);
7828 This pragma specifies the wide character encoding to be used in program
7829 source text appearing subsequently. It is a configuration pragma, but may
7830 also be used at any point that a pragma is allowed, and it is permissible
7831 to have more than one such pragma in a file, allowing multiple encodings
7832 to appear within the same file.
7834 The argument can be an identifier or a character literal. In the identifier
7835 case, it is one of @code{HEX}, @code{UPPER}, @code{SHIFT_JIS},
7836 @code{EUC}, @code{UTF8}, or @code{BRACKETS}. In the character literal
7837 case it is correspondingly one of the characters @samp{h}, @samp{u},
7838 @samp{s}, @samp{e}, @samp{8}, or @samp{b}.
7840 Note that when the pragma is used within a file, it affects only the
7841 encoding within that file, and does not affect withed units, specs,
7844 @node Implementation Defined Aspects
7845 @chapter Implementation Defined Aspects
7846 Ada defines (throughout the Ada 2012 reference manual, summarized
7847 in Annex K) a set of aspects that can be specified for certain entities.
7848 These language defined aspects are implemented in GNAT in Ada 2012 mode
7849 and work as described in the Ada 2012 Reference Manual.
7851 In addition, Ada 2012 allows implementations to define additional aspects
7852 whose meaning is defined by the implementation. GNAT provides
7853 a number of these implementation-defined aspects which can be used
7854 to extend and enhance the functionality of the compiler. This section of
7855 the GNAT reference manual describes these additional aspects.
7857 Note that any program using these aspects may not be portable to
7858 other compilers (although GNAT implements this set of aspects on all
7859 platforms). Therefore if portability to other compilers is an important
7860 consideration, you should minimize the use of these aspects.
7862 Note that for many of these aspects, the effect is essentially similar
7863 to the use of a pragma or attribute specification with the same name
7864 applied to the entity. For example, if we write:
7866 @smallexample @c ada
7867 type R is range 1 .. 100
7868 with Value_Size => 10;
7872 then the effect is the same as:
7874 @smallexample @c ada
7875 type R is range 1 .. 100;
7876 for R'Value_Size use 10;
7882 @smallexample @c ada
7883 type R is new Integer
7884 with Shared => True;
7888 then the effect is the same as:
7890 @smallexample @c ada
7891 type R is new Integer;
7896 In the documentation below, such cases are simply marked
7897 as being equivalent to the corresponding pragma or attribute definition
7901 * Aspect Abstract_State::
7902 * Aspect Contract_Cases::
7904 * Aspect Dimension::
7905 * Aspect Dimension_System::
7906 * Aspect Favor_Top_Level::
7908 * Aspect Initial_Condition::
7909 * Aspect Initializes::
7910 * Aspect Inline_Always::
7911 * Aspect Invariant::
7912 * Aspect Linker_Section::
7913 * Aspect Lock_Free::
7914 * Aspect Object_Size::
7915 * Aspect Persistent_BSS::
7916 * Aspect Predicate::
7917 * Aspect Preelaborate_05::
7920 * Aspect Pure_Function::
7921 * Aspect Refined_State::
7922 * Aspect Remote_Access_Type::
7923 * Aspect Scalar_Storage_Order::
7925 * Aspect Simple_Storage_Pool::
7926 * Aspect Simple_Storage_Pool_Type::
7927 * Aspect SPARK_Mode::
7928 * Aspect Suppress_Debug_Info::
7929 * Aspect Test_Case::
7930 * Aspect Universal_Aliasing::
7931 * Aspect Universal_Data::
7932 * Aspect Unmodified::
7933 * Aspect Unreferenced::
7934 * Aspect Unreferenced_Objects::
7935 * Aspect Value_Size::
7939 @node Aspect Abstract_State
7940 @unnumberedsec Aspect Abstract_State
7941 @findex Abstract_State
7943 This aspect is equivalent to pragma @code{Abstract_State}.
7945 @node Aspect Contract_Cases
7946 @unnumberedsec Aspect Contract_Cases
7947 @findex Contract_Cases
7949 This aspect is equivalent to pragma @code{Contract_Cases}, the sequence
7950 of clauses being enclosed in parentheses so that syntactically it is an
7953 @node Aspect Depends
7954 @unnumberedsec Aspect Depends
7957 This aspect is equivalent to pragma @code{Depends}.
7959 @node Aspect Dimension
7960 @unnumberedsec Aspect Dimension
7963 The @code{Dimension} aspect is used to specify the dimensions of a given
7964 subtype of a dimensioned numeric type. The aspect also specifies a symbol
7965 used when doing formatted output of dimensioned quantities. The syntax is:
7967 @smallexample @c ada
7969 ([Symbol =>] SYMBOL, DIMENSION_VALUE @{, DIMENSION_Value@})
7971 SYMBOL ::= STRING_LITERAL | CHARACTER_LITERAL
7975 | others => RATIONAL
7976 | DISCRETE_CHOICE_LIST => RATIONAL
7978 RATIONAL ::= [-] NUMERIC_LITERAL [/ NUMERIC_LITERAL]
7982 This aspect can only be applied to a subtype whose parent type has
7983 a @code{Dimension_Systen} aspect. The aspect must specify values for
7984 all dimensions of the system. The rational values are the powers of the
7985 corresponding dimensions that are used by the compiler to verify that
7986 physical (numeric) computations are dimensionally consistent. For example,
7987 the computation of a force must result in dimensions (L => 1, M => 1, T => -2).
7988 For further examples of the usage
7989 of this aspect, see package @code{System.Dim.Mks}.
7990 Note that when the dimensioned type is an integer type, then any
7991 dimension value must be an integer literal.
7993 @node Aspect Dimension_System
7994 @unnumberedsec Aspect Dimension_System
7995 @findex Dimension_System
7997 The @code{Dimension_System} aspect is used to define a system of
7998 dimensions that will be used in subsequent subtype declarations with
7999 @code{Dimension} aspects that reference this system. The syntax is:
8001 @smallexample @c ada
8002 with Dimension_System => (DIMENSION @{, DIMENSION@});
8004 DIMENSION ::= ([Unit_Name =>] IDENTIFIER,
8005 [Unit_Symbol =>] SYMBOL,
8006 [Dim_Symbol =>] SYMBOL)
8008 SYMBOL ::= CHARACTER_LITERAL | STRING_LITERAL
8012 This aspect is applied to a type, which must be a numeric derived type
8013 (typically a floating-point type), that
8014 will represent values within the dimension system. Each @code{DIMENSION}
8015 corresponds to one particular dimension. A maximum of 7 dimensions may
8016 be specified. @code{Unit_Name} is the name of the dimension (for example
8017 @code{Meter}). @code{Unit_Symbol} is the shorthand used for quantities
8018 of this dimension (for example @code{m} for @code{Meter}).
8019 @code{Dim_Symbol} gives
8020 the identification within the dimension system (typically this is a
8021 single letter, e.g. @code{L} standing for length for unit name @code{Meter}).
8022 The @code{Unit_Symbol} is used in formatted output of dimensioned quantities.
8023 The @code{Dim_Symbol} is used in error messages when numeric operations have
8024 inconsistent dimensions.
8026 GNAT provides the standard definition of the International MKS system in
8027 the run-time package @code{System.Dim.Mks}. You can easily define
8028 similar packages for cgs units or British units, and define conversion factors
8029 between values in different systems. The MKS system is characterized by the
8032 @smallexample @c ada
8033 type Mks_Type is new Long_Long_Float
8035 Dimension_System => (
8036 (Unit_Name => Meter, Unit_Symbol => 'm', Dim_Symbol => 'L'),
8037 (Unit_Name => Kilogram, Unit_Symbol => "kg", Dim_Symbol => 'M'),
8038 (Unit_Name => Second, Unit_Symbol => 's', Dim_Symbol => 'T'),
8039 (Unit_Name => Ampere, Unit_Symbol => 'A', Dim_Symbol => 'I'),
8040 (Unit_Name => Kelvin, Unit_Symbol => 'K', Dim_Symbol => "Theta"),
8041 (Unit_Name => Mole, Unit_Symbol => "mol", Dim_Symbol => 'N'),
8042 (Unit_Name => Candela, Unit_Symbol => "cd", Dim_Symbol => 'J'));
8046 See section ``Performing Dimensionality Analysis in GNAT'' in the GNAT Users
8047 Guide for detailed examples of use of the dimension system.
8049 @node Aspect Favor_Top_Level
8050 @unnumberedsec Aspect Favor_Top_Level
8051 @findex Favor_Top_Level
8053 This aspect is equivalent to pragma @code{Favor_Top_Level}.
8056 @unnumberedsec Aspect Global
8059 This aspect is equivalent to pragma @code{Global}.
8061 @node Aspect Initial_Condition
8062 @unnumberedsec Aspect Initial_Condition
8063 @findex Initial_Condition
8065 This aspect is equivalent to pragma @code{Initial_Condition}.
8067 @node Aspect Initializes
8068 @unnumberedsec Aspect Initializes
8071 This aspect is equivalent to pragma @code{Initializes}.
8073 @node Aspect Inline_Always
8074 @unnumberedsec Aspect Inline_Always
8075 @findex Inline_Always
8077 This aspect is equivalent to pragma @code{Inline_Always}.
8079 @node Aspect Invariant
8080 @unnumberedsec Aspect Invariant
8083 This aspect is equivalent to pragma @code{Invariant}. It is a
8084 synonym for the language defined aspect @code{Type_Invariant} except
8085 that it is separately controllable using pragma @code{Assertion_Policy}.
8087 @node Aspect Linker_Section
8088 @unnumberedsec Aspect Linker_Section
8089 @findex Linker_Section
8091 This aspect is equivalent to an @code{Linker_Section} pragma.
8093 @node Aspect Lock_Free
8094 @unnumberedsec Aspect Lock_Free
8097 This aspect is equivalent to pragma @code{Lock_Free}.
8099 @node Aspect Object_Size
8100 @unnumberedsec Aspect Object_Size
8103 This aspect is equivalent to an @code{Object_Size} attribute definition
8106 @node Aspect Persistent_BSS
8107 @unnumberedsec Aspect Persistent_BSS
8108 @findex Persistent_BSS
8110 This aspect is equivalent to pragma @code{Persistent_BSS}.
8112 @node Aspect Predicate
8113 @unnumberedsec Aspect Predicate
8116 This aspect is equivalent to pragma @code{Predicate}. It is thus
8117 similar to the language defined aspects @code{Dynamic_Predicate}
8118 and @code{Static_Predicate} except that whether the resulting
8119 predicate is static or dynamic is controlled by the form of the
8120 expression. It is also separately controllable using pragma
8121 @code{Assertion_Policy}.
8123 @node Aspect Preelaborate_05
8124 @unnumberedsec Aspect Preelaborate_05
8125 @findex Preelaborate_05
8127 This aspect is equivalent to pragma @code{Preelaborate_05}.
8129 @node Aspect Pure_05
8130 @unnumberedsec Aspect Pure_05
8133 This aspect is equivalent to pragma @code{Pure_05}.
8135 @node Aspect Pure_12
8136 @unnumberedsec Aspect Pure_12
8139 This aspect is equivalent to pragma @code{Pure_12}.
8141 @node Aspect Pure_Function
8142 @unnumberedsec Aspect Pure_Function
8143 @findex Pure_Function
8145 This aspect is equivalent to pragma @code{Pure_Function}.
8147 @node Aspect Refined_State
8148 @unnumberedsec Aspect Refined_State
8149 @findex Refined_State
8151 This aspect is equivalent to pragma @code{Refined_State}.
8153 @node Aspect Remote_Access_Type
8154 @unnumberedsec Aspect Remote_Access_Type
8155 @findex Remote_Access_Type
8157 This aspect is equivalent to pragma @code{Remote_Access_Type}.
8159 @node Aspect Scalar_Storage_Order
8160 @unnumberedsec Aspect Scalar_Storage_Order
8161 @findex Scalar_Storage_Order
8163 This aspect is equivalent to a @code{Scalar_Storage_Order}
8164 attribute definition clause.
8167 @unnumberedsec Aspect Shared
8170 This aspect is equivalent to pragma @code{Shared}, and is thus a synonym
8171 for aspect @code{Atomic}.
8173 @node Aspect Simple_Storage_Pool
8174 @unnumberedsec Aspect Simple_Storage_Pool
8175 @findex Simple_Storage_Pool
8177 This aspect is equivalent to a @code{Simple_Storage_Pool}
8178 attribute definition clause.
8180 @node Aspect Simple_Storage_Pool_Type
8181 @unnumberedsec Aspect Simple_Storage_Pool_Type
8182 @findex Simple_Storage_Pool_Type
8184 This aspect is equivalent to pragma @code{Simple_Storage_Pool_Type}.
8186 @node Aspect SPARK_Mode
8187 @unnumberedsec Aspect SPARK_Mode
8190 This aspect is equivalent to pragma @code{SPARK_Mode} and
8191 may be specified for either or both of the specification and body
8192 of a subprogram or package.
8194 @node Aspect Suppress_Debug_Info
8195 @unnumberedsec Aspect Suppress_Debug_Info
8196 @findex Suppress_Debug_Info
8198 This aspect is equivalent to pragma @code{Suppress_Debug_Info}.
8200 @node Aspect Test_Case
8201 @unnumberedsec Aspect Test_Case
8204 This aspect is equivalent to pragma @code{Test_Case}.
8206 @node Aspect Universal_Aliasing
8207 @unnumberedsec Aspect Universal_Aliasing
8208 @findex Universal_Aliasing
8210 This aspect is equivalent to pragma @code{Universal_Aliasing}.
8212 @node Aspect Universal_Data
8213 @unnumberedsec Aspect Universal_Data
8214 @findex Universal_Data
8216 This aspect is equivalent to pragma @code{Universal_Data}.
8218 @node Aspect Unmodified
8219 @unnumberedsec Aspect Unmodified
8222 This aspect is equivalent to pragma @code{Unmodified}.
8224 @node Aspect Unreferenced
8225 @unnumberedsec Aspect Unreferenced
8226 @findex Unreferenced
8228 This aspect is equivalent to pragma @code{Unreferenced}.
8230 @node Aspect Unreferenced_Objects
8231 @unnumberedsec Aspect Unreferenced_Objects
8232 @findex Unreferenced_Objects
8234 This aspect is equivalent to pragma @code{Unreferenced_Objects}.
8236 @node Aspect Value_Size
8237 @unnumberedsec Aspect Value_Size
8240 This aspect is equivalent to a @code{Value_Size}
8241 attribute definition clause.
8243 @node Aspect Warnings
8244 @unnumberedsec Aspect Warnings
8247 This aspect is equivalent to the two argument form of pragma @code{Warnings},
8248 where the first argument is @code{ON} or @code{OFF} and the second argument
8252 @node Implementation Defined Attributes
8253 @chapter Implementation Defined Attributes
8254 Ada defines (throughout the Ada reference manual,
8255 summarized in Annex K),
8256 a set of attributes that provide useful additional functionality in all
8257 areas of the language. These language defined attributes are implemented
8258 in GNAT and work as described in the Ada Reference Manual.
8260 In addition, Ada allows implementations to define additional
8261 attributes whose meaning is defined by the implementation. GNAT provides
8262 a number of these implementation-dependent attributes which can be used
8263 to extend and enhance the functionality of the compiler. This section of
8264 the GNAT reference manual describes these additional attributes.
8266 Note that any program using these attributes may not be portable to
8267 other compilers (although GNAT implements this set of attributes on all
8268 platforms). Therefore if portability to other compilers is an important
8269 consideration, you should minimize the use of these attributes.
8272 * Attribute Abort_Signal::
8273 * Attribute Address_Size::
8274 * Attribute Asm_Input::
8275 * Attribute Asm_Output::
8276 * Attribute AST_Entry::
8278 * Attribute Bit_Position::
8279 * Attribute Compiler_Version::
8280 * Attribute Code_Address::
8281 * Attribute Default_Bit_Order::
8282 * Attribute Descriptor_Size::
8283 * Attribute Elaborated::
8284 * Attribute Elab_Body::
8285 * Attribute Elab_Spec::
8286 * Attribute Elab_Subp_Body::
8288 * Attribute Enabled::
8289 * Attribute Enum_Rep::
8290 * Attribute Enum_Val::
8291 * Attribute Epsilon::
8292 * Attribute Fixed_Value::
8293 * Attribute Has_Access_Values::
8294 * Attribute Has_Discriminants::
8296 * Attribute Integer_Value::
8297 * Attribute Invalid_Value::
8299 * Attribute Library_Level::
8300 * Attribute Loop_Entry::
8301 * Attribute Machine_Size::
8302 * Attribute Mantissa::
8303 * Attribute Max_Interrupt_Priority::
8304 * Attribute Max_Priority::
8305 * Attribute Maximum_Alignment::
8306 * Attribute Mechanism_Code::
8307 * Attribute Null_Parameter::
8308 * Attribute Object_Size::
8309 * Attribute Passed_By_Reference::
8310 * Attribute Pool_Address::
8311 * Attribute Range_Length::
8313 * Attribute Restriction_Set::
8314 * Attribute Result::
8315 * Attribute Safe_Emax::
8316 * Attribute Safe_Large::
8317 * Attribute Scalar_Storage_Order::
8318 * Attribute Simple_Storage_Pool::
8320 * Attribute Storage_Unit::
8321 * Attribute Stub_Type::
8322 * Attribute System_Allocator_Alignment::
8323 * Attribute Target_Name::
8325 * Attribute To_Address::
8326 * Attribute Type_Class::
8327 * Attribute UET_Address::
8328 * Attribute Unconstrained_Array::
8329 * Attribute Universal_Literal_String::
8330 * Attribute Unrestricted_Access::
8331 * Attribute Update::
8332 * Attribute Valid_Scalars::
8333 * Attribute VADS_Size::
8334 * Attribute Value_Size::
8335 * Attribute Wchar_T_Size::
8336 * Attribute Word_Size::
8339 @node Attribute Abort_Signal
8340 @unnumberedsec Attribute Abort_Signal
8341 @findex Abort_Signal
8343 @code{Standard'Abort_Signal} (@code{Standard} is the only allowed
8344 prefix) provides the entity for the special exception used to signal
8345 task abort or asynchronous transfer of control. Normally this attribute
8346 should only be used in the tasking runtime (it is highly peculiar, and
8347 completely outside the normal semantics of Ada, for a user program to
8348 intercept the abort exception).
8350 @node Attribute Address_Size
8351 @unnumberedsec Attribute Address_Size
8352 @cindex Size of @code{Address}
8353 @findex Address_Size
8355 @code{Standard'Address_Size} (@code{Standard} is the only allowed
8356 prefix) is a static constant giving the number of bits in an
8357 @code{Address}. It is the same value as System.Address'Size,
8358 but has the advantage of being static, while a direct
8359 reference to System.Address'Size is non-static because Address
8362 @node Attribute Asm_Input
8363 @unnumberedsec Attribute Asm_Input
8366 The @code{Asm_Input} attribute denotes a function that takes two
8367 parameters. The first is a string, the second is an expression of the
8368 type designated by the prefix. The first (string) argument is required
8369 to be a static expression, and is the constraint for the parameter,
8370 (e.g.@: what kind of register is required). The second argument is the
8371 value to be used as the input argument. The possible values for the
8372 constant are the same as those used in the RTL, and are dependent on
8373 the configuration file used to built the GCC back end.
8374 @ref{Machine Code Insertions}
8376 @node Attribute Asm_Output
8377 @unnumberedsec Attribute Asm_Output
8380 The @code{Asm_Output} attribute denotes a function that takes two
8381 parameters. The first is a string, the second is the name of a variable
8382 of the type designated by the attribute prefix. The first (string)
8383 argument is required to be a static expression and designates the
8384 constraint for the parameter (e.g.@: what kind of register is
8385 required). The second argument is the variable to be updated with the
8386 result. The possible values for constraint are the same as those used in
8387 the RTL, and are dependent on the configuration file used to build the
8388 GCC back end. If there are no output operands, then this argument may
8389 either be omitted, or explicitly given as @code{No_Output_Operands}.
8390 @ref{Machine Code Insertions}
8392 @node Attribute AST_Entry
8393 @unnumberedsec Attribute AST_Entry
8397 This attribute is implemented only in OpenVMS versions of GNAT@. Applied to
8398 the name of an entry, it yields a value of the predefined type AST_Handler
8399 (declared in the predefined package System, as extended by the use of
8400 pragma @code{Extend_System (Aux_DEC)}). This value enables the given entry to
8401 be called when an AST occurs. For further details, refer to the @cite{DEC Ada
8402 Language Reference Manual}, section 9.12a.
8405 @unnumberedsec Attribute Bit
8407 @code{@var{obj}'Bit}, where @var{obj} is any object, yields the bit
8408 offset within the storage unit (byte) that contains the first bit of
8409 storage allocated for the object. The value of this attribute is of the
8410 type @code{Universal_Integer}, and is always a non-negative number not
8411 exceeding the value of @code{System.Storage_Unit}.
8413 For an object that is a variable or a constant allocated in a register,
8414 the value is zero. (The use of this attribute does not force the
8415 allocation of a variable to memory).
8417 For an object that is a formal parameter, this attribute applies
8418 to either the matching actual parameter or to a copy of the
8419 matching actual parameter.
8421 For an access object the value is zero. Note that
8422 @code{@var{obj}.all'Bit} is subject to an @code{Access_Check} for the
8423 designated object. Similarly for a record component
8424 @code{@var{X}.@var{C}'Bit} is subject to a discriminant check and
8425 @code{@var{X}(@var{I}).Bit} and @code{@var{X}(@var{I1}..@var{I2})'Bit}
8426 are subject to index checks.
8428 This attribute is designed to be compatible with the DEC Ada 83 definition
8429 and implementation of the @code{Bit} attribute.
8431 @node Attribute Bit_Position
8432 @unnumberedsec Attribute Bit_Position
8433 @findex Bit_Position
8435 @code{@var{R.C}'Bit_Position}, where @var{R} is a record object and C is one
8436 of the fields of the record type, yields the bit
8437 offset within the record contains the first bit of
8438 storage allocated for the object. The value of this attribute is of the
8439 type @code{Universal_Integer}. The value depends only on the field
8440 @var{C} and is independent of the alignment of
8441 the containing record @var{R}.
8443 @node Attribute Compiler_Version
8444 @unnumberedsec Attribute Compiler_Version
8445 @findex Compiler_Version
8447 @code{Standard'Compiler_Version} (@code{Standard} is the only allowed
8448 prefix) yields a static string identifying the version of the compiler
8449 being used to compile the unit containing the attribute reference. A
8450 typical result would be something like "@value{EDITION} @i{version} (20090221)".
8452 @node Attribute Code_Address
8453 @unnumberedsec Attribute Code_Address
8454 @findex Code_Address
8455 @cindex Subprogram address
8456 @cindex Address of subprogram code
8459 attribute may be applied to subprograms in Ada 95 and Ada 2005, but the
8460 intended effect seems to be to provide
8461 an address value which can be used to call the subprogram by means of
8462 an address clause as in the following example:
8464 @smallexample @c ada
8465 procedure K is @dots{}
8468 for L'Address use K'Address;
8469 pragma Import (Ada, L);
8473 A call to @code{L} is then expected to result in a call to @code{K}@.
8474 In Ada 83, where there were no access-to-subprogram values, this was
8475 a common work-around for getting the effect of an indirect call.
8476 GNAT implements the above use of @code{Address} and the technique
8477 illustrated by the example code works correctly.
8479 However, for some purposes, it is useful to have the address of the start
8480 of the generated code for the subprogram. On some architectures, this is
8481 not necessarily the same as the @code{Address} value described above.
8482 For example, the @code{Address} value may reference a subprogram
8483 descriptor rather than the subprogram itself.
8485 The @code{'Code_Address} attribute, which can only be applied to
8486 subprogram entities, always returns the address of the start of the
8487 generated code of the specified subprogram, which may or may not be
8488 the same value as is returned by the corresponding @code{'Address}
8491 @node Attribute Default_Bit_Order
8492 @unnumberedsec Attribute Default_Bit_Order
8494 @cindex Little endian
8495 @findex Default_Bit_Order
8497 @code{Standard'Default_Bit_Order} (@code{Standard} is the only
8498 permissible prefix), provides the value @code{System.Default_Bit_Order}
8499 as a @code{Pos} value (0 for @code{High_Order_First}, 1 for
8500 @code{Low_Order_First}). This is used to construct the definition of
8501 @code{Default_Bit_Order} in package @code{System}.
8503 @node Attribute Descriptor_Size
8504 @unnumberedsec Attribute Descriptor_Size
8507 @findex Descriptor_Size
8509 Non-static attribute @code{Descriptor_Size} returns the size in bits of the
8510 descriptor allocated for a type. The result is non-zero only for unconstrained
8511 array types and the returned value is of type universal integer. In GNAT, an
8512 array descriptor contains bounds information and is located immediately before
8513 the first element of the array.
8515 @smallexample @c ada
8516 type Unconstr_Array is array (Positive range <>) of Boolean;
8517 Put_Line ("Descriptor size = " & Unconstr_Array'Descriptor_Size'Img);
8521 The attribute takes into account any additional padding due to type alignment.
8522 In the example above, the descriptor contains two values of type
8523 @code{Positive} representing the low and high bound. Since @code{Positive} has
8524 a size of 31 bits and an alignment of 4, the descriptor size is @code{2 *
8525 Positive'Size + 2} or 64 bits.
8527 @node Attribute Elaborated
8528 @unnumberedsec Attribute Elaborated
8531 The prefix of the @code{'Elaborated} attribute must be a unit name. The
8532 value is a Boolean which indicates whether or not the given unit has been
8533 elaborated. This attribute is primarily intended for internal use by the
8534 generated code for dynamic elaboration checking, but it can also be used
8535 in user programs. The value will always be True once elaboration of all
8536 units has been completed. An exception is for units which need no
8537 elaboration, the value is always False for such units.
8539 @node Attribute Elab_Body
8540 @unnumberedsec Attribute Elab_Body
8543 This attribute can only be applied to a program unit name. It returns
8544 the entity for the corresponding elaboration procedure for elaborating
8545 the body of the referenced unit. This is used in the main generated
8546 elaboration procedure by the binder and is not normally used in any
8547 other context. However, there may be specialized situations in which it
8548 is useful to be able to call this elaboration procedure from Ada code,
8549 e.g.@: if it is necessary to do selective re-elaboration to fix some
8552 @node Attribute Elab_Spec
8553 @unnumberedsec Attribute Elab_Spec
8556 This attribute can only be applied to a program unit name. It returns
8557 the entity for the corresponding elaboration procedure for elaborating
8558 the spec of the referenced unit. This is used in the main
8559 generated elaboration procedure by the binder and is not normally used
8560 in any other context. However, there may be specialized situations in
8561 which it is useful to be able to call this elaboration procedure from
8562 Ada code, e.g.@: if it is necessary to do selective re-elaboration to fix
8565 @node Attribute Elab_Subp_Body
8566 @unnumberedsec Attribute Elab_Subp_Body
8567 @findex Elab_Subp_Body
8569 This attribute can only be applied to a library level subprogram
8570 name and is only allowed in CodePeer mode. It returns the entity
8571 for the corresponding elaboration procedure for elaborating the body
8572 of the referenced subprogram unit. This is used in the main generated
8573 elaboration procedure by the binder in CodePeer mode only and is unrecognized
8576 @node Attribute Emax
8577 @unnumberedsec Attribute Emax
8578 @cindex Ada 83 attributes
8581 The @code{Emax} attribute is provided for compatibility with Ada 83. See
8582 the Ada 83 reference manual for an exact description of the semantics of
8585 @node Attribute Enabled
8586 @unnumberedsec Attribute Enabled
8589 The @code{Enabled} attribute allows an application program to check at compile
8590 time to see if the designated check is currently enabled. The prefix is a
8591 simple identifier, referencing any predefined check name (other than
8592 @code{All_Checks}) or a check name introduced by pragma Check_Name. If
8593 no argument is given for the attribute, the check is for the general state
8594 of the check, if an argument is given, then it is an entity name, and the
8595 check indicates whether an @code{Suppress} or @code{Unsuppress} has been
8596 given naming the entity (if not, then the argument is ignored).
8598 Note that instantiations inherit the check status at the point of the
8599 instantiation, so a useful idiom is to have a library package that
8600 introduces a check name with @code{pragma Check_Name}, and then contains
8601 generic packages or subprograms which use the @code{Enabled} attribute
8602 to see if the check is enabled. A user of this package can then issue
8603 a @code{pragma Suppress} or @code{pragma Unsuppress} before instantiating
8604 the package or subprogram, controlling whether the check will be present.
8606 @node Attribute Enum_Rep
8607 @unnumberedsec Attribute Enum_Rep
8608 @cindex Representation of enums
8611 For every enumeration subtype @var{S}, @code{@var{S}'Enum_Rep} denotes a
8612 function with the following spec:
8614 @smallexample @c ada
8615 function @var{S}'Enum_Rep (Arg : @var{S}'Base)
8616 return @i{Universal_Integer};
8620 It is also allowable to apply @code{Enum_Rep} directly to an object of an
8621 enumeration type or to a non-overloaded enumeration
8622 literal. In this case @code{@var{S}'Enum_Rep} is equivalent to
8623 @code{@var{typ}'Enum_Rep(@var{S})} where @var{typ} is the type of the
8624 enumeration literal or object.
8626 The function returns the representation value for the given enumeration
8627 value. This will be equal to value of the @code{Pos} attribute in the
8628 absence of an enumeration representation clause. This is a static
8629 attribute (i.e.@: the result is static if the argument is static).
8631 @code{@var{S}'Enum_Rep} can also be used with integer types and objects,
8632 in which case it simply returns the integer value. The reason for this
8633 is to allow it to be used for @code{(<>)} discrete formal arguments in
8634 a generic unit that can be instantiated with either enumeration types
8635 or integer types. Note that if @code{Enum_Rep} is used on a modular
8636 type whose upper bound exceeds the upper bound of the largest signed
8637 integer type, and the argument is a variable, so that the universal
8638 integer calculation is done at run time, then the call to @code{Enum_Rep}
8639 may raise @code{Constraint_Error}.
8641 @node Attribute Enum_Val
8642 @unnumberedsec Attribute Enum_Val
8643 @cindex Representation of enums
8646 For every enumeration subtype @var{S}, @code{@var{S}'Enum_Val} denotes a
8647 function with the following spec:
8649 @smallexample @c ada
8650 function @var{S}'Enum_Val (Arg : @i{Universal_Integer)
8651 return @var{S}'Base};
8655 The function returns the enumeration value whose representation matches the
8656 argument, or raises Constraint_Error if no enumeration literal of the type
8657 has the matching value.
8658 This will be equal to value of the @code{Val} attribute in the
8659 absence of an enumeration representation clause. This is a static
8660 attribute (i.e.@: the result is static if the argument is static).
8662 @node Attribute Epsilon
8663 @unnumberedsec Attribute Epsilon
8664 @cindex Ada 83 attributes
8667 The @code{Epsilon} attribute is provided for compatibility with Ada 83. See
8668 the Ada 83 reference manual for an exact description of the semantics of
8671 @node Attribute Fixed_Value
8672 @unnumberedsec Attribute Fixed_Value
8675 For every fixed-point type @var{S}, @code{@var{S}'Fixed_Value} denotes a
8676 function with the following specification:
8678 @smallexample @c ada
8679 function @var{S}'Fixed_Value (Arg : @i{Universal_Integer})
8684 The value returned is the fixed-point value @var{V} such that
8686 @smallexample @c ada
8687 @var{V} = Arg * @var{S}'Small
8691 The effect is thus similar to first converting the argument to the
8692 integer type used to represent @var{S}, and then doing an unchecked
8693 conversion to the fixed-point type. The difference is
8694 that there are full range checks, to ensure that the result is in range.
8695 This attribute is primarily intended for use in implementation of the
8696 input-output functions for fixed-point values.
8698 @node Attribute Has_Access_Values
8699 @unnumberedsec Attribute Has_Access_Values
8700 @cindex Access values, testing for
8701 @findex Has_Access_Values
8703 The prefix of the @code{Has_Access_Values} attribute is a type. The result
8704 is a Boolean value which is True if the is an access type, or is a composite
8705 type with a component (at any nesting depth) that is an access type, and is
8707 The intended use of this attribute is in conjunction with generic
8708 definitions. If the attribute is applied to a generic private type, it
8709 indicates whether or not the corresponding actual type has access values.
8711 @node Attribute Has_Discriminants
8712 @unnumberedsec Attribute Has_Discriminants
8713 @cindex Discriminants, testing for
8714 @findex Has_Discriminants
8716 The prefix of the @code{Has_Discriminants} attribute is a type. The result
8717 is a Boolean value which is True if the type has discriminants, and False
8718 otherwise. The intended use of this attribute is in conjunction with generic
8719 definitions. If the attribute is applied to a generic private type, it
8720 indicates whether or not the corresponding actual type has discriminants.
8723 @unnumberedsec Attribute Img
8726 The @code{Img} attribute differs from @code{Image} in that it is applied
8727 directly to an object, and yields the same result as
8728 @code{Image} for the subtype of the object. This is convenient for
8731 @smallexample @c ada
8732 Put_Line ("X = " & X'Img);
8736 has the same meaning as the more verbose:
8738 @smallexample @c ada
8739 Put_Line ("X = " & @var{T}'Image (X));
8743 where @var{T} is the (sub)type of the object @code{X}.
8745 Note that technically, in analogy to @code{Image},
8746 @code{X'Img} returns a parameterless function
8747 that returns the appropriate string when called. This means that
8748 @code{X'Img} can be renamed as a function-returning-string, or used
8749 in an instantiation as a function parameter.
8751 @node Attribute Integer_Value
8752 @unnumberedsec Attribute Integer_Value
8753 @findex Integer_Value
8755 For every integer type @var{S}, @code{@var{S}'Integer_Value} denotes a
8756 function with the following spec:
8758 @smallexample @c ada
8759 function @var{S}'Integer_Value (Arg : @i{Universal_Fixed})
8764 The value returned is the integer value @var{V}, such that
8766 @smallexample @c ada
8767 Arg = @var{V} * @var{T}'Small
8771 where @var{T} is the type of @code{Arg}.
8772 The effect is thus similar to first doing an unchecked conversion from
8773 the fixed-point type to its corresponding implementation type, and then
8774 converting the result to the target integer type. The difference is
8775 that there are full range checks, to ensure that the result is in range.
8776 This attribute is primarily intended for use in implementation of the
8777 standard input-output functions for fixed-point values.
8779 @node Attribute Invalid_Value
8780 @unnumberedsec Attribute Invalid_Value
8781 @findex Invalid_Value
8783 For every scalar type S, S'Invalid_Value returns an undefined value of the
8784 type. If possible this value is an invalid representation for the type. The
8785 value returned is identical to the value used to initialize an otherwise
8786 uninitialized value of the type if pragma Initialize_Scalars is used,
8787 including the ability to modify the value with the binder -Sxx flag and
8788 relevant environment variables at run time.
8790 @node Attribute Large
8791 @unnumberedsec Attribute Large
8792 @cindex Ada 83 attributes
8795 The @code{Large} attribute is provided for compatibility with Ada 83. See
8796 the Ada 83 reference manual for an exact description of the semantics of
8799 @node Attribute Library_Level
8800 @unnumberedsec Attribute Library_Level
8801 @findex Library_Level
8804 @code{P'Library_Level}, where P is an entity name,
8805 returns a Boolean value which is True if the entity is declared
8806 at the library level, and False otherwise. Note that within a
8807 generic instantition, the name of the generic unit denotes the
8808 instance, which means that this attribute can be used to test
8809 if a generic is instantiated at the library level, as shown
8812 @smallexample @c ada
8816 pragma Compile_Time_Error
8817 (not Gen'Library_Level,
8818 "Gen can only be instantiated at library level");
8823 @node Attribute Loop_Entry
8824 @unnumberedsec Attribute Loop_Entry
8829 @smallexample @c ada
8830 X'Loop_Entry [(loop_name)]
8834 The @code{Loop_Entry} attribute is used to refer to the value that an
8835 expression had upon entry to a given loop in much the same way that the
8836 @code{Old} attribute in a subprogram postcondition can be used to refer
8837 to the value an expression had upon entry to the subprogram. The
8838 relevant loop is either identified by the given loop name, or it is the
8839 innermost enclosing loop when no loop name is given.
8842 A @code{Loop_Entry} attribute can only occur within a
8843 @code{Loop_Variant} or @code{Loop_Invariant} pragma. A common use of
8844 @code{Loop_Entry} is to compare the current value of objects with their
8845 initial value at loop entry, in a @code{Loop_Invariant} pragma.
8848 The effect of using @code{X'Loop_Entry} is the same as declaring
8849 a constant initialized with the initial value of @code{X} at loop
8850 entry. This copy is not performed if the loop is not entered, or if the
8851 corresponding pragmas are ignored or disabled.
8853 @node Attribute Machine_Size
8854 @unnumberedsec Attribute Machine_Size
8855 @findex Machine_Size
8857 This attribute is identical to the @code{Object_Size} attribute. It is
8858 provided for compatibility with the DEC Ada 83 attribute of this name.
8860 @node Attribute Mantissa
8861 @unnumberedsec Attribute Mantissa
8862 @cindex Ada 83 attributes
8865 The @code{Mantissa} attribute is provided for compatibility with Ada 83. See
8866 the Ada 83 reference manual for an exact description of the semantics of
8869 @node Attribute Max_Interrupt_Priority
8870 @unnumberedsec Attribute Max_Interrupt_Priority
8871 @cindex Interrupt priority, maximum
8872 @findex Max_Interrupt_Priority
8874 @code{Standard'Max_Interrupt_Priority} (@code{Standard} is the only
8875 permissible prefix), provides the same value as
8876 @code{System.Max_Interrupt_Priority}.
8878 @node Attribute Max_Priority
8879 @unnumberedsec Attribute Max_Priority
8880 @cindex Priority, maximum
8881 @findex Max_Priority
8883 @code{Standard'Max_Priority} (@code{Standard} is the only permissible
8884 prefix) provides the same value as @code{System.Max_Priority}.
8886 @node Attribute Maximum_Alignment
8887 @unnumberedsec Attribute Maximum_Alignment
8888 @cindex Alignment, maximum
8889 @findex Maximum_Alignment
8891 @code{Standard'Maximum_Alignment} (@code{Standard} is the only
8892 permissible prefix) provides the maximum useful alignment value for the
8893 target. This is a static value that can be used to specify the alignment
8894 for an object, guaranteeing that it is properly aligned in all
8897 @node Attribute Mechanism_Code
8898 @unnumberedsec Attribute Mechanism_Code
8899 @cindex Return values, passing mechanism
8900 @cindex Parameters, passing mechanism
8901 @findex Mechanism_Code
8903 @code{@var{function}'Mechanism_Code} yields an integer code for the
8904 mechanism used for the result of function, and
8905 @code{@var{subprogram}'Mechanism_Code (@var{n})} yields the mechanism
8906 used for formal parameter number @var{n} (a static integer value with 1
8907 meaning the first parameter) of @var{subprogram}. The code returned is:
8915 by descriptor (default descriptor class)
8917 by descriptor (UBS: unaligned bit string)
8919 by descriptor (UBSB: aligned bit string with arbitrary bounds)
8921 by descriptor (UBA: unaligned bit array)
8923 by descriptor (S: string, also scalar access type parameter)
8925 by descriptor (SB: string with arbitrary bounds)
8927 by descriptor (A: contiguous array)
8929 by descriptor (NCA: non-contiguous array)
8933 Values from 3 through 10 are only relevant to Digital OpenVMS implementations.
8936 @node Attribute Null_Parameter
8937 @unnumberedsec Attribute Null_Parameter
8938 @cindex Zero address, passing
8939 @findex Null_Parameter
8941 A reference @code{@var{T}'Null_Parameter} denotes an imaginary object of
8942 type or subtype @var{T} allocated at machine address zero. The attribute
8943 is allowed only as the default expression of a formal parameter, or as
8944 an actual expression of a subprogram call. In either case, the
8945 subprogram must be imported.
8947 The identity of the object is represented by the address zero in the
8948 argument list, independent of the passing mechanism (explicit or
8951 This capability is needed to specify that a zero address should be
8952 passed for a record or other composite object passed by reference.
8953 There is no way of indicating this without the @code{Null_Parameter}
8956 @node Attribute Object_Size
8957 @unnumberedsec Attribute Object_Size
8958 @cindex Size, used for objects
8961 The size of an object is not necessarily the same as the size of the type
8962 of an object. This is because by default object sizes are increased to be
8963 a multiple of the alignment of the object. For example,
8964 @code{Natural'Size} is
8965 31, but by default objects of type @code{Natural} will have a size of 32 bits.
8966 Similarly, a record containing an integer and a character:
8968 @smallexample @c ada
8976 will have a size of 40 (that is @code{Rec'Size} will be 40). The
8977 alignment will be 4, because of the
8978 integer field, and so the default size of record objects for this type
8979 will be 64 (8 bytes).
8981 If the alignment of the above record is specified to be 1, then the
8982 object size will be 40 (5 bytes). This is true by default, and also
8983 an object size of 40 can be explicitly specified in this case.
8985 A consequence of this capability is that different object sizes can be
8986 given to subtypes that would otherwise be considered in Ada to be
8987 statically matching. But it makes no sense to consider such subtypes
8988 as statically matching. Consequently, in @code{GNAT} we add a rule
8989 to the static matching rules that requires object sizes to match.
8990 Consider this example:
8992 @smallexample @c ada
8993 1. procedure BadAVConvert is
8994 2. type R is new Integer;
8995 3. subtype R1 is R range 1 .. 10;
8996 4. subtype R2 is R range 1 .. 10;
8997 5. for R1'Object_Size use 8;
8998 6. for R2'Object_Size use 16;
8999 7. type R1P is access all R1;
9000 8. type R2P is access all R2;
9001 9. R1PV : R1P := new R1'(4);
9004 12. R2PV := R2P (R1PV);
9006 >>> target designated subtype not compatible with
9007 type "R1" defined at line 3
9013 In the absence of lines 5 and 6,
9014 types @code{R1} and @code{R2} statically match and
9015 hence the conversion on line 12 is legal. But since lines 5 and 6
9016 cause the object sizes to differ, @code{GNAT} considers that types
9017 @code{R1} and @code{R2} are not statically matching, and line 12
9018 generates the diagnostic shown above.
9021 Similar additional checks are performed in other contexts requiring
9022 statically matching subtypes.
9024 @node Attribute Passed_By_Reference
9025 @unnumberedsec Attribute Passed_By_Reference
9026 @cindex Parameters, when passed by reference
9027 @findex Passed_By_Reference
9029 @code{@var{type}'Passed_By_Reference} for any subtype @var{type} returns
9030 a value of type @code{Boolean} value that is @code{True} if the type is
9031 normally passed by reference and @code{False} if the type is normally
9032 passed by copy in calls. For scalar types, the result is always @code{False}
9033 and is static. For non-scalar types, the result is non-static.
9035 @node Attribute Pool_Address
9036 @unnumberedsec Attribute Pool_Address
9037 @cindex Parameters, when passed by reference
9038 @findex Pool_Address
9040 @code{@var{X}'Pool_Address} for any object @var{X} returns the address
9041 of X within its storage pool. This is the same as
9042 @code{@var{X}'Address}, except that for an unconstrained array whose
9043 bounds are allocated just before the first component,
9044 @code{@var{X}'Pool_Address} returns the address of those bounds,
9045 whereas @code{@var{X}'Address} returns the address of the first
9048 Here, we are interpreting ``storage pool'' broadly to mean ``wherever
9049 the object is allocated'', which could be a user-defined storage pool,
9050 the global heap, on the stack, or in a static memory area. For an
9051 object created by @code{new}, @code{@var{Ptr.all}'Pool_Address} is
9052 what is passed to @code{Allocate} and returned from @code{Deallocate}.
9054 @node Attribute Range_Length
9055 @unnumberedsec Attribute Range_Length
9056 @findex Range_Length
9058 @code{@var{type}'Range_Length} for any discrete type @var{type} yields
9059 the number of values represented by the subtype (zero for a null
9060 range). The result is static for static subtypes. @code{Range_Length}
9061 applied to the index subtype of a one dimensional array always gives the
9062 same result as @code{Length} applied to the array itself.
9065 @unnumberedsec Attribute Ref
9070 @node Attribute Restriction_Set
9071 @unnumberedsec Attribute Restriction_Set
9072 @findex Restriction_Set
9073 @cindex Restrictions
9075 This attribute allows compile time testing of restrictions that
9076 are currently in effect. It is primarily intended for specializing
9077 code in the run-time based on restrictions that are active (e.g.
9078 don't need to save fpt registers if restriction No_Floating_Point
9079 is known to be in effect), but can be used anywhere.
9081 There are two forms:
9083 @smallexample @c ada
9084 System'Restriction_Set (partition_boolean_restriction_NAME)
9085 System'Restriction_Set (No_Dependence => library_unit_NAME);
9089 In the case of the first form, the only restriction names
9090 allowed are parameterless restrictions that are checked
9091 for consistency at bind time. For a complete list see the
9092 subtype @code{System.Rident.Partition_Boolean_Restrictions}.
9094 The result returned is True if the restriction is known to
9095 be in effect, and False if the restriction is known not to
9096 be in effect. An important guarantee is that the value of
9097 a Restriction_Set attribute is known to be consistent throughout
9098 all the code of a partition.
9100 This is trivially achieved if the entire partition is compiled
9101 with a consistent set of restriction pragmas. However, the
9102 compilation model does not require this. It is possible to
9103 compile one set of units with one set of pragmas, and another
9104 set of units with another set of pragmas. It is even possible
9105 to compile a spec with one set of pragmas, and then WITH the
9106 same spec with a different set of pragmas. Inconsistencies
9107 in the actual use of the restriction are checked at bind time.
9109 In order to achieve the guarantee of consistency for the
9110 Restriction_Set pragma, we consider that a use of the pragma
9111 that yields False is equivalent to a violation of the
9114 So for example if you write
9116 @smallexample @c ada
9117 if System'Restriction_Set (No_Floating_Point) then
9125 And the result is False, so that the else branch is executed,
9126 you can assume that this restriction is not set for any unit
9127 in the partition. This is checked by considering this use of
9128 the restriction pragma to be a violation of the restriction
9129 No_Floating_Point. This means that no other unit can attempt
9130 to set this restriction (if some unit does attempt to set it,
9131 the binder will refuse to bind the partition).
9133 Technical note: The restriction name and the unit name are
9134 intepreted entirely syntactically, as in the corresponding
9135 Restrictions pragma, they are not analyzed semantically,
9136 so they do not have a type.
9138 @node Attribute Result
9139 @unnumberedsec Attribute Result
9142 @code{@var{function}'Result} can only be used with in a Postcondition pragma
9143 for a function. The prefix must be the name of the corresponding function. This
9144 is used to refer to the result of the function in the postcondition expression.
9145 For a further discussion of the use of this attribute and examples of its use,
9146 see the description of pragma Postcondition.
9148 @node Attribute Safe_Emax
9149 @unnumberedsec Attribute Safe_Emax
9150 @cindex Ada 83 attributes
9153 The @code{Safe_Emax} attribute is provided for compatibility with Ada 83. See
9154 the Ada 83 reference manual for an exact description of the semantics of
9157 @node Attribute Safe_Large
9158 @unnumberedsec Attribute Safe_Large
9159 @cindex Ada 83 attributes
9162 The @code{Safe_Large} attribute is provided for compatibility with Ada 83. See
9163 the Ada 83 reference manual for an exact description of the semantics of
9166 @node Attribute Scalar_Storage_Order
9167 @unnumberedsec Attribute Scalar_Storage_Order
9169 @cindex Scalar storage order
9170 @findex Scalar_Storage_Order
9172 For every array or record type @var{S}, the representation attribute
9173 @code{Scalar_Storage_Order} denotes the order in which storage elements
9174 that make up scalar components are ordered within S:
9176 @smallexample @c ada
9177 -- Component type definitions
9179 subtype Yr_Type is Natural range 0 .. 127;
9180 subtype Mo_Type is Natural range 1 .. 12;
9181 subtype Da_Type is Natural range 1 .. 31;
9183 -- Record declaration
9186 Years_Since_1980 : Yr_Type;
9188 Day_Of_Month : Da_Type;
9191 -- Record representation clause
9194 Years_Since_1980 at 0 range 0 .. 6;
9195 Month at 0 range 7 .. 10;
9196 Day_Of_Month at 0 range 11 .. 15;
9199 -- Attribute definition clauses
9201 for Date'Bit_Order use System.High_Order_First;
9202 for Date'Scalar_Storage_Order use System.High_Order_First;
9203 -- If Scalar_Storage_Order is specified, it must be consistent with
9204 -- Bit_Order, so it's best to always define the latter explicitly if
9205 -- the former is used.
9208 Other properties are
9209 as for standard representation attribute @code{Bit_Order}, as defined by
9210 Ada RM 13.5.3(4). The default is @code{System.Default_Bit_Order}.
9212 For a record type @var{S}, if @code{@var{S}'Scalar_Storage_Order} is
9213 specified explicitly, it shall be equal to @code{@var{S}'Bit_Order}. Note:
9214 this means that if a @code{Scalar_Storage_Order} attribute definition
9215 clause is not confirming, then the type's @code{Bit_Order} shall be
9216 specified explicitly and set to the same value.
9218 For a record extension, the derived type shall have the same scalar storage
9219 order as the parent type.
9221 If a component of @var{S} has itself a record or array type, then it shall also
9222 have a @code{Scalar_Storage_Order} attribute definition clause. In addition,
9223 if the component is a packed array, or does not start on a byte boundary, then
9224 the scalar storage order specified for S and for the nested component type shall
9227 If @var{S} appears as the type of a record or array component, the enclosing
9228 record or array shall also have a @code{Scalar_Storage_Order} attribute
9231 No component of a type that has a @code{Scalar_Storage_Order} attribute
9232 definition may be aliased.
9234 A confirming @code{Scalar_Storage_Order} attribute definition clause (i.e.
9235 with a value equal to @code{System.Default_Bit_Order}) has no effect.
9237 If the opposite storage order is specified, then whenever the value of
9238 a scalar component of an object of type @var{S} is read, the storage
9239 elements of the enclosing machine scalar are first reversed (before
9240 retrieving the component value, possibly applying some shift and mask
9241 operatings on the enclosing machine scalar), and the opposite operation
9244 In that case, the restrictions set forth in 13.5.1(10.3/2) for scalar components
9245 are relaxed. Instead, the following rules apply:
9248 @item the underlying storage elements are those at positions
9249 @code{(position + first_bit / storage_element_size) ..
9250 (position + (last_bit + storage_element_size - 1) /
9251 storage_element_size)}
9252 @item the sequence of underlying storage elements shall have
9253 a size no greater than the largest machine scalar
9254 @item the enclosing machine scalar is defined as the smallest machine
9255 scalar starting at a position no greater than
9256 @code{position + first_bit / storage_element_size} and covering
9257 storage elements at least up to @code{position + (last_bit +
9258 storage_element_size - 1) / storage_element_size}
9259 @item the position of the component is interpreted relative to that machine
9264 @node Attribute Simple_Storage_Pool
9265 @unnumberedsec Attribute Simple_Storage_Pool
9266 @cindex Storage pool, simple
9267 @cindex Simple storage pool
9268 @findex Simple_Storage_Pool
9270 For every nonformal, nonderived access-to-object type @var{Acc}, the
9271 representation attribute @code{Simple_Storage_Pool} may be specified
9272 via an attribute_definition_clause (or by specifying the equivalent aspect):
9274 @smallexample @c ada
9276 My_Pool : My_Simple_Storage_Pool_Type;
9278 type Acc is access My_Data_Type;
9280 for Acc'Simple_Storage_Pool use My_Pool;
9285 The name given in an attribute_definition_clause for the
9286 @code{Simple_Storage_Pool} attribute shall denote a variable of
9287 a ``simple storage pool type'' (see pragma @code{Simple_Storage_Pool_Type}).
9289 The use of this attribute is only allowed for a prefix denoting a type
9290 for which it has been specified. The type of the attribute is the type
9291 of the variable specified as the simple storage pool of the access type,
9292 and the attribute denotes that variable.
9294 It is illegal to specify both @code{Storage_Pool} and @code{Simple_Storage_Pool}
9295 for the same access type.
9297 If the @code{Simple_Storage_Pool} attribute has been specified for an access
9298 type, then applying the @code{Storage_Pool} attribute to the type is flagged
9299 with a warning and its evaluation raises the exception @code{Program_Error}.
9301 If the Simple_Storage_Pool attribute has been specified for an access
9302 type @var{S}, then the evaluation of the attribute @code{@var{S}'Storage_Size}
9303 returns the result of calling @code{Storage_Size (@var{S}'Simple_Storage_Pool)},
9304 which is intended to indicate the number of storage elements reserved for
9305 the simple storage pool. If the Storage_Size function has not been defined
9306 for the simple storage pool type, then this attribute returns zero.
9308 If an access type @var{S} has a specified simple storage pool of type
9309 @var{SSP}, then the evaluation of an allocator for that access type calls
9310 the primitive @code{Allocate} procedure for type @var{SSP}, passing
9311 @code{@var{S}'Simple_Storage_Pool} as the pool parameter. The detailed
9312 semantics of such allocators is the same as those defined for allocators
9313 in section 13.11 of the Ada Reference Manual, with the term
9314 ``simple storage pool'' substituted for ``storage pool''.
9316 If an access type @var{S} has a specified simple storage pool of type
9317 @var{SSP}, then a call to an instance of the @code{Ada.Unchecked_Deallocation}
9318 for that access type invokes the primitive @code{Deallocate} procedure
9319 for type @var{SSP}, passing @code{@var{S}'Simple_Storage_Pool} as the pool
9320 parameter. The detailed semantics of such unchecked deallocations is the same
9321 as defined in section 13.11.2 of the Ada Reference Manual, except that the
9322 term ``simple storage pool'' is substituted for ``storage pool''.
9324 @node Attribute Small
9325 @unnumberedsec Attribute Small
9326 @cindex Ada 83 attributes
9329 The @code{Small} attribute is defined in Ada 95 (and Ada 2005) only for
9331 GNAT also allows this attribute to be applied to floating-point types
9332 for compatibility with Ada 83. See
9333 the Ada 83 reference manual for an exact description of the semantics of
9334 this attribute when applied to floating-point types.
9336 @node Attribute Storage_Unit
9337 @unnumberedsec Attribute Storage_Unit
9338 @findex Storage_Unit
9340 @code{Standard'Storage_Unit} (@code{Standard} is the only permissible
9341 prefix) provides the same value as @code{System.Storage_Unit}.
9343 @node Attribute Stub_Type
9344 @unnumberedsec Attribute Stub_Type
9347 The GNAT implementation of remote access-to-classwide types is
9348 organized as described in AARM section E.4 (20.t): a value of an RACW type
9349 (designating a remote object) is represented as a normal access
9350 value, pointing to a "stub" object which in turn contains the
9351 necessary information to contact the designated remote object. A
9352 call on any dispatching operation of such a stub object does the
9353 remote call, if necessary, using the information in the stub object
9354 to locate the target partition, etc.
9356 For a prefix @code{T} that denotes a remote access-to-classwide type,
9357 @code{T'Stub_Type} denotes the type of the corresponding stub objects.
9359 By construction, the layout of @code{T'Stub_Type} is identical to that of
9360 type @code{RACW_Stub_Type} declared in the internal implementation-defined
9361 unit @code{System.Partition_Interface}. Use of this attribute will create
9362 an implicit dependency on this unit.
9364 @node Attribute System_Allocator_Alignment
9365 @unnumberedsec Attribute System_Allocator_Alignment
9366 @cindex Alignment, allocator
9367 @findex System_Allocator_Alignment
9369 @code{Standard'System_Allocator_Alignment} (@code{Standard} is the only
9370 permissible prefix) provides the observable guaranted to be honored by
9371 the system allocator (malloc). This is a static value that can be used
9372 in user storage pools based on malloc either to reject allocation
9373 with alignment too large or to enable a realignment circuitry if the
9374 alignment request is larger than this value.
9376 @node Attribute Target_Name
9377 @unnumberedsec Attribute Target_Name
9380 @code{Standard'Target_Name} (@code{Standard} is the only permissible
9381 prefix) provides a static string value that identifies the target
9382 for the current compilation. For GCC implementations, this is the
9383 standard gcc target name without the terminating slash (for
9384 example, GNAT 5.0 on windows yields "i586-pc-mingw32msv").
9386 @node Attribute Tick
9387 @unnumberedsec Attribute Tick
9390 @code{Standard'Tick} (@code{Standard} is the only permissible prefix)
9391 provides the same value as @code{System.Tick},
9393 @node Attribute To_Address
9394 @unnumberedsec Attribute To_Address
9397 The @code{System'To_Address}
9398 (@code{System} is the only permissible prefix)
9399 denotes a function identical to
9400 @code{System.Storage_Elements.To_Address} except that
9401 it is a static attribute. This means that if its argument is
9402 a static expression, then the result of the attribute is a
9403 static expression. This means that such an expression can be
9404 used in contexts (e.g.@: preelaborable packages) which require a
9405 static expression and where the function call could not be used
9406 (since the function call is always non-static, even if its
9407 argument is static). The argument must be in the range
9408 -(2**(m-1) .. 2**m-1, where m is the memory size
9409 (typically 32 or 64). Negative values are intepreted in a
9410 modular manner (e.g. -1 means the same as 16#FFFF_FFFF# on
9413 @node Attribute Type_Class
9414 @unnumberedsec Attribute Type_Class
9417 @code{@var{type}'Type_Class} for any type or subtype @var{type} yields
9418 the value of the type class for the full type of @var{type}. If
9419 @var{type} is a generic formal type, the value is the value for the
9420 corresponding actual subtype. The value of this attribute is of type
9421 @code{System.Aux_DEC.Type_Class}, which has the following definition:
9423 @smallexample @c ada
9425 (Type_Class_Enumeration,
9427 Type_Class_Fixed_Point,
9428 Type_Class_Floating_Point,
9433 Type_Class_Address);
9437 Protected types yield the value @code{Type_Class_Task}, which thus
9438 applies to all concurrent types. This attribute is designed to
9439 be compatible with the DEC Ada 83 attribute of the same name.
9441 @node Attribute UET_Address
9442 @unnumberedsec Attribute UET_Address
9445 The @code{UET_Address} attribute can only be used for a prefix which
9446 denotes a library package. It yields the address of the unit exception
9447 table when zero cost exception handling is used. This attribute is
9448 intended only for use within the GNAT implementation. See the unit
9449 @code{Ada.Exceptions} in files @file{a-except.ads} and @file{a-except.adb}
9450 for details on how this attribute is used in the implementation.
9452 @node Attribute Unconstrained_Array
9453 @unnumberedsec Attribute Unconstrained_Array
9454 @findex Unconstrained_Array
9456 The @code{Unconstrained_Array} attribute can be used with a prefix that
9457 denotes any type or subtype. It is a static attribute that yields
9458 @code{True} if the prefix designates an unconstrained array,
9459 and @code{False} otherwise. In a generic instance, the result is
9460 still static, and yields the result of applying this test to the
9463 @node Attribute Universal_Literal_String
9464 @unnumberedsec Attribute Universal_Literal_String
9465 @cindex Named numbers, representation of
9466 @findex Universal_Literal_String
9468 The prefix of @code{Universal_Literal_String} must be a named
9469 number. The static result is the string consisting of the characters of
9470 the number as defined in the original source. This allows the user
9471 program to access the actual text of named numbers without intermediate
9472 conversions and without the need to enclose the strings in quotes (which
9473 would preclude their use as numbers).
9475 For example, the following program prints the first 50 digits of pi:
9477 @smallexample @c ada
9478 with Text_IO; use Text_IO;
9482 Put (Ada.Numerics.Pi'Universal_Literal_String);
9486 @node Attribute Unrestricted_Access
9487 @unnumberedsec Attribute Unrestricted_Access
9488 @cindex @code{Access}, unrestricted
9489 @findex Unrestricted_Access
9491 The @code{Unrestricted_Access} attribute is similar to @code{Access}
9492 except that all accessibility and aliased view checks are omitted. This
9493 is a user-beware attribute. It is similar to
9494 @code{Address}, for which it is a desirable replacement where the value
9495 desired is an access type. In other words, its effect is identical to
9496 first applying the @code{Address} attribute and then doing an unchecked
9497 conversion to a desired access type. In GNAT, but not necessarily in
9498 other implementations, the use of static chains for inner level
9499 subprograms means that @code{Unrestricted_Access} applied to a
9500 subprogram yields a value that can be called as long as the subprogram
9501 is in scope (normal Ada accessibility rules restrict this usage).
9503 It is possible to use @code{Unrestricted_Access} for any type, but care
9504 must be exercised if it is used to create pointers to unconstrained
9505 objects. In this case, the resulting pointer has the same scope as the
9506 context of the attribute, and may not be returned to some enclosing
9507 scope. For instance, a function cannot use @code{Unrestricted_Access}
9508 to create a unconstrained pointer and then return that value to the
9511 @node Attribute Update
9512 @unnumberedsec Attribute Update
9515 The @code{Update} attribute creates a copy of an array or record value
9516 with one or more modified components. The syntax is:
9518 @smallexample @c ada
9519 PREFIX'Update ( RECORD_COMPONENT_ASSOCIATION_LIST )
9520 PREFIX'Update ( ARRAY_COMPONENT_ASSOCIATION @{, ARRAY_COMPONENT_ASSOCIATION @} )
9521 PREFIX'Update ( MULTIDIMENSIONAL_ARRAY_COMPONENT_ASSOCIATION
9522 @{, MULTIDIMENSIONAL_ARRAY_COMPONENT_ASSOCIATION @} )
9524 MULTIDIMENSIONAL_ARRAY_COMPONENT_ASSOCIATION ::= INDEX_EXPRESSION_LIST_LIST => EXPRESSION
9525 INDEX_EXPRESSION_LIST_LIST ::= INDEX_EXPRESSION_LIST @{| INDEX_EXPRESSION_LIST @}
9526 INDEX_EXPRESSION_LIST ::= ( EXPRESSION @{, EXPRESSION @} )
9530 where @code{PREFIX} is the name of an array or record object, and
9531 the association list in parentheses does not contain an @code{others}
9532 choice. The effect is to yield a copy of the array or record value which
9533 is unchanged apart from the components mentioned in the association list, which
9534 are changed to the indicated value. The original value of the array or
9535 record value is not affected. For example:
9537 @smallexample @c ada
9538 type Arr is Array (1 .. 5) of Integer;
9540 Avar1 : Arr := (1,2,3,4,5);
9541 Avar2 : Arr := Avar1'Update (2 => 10, 3 .. 4 => 20);
9545 yields a value for @code{Avar2} of 1,10,20,20,5 with @code{Avar1}
9546 begin unmodified. Similarly:
9548 @smallexample @c ada
9549 type Rec is A, B, C : Integer;
9551 Rvar1 : Rec := (A => 1, B => 2, C => 3);
9552 Rvar2 : Rec := Rvar1'Update (B => 20);
9556 yields a value for @code{Rvar2} of (A => 1, B => 20, C => 3),
9557 with @code{Rvar1} being unmodifed.
9558 Note that the value of the attribute reference is computed
9559 completely before it is used. This means that if you write:
9561 @smallexample @c ada
9562 Avar1 := Avar1'Update (1 => 10, 2 => Function_Call);
9566 then the value of @code{Avar1} is not modified if @code{Function_Call}
9567 raises an exception, unlike the effect of a series of direct assignments
9568 to elements of @code{Avar1}. In general this requires that
9569 two extra complete copies of the object are required, which should be
9570 kept in mind when considering efficiency.
9572 The @code{Update} attribute cannot be applied to prefixes of a limited
9573 type, and cannot reference discriminants in the case of a record type.
9574 The accessibility level of an Update attribute result object is defined
9575 as for an aggregate.
9577 In the record case, no component can be mentioned more than once. In
9578 the array case, two overlapping ranges can appear in the association list,
9579 in which case the modifications are processed left to right.
9581 Multi-dimensional arrays can be modified, as shown by this example:
9583 @smallexample @c ada
9584 A : array (1 .. 10, 1 .. 10) of Integer;
9586 A := A'Update ((1, 2) => 20, (3, 4) => 30);
9590 which changes element (1,2) to 20 and (3,4) to 30.
9592 @node Attribute Valid_Scalars
9593 @unnumberedsec Attribute Valid_Scalars
9594 @findex Valid_Scalars
9596 The @code{'Valid_Scalars} attribute is intended to make it easier to
9597 check the validity of scalar subcomponents of composite objects. It
9598 is defined for any prefix @code{X} that denotes an object.
9599 The value of this attribute is of the predefined type Boolean.
9600 @code{X'Valid_Scalars} yields True if and only if evaluation of
9601 @code{P'Valid} yields True for every scalar part P of X or if X has
9602 no scalar parts. It is not specified in what order the scalar parts
9603 are checked, nor whether any more are checked after any one of them
9604 is determined to be invalid. If the prefix @code{X} is of a class-wide
9605 type @code{T'Class} (where @code{T} is the associated specific type),
9606 or if the prefix @code{X} is of a specific tagged type @code{T}, then
9607 only the scalar parts of components of @code{T} are traversed; in other
9608 words, components of extensions of @code{T} are not traversed even if
9609 @code{T'Class (X)'Tag /= T'Tag} . The compiler will issue a warning if it can
9610 be determined at compile time that the prefix of the attribute has no
9611 scalar parts (e.g., if the prefix is of an access type, an interface type,
9612 an undiscriminated task type, or an undiscriminated protected type).
9614 @node Attribute VADS_Size
9615 @unnumberedsec Attribute VADS_Size
9616 @cindex @code{Size}, VADS compatibility
9619 The @code{'VADS_Size} attribute is intended to make it easier to port
9620 legacy code which relies on the semantics of @code{'Size} as implemented
9621 by the VADS Ada 83 compiler. GNAT makes a best effort at duplicating the
9622 same semantic interpretation. In particular, @code{'VADS_Size} applied
9623 to a predefined or other primitive type with no Size clause yields the
9624 Object_Size (for example, @code{Natural'Size} is 32 rather than 31 on
9625 typical machines). In addition @code{'VADS_Size} applied to an object
9626 gives the result that would be obtained by applying the attribute to
9627 the corresponding type.
9629 @node Attribute Value_Size
9630 @unnumberedsec Attribute Value_Size
9631 @cindex @code{Size}, setting for not-first subtype
9633 @code{@var{type}'Value_Size} is the number of bits required to represent
9634 a value of the given subtype. It is the same as @code{@var{type}'Size},
9635 but, unlike @code{Size}, may be set for non-first subtypes.
9637 @node Attribute Wchar_T_Size
9638 @unnumberedsec Attribute Wchar_T_Size
9639 @findex Wchar_T_Size
9640 @code{Standard'Wchar_T_Size} (@code{Standard} is the only permissible
9641 prefix) provides the size in bits of the C @code{wchar_t} type
9642 primarily for constructing the definition of this type in
9643 package @code{Interfaces.C}.
9645 @node Attribute Word_Size
9646 @unnumberedsec Attribute Word_Size
9648 @code{Standard'Word_Size} (@code{Standard} is the only permissible
9649 prefix) provides the value @code{System.Word_Size}.
9651 @node Standard and Implementation Defined Restrictions
9652 @chapter Standard and Implementation Defined Restrictions
9655 All RM defined Restriction identifiers are implemented:
9658 @item language-defined restrictions (see 13.12.1)
9659 @item tasking restrictions (see D.7)
9660 @item high integrity restrictions (see H.4)
9664 GNAT implements additional restriction identifiers. All restrictions, whether
9665 language defined or GNAT-specific, are listed in the following.
9668 * Partition-Wide Restrictions::
9669 * Program Unit Level Restrictions::
9672 @node Partition-Wide Restrictions
9673 @section Partition-Wide Restrictions
9675 There are two separate lists of restriction identifiers. The first
9676 set requires consistency throughout a partition (in other words, if the
9677 restriction identifier is used for any compilation unit in the partition,
9678 then all compilation units in the partition must obey the restriction).
9681 * Immediate_Reclamation::
9682 * Max_Asynchronous_Select_Nesting::
9683 * Max_Entry_Queue_Length::
9684 * Max_Protected_Entries::
9685 * Max_Select_Alternatives::
9686 * Max_Storage_At_Blocking::
9687 * Max_Task_Entries::
9689 * No_Abort_Statements::
9690 * No_Access_Parameter_Allocators::
9691 * No_Access_Subprograms::
9693 * No_Anonymous_Allocators::
9696 * No_Default_Initialization::
9699 * No_Direct_Boolean_Operators::
9701 * No_Dispatching_Calls::
9702 * No_Dynamic_Attachment::
9703 * No_Dynamic_Priorities::
9704 * No_Entry_Calls_In_Elaboration_Code::
9705 * No_Enumeration_Maps::
9706 * No_Exception_Handlers::
9707 * No_Exception_Propagation::
9708 * No_Exception_Registration::
9712 * No_Floating_Point::
9713 * No_Implicit_Conditionals::
9714 * No_Implicit_Dynamic_Code::
9715 * No_Implicit_Heap_Allocations::
9716 * No_Implicit_Loops::
9717 * No_Initialize_Scalars::
9719 * No_Local_Allocators::
9720 * No_Local_Protected_Objects::
9721 * No_Local_Timing_Events::
9722 * No_Nested_Finalization::
9723 * No_Protected_Type_Allocators::
9724 * No_Protected_Types::
9727 * No_Relative_Delay::
9728 * No_Requeue_Statements::
9729 * No_Secondary_Stack::
9730 * No_Select_Statements::
9731 * No_Specific_Termination_Handlers::
9732 * No_Specification_of_Aspect::
9733 * No_Standard_Allocators_After_Elaboration::
9734 * No_Standard_Storage_Pools::
9735 * No_Stream_Optimizations::
9737 * No_Task_Allocators::
9738 * No_Task_Attributes_Package::
9739 * No_Task_Hierarchy::
9740 * No_Task_Termination::
9742 * No_Terminate_Alternatives::
9743 * No_Unchecked_Access::
9745 * Static_Priorities::
9746 * Static_Storage_Size::
9749 @node Immediate_Reclamation
9750 @unnumberedsubsec Immediate_Reclamation
9751 @findex Immediate_Reclamation
9752 [RM H.4] This restriction ensures that, except for storage occupied by
9753 objects created by allocators and not deallocated via unchecked
9754 deallocation, any storage reserved at run time for an object is
9755 immediately reclaimed when the object no longer exists.
9757 @node Max_Asynchronous_Select_Nesting
9758 @unnumberedsubsec Max_Asynchronous_Select_Nesting
9759 @findex Max_Asynchronous_Select_Nesting
9760 [RM D.7] Specifies the maximum dynamic nesting level of asynchronous
9761 selects. Violations of this restriction with a value of zero are
9762 detected at compile time. Violations of this restriction with values
9763 other than zero cause Storage_Error to be raised.
9765 @node Max_Entry_Queue_Length
9766 @unnumberedsubsec Max_Entry_Queue_Length
9767 @findex Max_Entry_Queue_Length
9768 [RM D.7] This restriction is a declaration that any protected entry compiled in
9769 the scope of the restriction has at most the specified number of
9770 tasks waiting on the entry at any one time, and so no queue is required.
9771 Note that this restriction is checked at run time. Violation of this
9772 restriction results in the raising of Program_Error exception at the point of
9775 @findex Max_Entry_Queue_Depth
9776 The restriction @code{Max_Entry_Queue_Depth} is recognized as a
9777 synonym for @code{Max_Entry_Queue_Length}. This is retained for historical
9778 compatibility purposes (and a warning will be generated for its use if
9779 warnings on obsolescent features are activated).
9781 @node Max_Protected_Entries
9782 @unnumberedsubsec Max_Protected_Entries
9783 @findex Max_Protected_Entries
9784 [RM D.7] Specifies the maximum number of entries per protected type. The
9785 bounds of every entry family of a protected unit shall be static, or shall be
9786 defined by a discriminant of a subtype whose corresponding bound is static.
9788 @node Max_Select_Alternatives
9789 @unnumberedsubsec Max_Select_Alternatives
9790 @findex Max_Select_Alternatives
9791 [RM D.7] Specifies the maximum number of alternatives in a selective accept.
9793 @node Max_Storage_At_Blocking
9794 @unnumberedsubsec Max_Storage_At_Blocking
9795 @findex Max_Storage_At_Blocking
9796 [RM D.7] Specifies the maximum portion (in storage elements) of a task's
9797 Storage_Size that can be retained by a blocked task. A violation of this
9798 restriction causes Storage_Error to be raised.
9800 @node Max_Task_Entries
9801 @unnumberedsubsec Max_Task_Entries
9802 @findex Max_Task_Entries
9803 [RM D.7] Specifies the maximum number of entries
9804 per task. The bounds of every entry family
9805 of a task unit shall be static, or shall be
9806 defined by a discriminant of a subtype whose
9807 corresponding bound is static.
9810 @unnumberedsubsec Max_Tasks
9812 [RM D.7] Specifies the maximum number of task that may be created, not
9813 counting the creation of the environment task. Violations of this
9814 restriction with a value of zero are detected at compile
9815 time. Violations of this restriction with values other than zero cause
9816 Storage_Error to be raised.
9818 @node No_Abort_Statements
9819 @unnumberedsubsec No_Abort_Statements
9820 @findex No_Abort_Statements
9821 [RM D.7] There are no abort_statements, and there are
9822 no calls to Task_Identification.Abort_Task.
9824 @node No_Access_Parameter_Allocators
9825 @unnumberedsubsec No_Access_Parameter_Allocators
9826 @findex No_Access_Parameter_Allocators
9827 [RM H.4] This restriction ensures at compile time that there are no
9828 occurrences of an allocator as the actual parameter to an access
9831 @node No_Access_Subprograms
9832 @unnumberedsubsec No_Access_Subprograms
9833 @findex No_Access_Subprograms
9834 [RM H.4] This restriction ensures at compile time that there are no
9835 declarations of access-to-subprogram types.
9838 @unnumberedsubsec No_Allocators
9839 @findex No_Allocators
9840 [RM H.4] This restriction ensures at compile time that there are no
9841 occurrences of an allocator.
9843 @node No_Anonymous_Allocators
9844 @unnumberedsubsec No_Anonymous_Allocators
9845 @findex No_Anonymous_Allocators
9846 [RM H.4] This restriction ensures at compile time that there are no
9847 occurrences of an allocator of anonymous access type.
9850 @unnumberedsubsec No_Calendar
9852 [GNAT] This restriction ensures at compile time that there is no implicit or
9853 explicit dependence on the package @code{Ada.Calendar}.
9855 @node No_Coextensions
9856 @unnumberedsubsec No_Coextensions
9857 @findex No_Coextensions
9858 [RM H.4] This restriction ensures at compile time that there are no
9859 coextensions. See 3.10.2.
9861 @node No_Default_Initialization
9862 @unnumberedsubsec No_Default_Initialization
9863 @findex No_Default_Initialization
9865 [GNAT] This restriction prohibits any instance of default initialization
9866 of variables. The binder implements a consistency rule which prevents
9867 any unit compiled without the restriction from with'ing a unit with the
9868 restriction (this allows the generation of initialization procedures to
9869 be skipped, since you can be sure that no call is ever generated to an
9870 initialization procedure in a unit with the restriction active). If used
9871 in conjunction with Initialize_Scalars or Normalize_Scalars, the effect
9872 is to prohibit all cases of variables declared without a specific
9873 initializer (including the case of OUT scalar parameters).
9876 @unnumberedsubsec No_Delay
9878 [RM H.4] This restriction ensures at compile time that there are no
9879 delay statements and no dependences on package Calendar.
9882 @unnumberedsubsec No_Dependence
9883 @findex No_Dependence
9884 [RM 13.12.1] This restriction checks at compile time that there are no
9885 dependence on a library unit.
9887 @node No_Direct_Boolean_Operators
9888 @unnumberedsubsec No_Direct_Boolean_Operators
9889 @findex No_Direct_Boolean_Operators
9890 [GNAT] This restriction ensures that no logical operators (and/or/xor)
9891 are used on operands of type Boolean (or any type derived from Boolean).
9892 This is intended for use in safety critical programs where the certification
9893 protocol requires the use of short-circuit (and then, or else) forms for all
9894 composite boolean operations.
9897 @unnumberedsubsec No_Dispatch
9899 [RM H.4] This restriction ensures at compile time that there are no
9900 occurrences of @code{T'Class}, for any (tagged) subtype @code{T}.
9902 @node No_Dispatching_Calls
9903 @unnumberedsubsec No_Dispatching_Calls
9904 @findex No_Dispatching_Calls
9905 [GNAT] This restriction ensures at compile time that the code generated by the
9906 compiler involves no dispatching calls. The use of this restriction allows the
9907 safe use of record extensions, classwide membership tests and other classwide
9908 features not involving implicit dispatching. This restriction ensures that
9909 the code contains no indirect calls through a dispatching mechanism. Note that
9910 this includes internally-generated calls created by the compiler, for example
9911 in the implementation of class-wide objects assignments. The
9912 membership test is allowed in the presence of this restriction, because its
9913 implementation requires no dispatching.
9914 This restriction is comparable to the official Ada restriction
9915 @code{No_Dispatch} except that it is a bit less restrictive in that it allows
9916 all classwide constructs that do not imply dispatching.
9917 The following example indicates constructs that violate this restriction.
9921 type T is tagged record
9924 procedure P (X : T);
9926 type DT is new T with record
9927 More_Data : Natural;
9929 procedure Q (X : DT);
9933 procedure Example is
9934 procedure Test (O : T'Class) is
9935 N : Natural := O'Size;-- Error: Dispatching call
9936 C : T'Class := O; -- Error: implicit Dispatching Call
9938 if O in DT'Class then -- OK : Membership test
9939 Q (DT (O)); -- OK : Type conversion plus direct call
9941 P (O); -- Error: Dispatching call
9947 P (Obj); -- OK : Direct call
9948 P (T (Obj)); -- OK : Type conversion plus direct call
9949 P (T'Class (Obj)); -- Error: Dispatching call
9951 Test (Obj); -- OK : Type conversion
9953 if Obj in T'Class then -- OK : Membership test
9959 @node No_Dynamic_Attachment
9960 @unnumberedsubsec No_Dynamic_Attachment
9961 @findex No_Dynamic_Attachment
9962 [RM D.7] This restriction ensures that there is no call to any of the
9963 operations defined in package Ada.Interrupts
9964 (Is_Reserved, Is_Attached, Current_Handler, Attach_Handler, Exchange_Handler,
9965 Detach_Handler, and Reference).
9967 @findex No_Dynamic_Interrupts
9968 The restriction @code{No_Dynamic_Interrupts} is recognized as a
9969 synonym for @code{No_Dynamic_Attachment}. This is retained for historical
9970 compatibility purposes (and a warning will be generated for its use if
9971 warnings on obsolescent features are activated).
9973 @node No_Dynamic_Priorities
9974 @unnumberedsubsec No_Dynamic_Priorities
9975 @findex No_Dynamic_Priorities
9976 [RM D.7] There are no semantic dependencies on the package Dynamic_Priorities.
9978 @node No_Entry_Calls_In_Elaboration_Code
9979 @unnumberedsubsec No_Entry_Calls_In_Elaboration_Code
9980 @findex No_Entry_Calls_In_Elaboration_Code
9981 [GNAT] This restriction ensures at compile time that no task or protected entry
9982 calls are made during elaboration code. As a result of the use of this
9983 restriction, the compiler can assume that no code past an accept statement
9984 in a task can be executed at elaboration time.
9986 @node No_Enumeration_Maps
9987 @unnumberedsubsec No_Enumeration_Maps
9988 @findex No_Enumeration_Maps
9989 [GNAT] This restriction ensures at compile time that no operations requiring
9990 enumeration maps are used (that is Image and Value attributes applied
9991 to enumeration types).
9993 @node No_Exception_Handlers
9994 @unnumberedsubsec No_Exception_Handlers
9995 @findex No_Exception_Handlers
9996 [GNAT] This restriction ensures at compile time that there are no explicit
9997 exception handlers. It also indicates that no exception propagation will
9998 be provided. In this mode, exceptions may be raised but will result in
9999 an immediate call to the last chance handler, a routine that the user
10000 must define with the following profile:
10002 @smallexample @c ada
10003 procedure Last_Chance_Handler
10004 (Source_Location : System.Address; Line : Integer);
10005 pragma Export (C, Last_Chance_Handler,
10006 "__gnat_last_chance_handler");
10009 The parameter is a C null-terminated string representing a message to be
10010 associated with the exception (typically the source location of the raise
10011 statement generated by the compiler). The Line parameter when nonzero
10012 represents the line number in the source program where the raise occurs.
10014 @node No_Exception_Propagation
10015 @unnumberedsubsec No_Exception_Propagation
10016 @findex No_Exception_Propagation
10017 [GNAT] This restriction guarantees that exceptions are never propagated
10018 to an outer subprogram scope. The only case in which an exception may
10019 be raised is when the handler is statically in the same subprogram, so
10020 that the effect of a raise is essentially like a goto statement. Any
10021 other raise statement (implicit or explicit) will be considered
10022 unhandled. Exception handlers are allowed, but may not contain an
10023 exception occurrence identifier (exception choice). In addition, use of
10024 the package GNAT.Current_Exception is not permitted, and reraise
10025 statements (raise with no operand) are not permitted.
10027 @node No_Exception_Registration
10028 @unnumberedsubsec No_Exception_Registration
10029 @findex No_Exception_Registration
10030 [GNAT] This restriction ensures at compile time that no stream operations for
10031 types Exception_Id or Exception_Occurrence are used. This also makes it
10032 impossible to pass exceptions to or from a partition with this restriction
10033 in a distributed environment. If this exception is active, then the generated
10034 code is simplified by omitting the otherwise-required global registration
10035 of exceptions when they are declared.
10037 @node No_Exceptions
10038 @unnumberedsubsec No_Exceptions
10039 @findex No_Exceptions
10040 [RM H.4] This restriction ensures at compile time that there are no
10041 raise statements and no exception handlers.
10043 @node No_Finalization
10044 @unnumberedsubsec No_Finalization
10045 @findex No_Finalization
10046 [GNAT] This restriction disables the language features described in
10047 chapter 7.6 of the Ada 2005 RM as well as all form of code generation
10048 performed by the compiler to support these features. The following types
10049 are no longer considered controlled when this restriction is in effect:
10052 @code{Ada.Finalization.Controlled}
10054 @code{Ada.Finalization.Limited_Controlled}
10056 Derivations from @code{Controlled} or @code{Limited_Controlled}
10064 Array and record types with controlled components
10066 The compiler no longer generates code to initialize, finalize or adjust an
10067 object or a nested component, either declared on the stack or on the heap. The
10068 deallocation of a controlled object no longer finalizes its contents.
10070 @node No_Fixed_Point
10071 @unnumberedsubsec No_Fixed_Point
10072 @findex No_Fixed_Point
10073 [RM H.4] This restriction ensures at compile time that there are no
10074 occurrences of fixed point types and operations.
10076 @node No_Floating_Point
10077 @unnumberedsubsec No_Floating_Point
10078 @findex No_Floating_Point
10079 [RM H.4] This restriction ensures at compile time that there are no
10080 occurrences of floating point types and operations.
10082 @node No_Implicit_Conditionals
10083 @unnumberedsubsec No_Implicit_Conditionals
10084 @findex No_Implicit_Conditionals
10085 [GNAT] This restriction ensures that the generated code does not contain any
10086 implicit conditionals, either by modifying the generated code where possible,
10087 or by rejecting any construct that would otherwise generate an implicit
10088 conditional. Note that this check does not include run time constraint
10089 checks, which on some targets may generate implicit conditionals as
10090 well. To control the latter, constraint checks can be suppressed in the
10091 normal manner. Constructs generating implicit conditionals include comparisons
10092 of composite objects and the Max/Min attributes.
10094 @node No_Implicit_Dynamic_Code
10095 @unnumberedsubsec No_Implicit_Dynamic_Code
10096 @findex No_Implicit_Dynamic_Code
10098 [GNAT] This restriction prevents the compiler from building ``trampolines''.
10099 This is a structure that is built on the stack and contains dynamic
10100 code to be executed at run time. On some targets, a trampoline is
10101 built for the following features: @code{Access},
10102 @code{Unrestricted_Access}, or @code{Address} of a nested subprogram;
10103 nested task bodies; primitive operations of nested tagged types.
10104 Trampolines do not work on machines that prevent execution of stack
10105 data. For example, on windows systems, enabling DEP (data execution
10106 protection) will cause trampolines to raise an exception.
10107 Trampolines are also quite slow at run time.
10109 On many targets, trampolines have been largely eliminated. Look at the
10110 version of system.ads for your target --- if it has
10111 Always_Compatible_Rep equal to False, then trampolines are largely
10112 eliminated. In particular, a trampoline is built for the following
10113 features: @code{Address} of a nested subprogram;
10114 @code{Access} or @code{Unrestricted_Access} of a nested subprogram,
10115 but only if pragma Favor_Top_Level applies, or the access type has a
10116 foreign-language convention; primitive operations of nested tagged
10119 @node No_Implicit_Heap_Allocations
10120 @unnumberedsubsec No_Implicit_Heap_Allocations
10121 @findex No_Implicit_Heap_Allocations
10122 [RM D.7] No constructs are allowed to cause implicit heap allocation.
10124 @node No_Implicit_Loops
10125 @unnumberedsubsec No_Implicit_Loops
10126 @findex No_Implicit_Loops
10127 [GNAT] This restriction ensures that the generated code does not contain any
10128 implicit @code{for} loops, either by modifying
10129 the generated code where possible,
10130 or by rejecting any construct that would otherwise generate an implicit
10131 @code{for} loop. If this restriction is active, it is possible to build
10132 large array aggregates with all static components without generating an
10133 intermediate temporary, and without generating a loop to initialize individual
10134 components. Otherwise, a loop is created for arrays larger than about 5000
10137 @node No_Initialize_Scalars
10138 @unnumberedsubsec No_Initialize_Scalars
10139 @findex No_Initialize_Scalars
10140 [GNAT] This restriction ensures that no unit in the partition is compiled with
10141 pragma Initialize_Scalars. This allows the generation of more efficient
10142 code, and in particular eliminates dummy null initialization routines that
10143 are otherwise generated for some record and array types.
10146 @unnumberedsubsec No_IO
10148 [RM H.4] This restriction ensures at compile time that there are no
10149 dependences on any of the library units Sequential_IO, Direct_IO,
10150 Text_IO, Wide_Text_IO, Wide_Wide_Text_IO, or Stream_IO.
10152 @node No_Local_Allocators
10153 @unnumberedsubsec No_Local_Allocators
10154 @findex No_Local_Allocators
10155 [RM H.4] This restriction ensures at compile time that there are no
10156 occurrences of an allocator in subprograms, generic subprograms, tasks,
10159 @node No_Local_Protected_Objects
10160 @unnumberedsubsec No_Local_Protected_Objects
10161 @findex No_Local_Protected_Objects
10162 [RM D.7] This restriction ensures at compile time that protected objects are
10163 only declared at the library level.
10165 @node No_Local_Timing_Events
10166 @unnumberedsubsec No_Local_Timing_Events
10167 @findex No_Local_Timing_Events
10168 [RM D.7] All objects of type Ada.Timing_Events.Timing_Event are
10169 declared at the library level.
10171 @node No_Nested_Finalization
10172 @unnumberedsubsec No_Nested_Finalization
10173 @findex No_Nested_Finalization
10174 [RM D.7] All objects requiring finalization are declared at the library level.
10176 @node No_Protected_Type_Allocators
10177 @unnumberedsubsec No_Protected_Type_Allocators
10178 @findex No_Protected_Type_Allocators
10179 [RM D.7] This restriction ensures at compile time that there are no allocator
10180 expressions that attempt to allocate protected objects.
10182 @node No_Protected_Types
10183 @unnumberedsubsec No_Protected_Types
10184 @findex No_Protected_Types
10185 [RM H.4] This restriction ensures at compile time that there are no
10186 declarations of protected types or protected objects.
10189 @unnumberedsubsec No_Recursion
10190 @findex No_Recursion
10191 [RM H.4] A program execution is erroneous if a subprogram is invoked as
10192 part of its execution.
10194 @node No_Reentrancy
10195 @unnumberedsubsec No_Reentrancy
10196 @findex No_Reentrancy
10197 [RM H.4] A program execution is erroneous if a subprogram is executed by
10198 two tasks at the same time.
10200 @node No_Relative_Delay
10201 @unnumberedsubsec No_Relative_Delay
10202 @findex No_Relative_Delay
10203 [RM D.7] This restriction ensures at compile time that there are no delay
10204 relative statements and prevents expressions such as @code{delay 1.23;} from
10205 appearing in source code.
10207 @node No_Requeue_Statements
10208 @unnumberedsubsec No_Requeue_Statements
10209 @findex No_Requeue_Statements
10210 [RM D.7] This restriction ensures at compile time that no requeue statements
10211 are permitted and prevents keyword @code{requeue} from being used in source
10215 The restriction @code{No_Requeue} is recognized as a
10216 synonym for @code{No_Requeue_Statements}. This is retained for historical
10217 compatibility purposes (and a warning will be generated for its use if
10218 warnings on oNobsolescent features are activated).
10220 @node No_Secondary_Stack
10221 @unnumberedsubsec No_Secondary_Stack
10222 @findex No_Secondary_Stack
10223 [GNAT] This restriction ensures at compile time that the generated code
10224 does not contain any reference to the secondary stack. The secondary
10225 stack is used to implement functions returning unconstrained objects
10226 (arrays or records) on some targets.
10228 @node No_Select_Statements
10229 @unnumberedsubsec No_Select_Statements
10230 @findex No_Select_Statements
10231 [RM D.7] This restriction ensures at compile time no select statements of any
10232 kind are permitted, that is the keyword @code{select} may not appear.
10234 @node No_Specific_Termination_Handlers
10235 @unnumberedsubsec No_Specific_Termination_Handlers
10236 @findex No_Specific_Termination_Handlers
10237 [RM D.7] There are no calls to Ada.Task_Termination.Set_Specific_Handler
10238 or to Ada.Task_Termination.Specific_Handler.
10240 @node No_Specification_of_Aspect
10241 @unnumberedsubsec No_Specification_of_Aspect
10242 @findex No_Specification_of_Aspect
10243 [RM 13.12.1] This restriction checks at compile time that no aspect
10244 specification, attribute definition clause, or pragma is given for a
10247 @node No_Standard_Allocators_After_Elaboration
10248 @unnumberedsubsec No_Standard_Allocators_After_Elaboration
10249 @findex No_Standard_Allocators_After_Elaboration
10250 [RM D.7] Specifies that an allocator using a standard storage pool
10251 should never be evaluated at run time after the elaboration of the
10252 library items of the partition has completed. Otherwise, Storage_Error
10255 @node No_Standard_Storage_Pools
10256 @unnumberedsubsec No_Standard_Storage_Pools
10257 @findex No_Standard_Storage_Pools
10258 [GNAT] This restriction ensures at compile time that no access types
10259 use the standard default storage pool. Any access type declared must
10260 have an explicit Storage_Pool attribute defined specifying a
10261 user-defined storage pool.
10263 @node No_Stream_Optimizations
10264 @unnumberedsubsec No_Stream_Optimizations
10265 @findex No_Stream_Optimizations
10266 [GNAT] This restriction affects the performance of stream operations on types
10267 @code{String}, @code{Wide_String} and @code{Wide_Wide_String}. By default, the
10268 compiler uses block reads and writes when manipulating @code{String} objects
10269 due to their supperior performance. When this restriction is in effect, the
10270 compiler performs all IO operations on a per-character basis.
10273 @unnumberedsubsec No_Streams
10275 [GNAT] This restriction ensures at compile/bind time that there are no
10276 stream objects created and no use of stream attributes.
10277 This restriction does not forbid dependences on the package
10278 @code{Ada.Streams}. So it is permissible to with
10279 @code{Ada.Streams} (or another package that does so itself)
10280 as long as no actual stream objects are created and no
10281 stream attributes are used.
10283 Note that the use of restriction allows optimization of tagged types,
10284 since they do not need to worry about dispatching stream operations.
10285 To take maximum advantage of this space-saving optimization, any
10286 unit declaring a tagged type should be compiled with the restriction,
10287 though this is not required.
10289 @node No_Task_Allocators
10290 @unnumberedsubsec No_Task_Allocators
10291 @findex No_Task_Allocators
10292 [RM D.7] There are no allocators for task types
10293 or types containing task subcomponents.
10295 @node No_Task_Attributes_Package
10296 @unnumberedsubsec No_Task_Attributes_Package
10297 @findex No_Task_Attributes_Package
10298 [GNAT] This restriction ensures at compile time that there are no implicit or
10299 explicit dependencies on the package @code{Ada.Task_Attributes}.
10301 @findex No_Task_Attributes
10302 The restriction @code{No_Task_Attributes} is recognized as a synonym
10303 for @code{No_Task_Attributes_Package}. This is retained for historical
10304 compatibility purposes (and a warning will be generated for its use if
10305 warnings on obsolescent features are activated).
10307 @node No_Task_Hierarchy
10308 @unnumberedsubsec No_Task_Hierarchy
10309 @findex No_Task_Hierarchy
10310 [RM D.7] All (non-environment) tasks depend
10311 directly on the environment task of the partition.
10313 @node No_Task_Termination
10314 @unnumberedsubsec No_Task_Termination
10315 @findex No_Task_Termination
10316 [RM D.7] Tasks which terminate are erroneous.
10319 @unnumberedsubsec No_Tasking
10321 [GNAT] This restriction prevents the declaration of tasks or task types
10322 throughout the partition. It is similar in effect to the use of
10323 @code{Max_Tasks => 0} except that violations are caught at compile time
10324 and cause an error message to be output either by the compiler or
10327 @node No_Terminate_Alternatives
10328 @unnumberedsubsec No_Terminate_Alternatives
10329 @findex No_Terminate_Alternatives
10330 [RM D.7] There are no selective accepts with terminate alternatives.
10332 @node No_Unchecked_Access
10333 @unnumberedsubsec No_Unchecked_Access
10334 @findex No_Unchecked_Access
10335 [RM H.4] This restriction ensures at compile time that there are no
10336 occurrences of the Unchecked_Access attribute.
10338 @node Simple_Barriers
10339 @unnumberedsubsec Simple_Barriers
10340 @findex Simple_Barriers
10341 [RM D.7] This restriction ensures at compile time that barriers in entry
10342 declarations for protected types are restricted to either static boolean
10343 expressions or references to simple boolean variables defined in the private
10344 part of the protected type. No other form of entry barriers is permitted.
10346 @findex Boolean_Entry_Barriers
10347 The restriction @code{Boolean_Entry_Barriers} is recognized as a
10348 synonym for @code{Simple_Barriers}. This is retained for historical
10349 compatibility purposes (and a warning will be generated for its use if
10350 warnings on obsolescent features are activated).
10352 @node Static_Priorities
10353 @unnumberedsubsec Static_Priorities
10354 @findex Static_Priorities
10355 [GNAT] This restriction ensures at compile time that all priority expressions
10356 are static, and that there are no dependences on the package
10357 @code{Ada.Dynamic_Priorities}.
10359 @node Static_Storage_Size
10360 @unnumberedsubsec Static_Storage_Size
10361 @findex Static_Storage_Size
10362 [GNAT] This restriction ensures at compile time that any expression appearing
10363 in a Storage_Size pragma or attribute definition clause is static.
10365 @node Program Unit Level Restrictions
10366 @section Program Unit Level Restrictions
10369 The second set of restriction identifiers
10370 does not require partition-wide consistency.
10371 The restriction may be enforced for a single
10372 compilation unit without any effect on any of the
10373 other compilation units in the partition.
10376 * No_Elaboration_Code::
10378 * No_Implementation_Aspect_Specifications::
10379 * No_Implementation_Attributes::
10380 * No_Implementation_Identifiers::
10381 * No_Implementation_Pragmas::
10382 * No_Implementation_Restrictions::
10383 * No_Implementation_Units::
10384 * No_Implicit_Aliasing::
10385 * No_Obsolescent_Features::
10386 * No_Wide_Characters::
10390 @node No_Elaboration_Code
10391 @unnumberedsubsec No_Elaboration_Code
10392 @findex No_Elaboration_Code
10393 [GNAT] This restriction ensures at compile time that no elaboration code is
10394 generated. Note that this is not the same condition as is enforced
10395 by pragma @code{Preelaborate}. There are cases in which pragma
10396 @code{Preelaborate} still permits code to be generated (e.g.@: code
10397 to initialize a large array to all zeroes), and there are cases of units
10398 which do not meet the requirements for pragma @code{Preelaborate},
10399 but for which no elaboration code is generated. Generally, it is
10400 the case that preelaborable units will meet the restrictions, with
10401 the exception of large aggregates initialized with an others_clause,
10402 and exception declarations (which generate calls to a run-time
10403 registry procedure). This restriction is enforced on
10404 a unit by unit basis, it need not be obeyed consistently
10405 throughout a partition.
10407 In the case of aggregates with others, if the aggregate has a dynamic
10408 size, there is no way to eliminate the elaboration code (such dynamic
10409 bounds would be incompatible with @code{Preelaborate} in any case). If
10410 the bounds are static, then use of this restriction actually modifies
10411 the code choice of the compiler to avoid generating a loop, and instead
10412 generate the aggregate statically if possible, no matter how many times
10413 the data for the others clause must be repeatedly generated.
10415 It is not possible to precisely document
10416 the constructs which are compatible with this restriction, since,
10417 unlike most other restrictions, this is not a restriction on the
10418 source code, but a restriction on the generated object code. For
10419 example, if the source contains a declaration:
10422 Val : constant Integer := X;
10426 where X is not a static constant, it may be possible, depending
10427 on complex optimization circuitry, for the compiler to figure
10428 out the value of X at compile time, in which case this initialization
10429 can be done by the loader, and requires no initialization code. It
10430 is not possible to document the precise conditions under which the
10431 optimizer can figure this out.
10433 Note that this the implementation of this restriction requires full
10434 code generation. If it is used in conjunction with "semantics only"
10435 checking, then some cases of violations may be missed.
10437 @node No_Entry_Queue
10438 @unnumberedsubsec No_Entry_Queue
10439 @findex No_Entry_Queue
10440 [GNAT] This restriction is a declaration that any protected entry compiled in
10441 the scope of the restriction has at most one task waiting on the entry
10442 at any one time, and so no queue is required. This restriction is not
10443 checked at compile time. A program execution is erroneous if an attempt
10444 is made to queue a second task on such an entry.
10446 @node No_Implementation_Aspect_Specifications
10447 @unnumberedsubsec No_Implementation_Aspect_Specifications
10448 @findex No_Implementation_Aspect_Specifications
10449 [RM 13.12.1] This restriction checks at compile time that no
10450 GNAT-defined aspects are present. With this restriction, the only
10451 aspects that can be used are those defined in the Ada Reference Manual.
10453 @node No_Implementation_Attributes
10454 @unnumberedsubsec No_Implementation_Attributes
10455 @findex No_Implementation_Attributes
10456 [RM 13.12.1] This restriction checks at compile time that no
10457 GNAT-defined attributes are present. With this restriction, the only
10458 attributes that can be used are those defined in the Ada Reference
10461 @node No_Implementation_Identifiers
10462 @unnumberedsubsec No_Implementation_Identifiers
10463 @findex No_Implementation_Identifiers
10464 [RM 13.12.1] This restriction checks at compile time that no
10465 implementation-defined identifiers (marked with pragma Implementation_Defined)
10466 occur within language-defined packages.
10468 @node No_Implementation_Pragmas
10469 @unnumberedsubsec No_Implementation_Pragmas
10470 @findex No_Implementation_Pragmas
10471 [RM 13.12.1] This restriction checks at compile time that no
10472 GNAT-defined pragmas are present. With this restriction, the only
10473 pragmas that can be used are those defined in the Ada Reference Manual.
10475 @node No_Implementation_Restrictions
10476 @unnumberedsubsec No_Implementation_Restrictions
10477 @findex No_Implementation_Restrictions
10478 [GNAT] This restriction checks at compile time that no GNAT-defined restriction
10479 identifiers (other than @code{No_Implementation_Restrictions} itself)
10480 are present. With this restriction, the only other restriction identifiers
10481 that can be used are those defined in the Ada Reference Manual.
10483 @node No_Implementation_Units
10484 @unnumberedsubsec No_Implementation_Units
10485 @findex No_Implementation_Units
10486 [RM 13.12.1] This restriction checks at compile time that there is no
10487 mention in the context clause of any implementation-defined descendants
10488 of packages Ada, Interfaces, or System.
10490 @node No_Implicit_Aliasing
10491 @unnumberedsubsec No_Implicit_Aliasing
10492 @findex No_Implicit_Aliasing
10493 [GNAT] This restriction, which is not required to be partition-wide consistent,
10494 requires an explicit aliased keyword for an object to which 'Access,
10495 'Unchecked_Access, or 'Address is applied, and forbids entirely the use of
10496 the 'Unrestricted_Access attribute for objects. Note: the reason that
10497 Unrestricted_Access is forbidden is that it would require the prefix
10498 to be aliased, and in such cases, it can always be replaced by
10499 the standard attribute Unchecked_Access which is preferable.
10501 @node No_Obsolescent_Features
10502 @unnumberedsubsec No_Obsolescent_Features
10503 @findex No_Obsolescent_Features
10504 [RM 13.12.1] This restriction checks at compile time that no obsolescent
10505 features are used, as defined in Annex J of the Ada Reference Manual.
10507 @node No_Wide_Characters
10508 @unnumberedsubsec No_Wide_Characters
10509 @findex No_Wide_Characters
10510 [GNAT] This restriction ensures at compile time that no uses of the types
10511 @code{Wide_Character} or @code{Wide_String} or corresponding wide
10513 appear, and that no wide or wide wide string or character literals
10514 appear in the program (that is literals representing characters not in
10515 type @code{Character}).
10518 @unnumberedsubsec SPARK_05
10520 [GNAT] This restriction checks at compile time that some constructs
10521 forbidden in SPARK 2005 are not present. Error messages related to
10522 SPARK restriction have the form:
10525 The restriction @code{SPARK} is recognized as a
10526 synonym for @code{SPARK_05}. This is retained for historical
10527 compatibility purposes (and an unconditional warning will be generated
10528 for its use, advising replacement by @code{SPARK}.
10531 violation of restriction "SPARK" at <file>
10535 This is not a replacement for the semantic checks performed by the
10536 SPARK Examiner tool, as the compiler only deals currently with code,
10537 not at all with SPARK 2005 annotations and does not guarantee catching all
10538 cases of constructs forbidden by SPARK 2005.
10540 Thus it may well be the case that code which passes the compiler with
10541 the SPARK restriction is rejected by the SPARK Examiner, e.g. due to
10542 the different visibility rules of the Examiner based on SPARK 2005
10543 @code{inherit} annotations.
10545 This restriction can be useful in providing an initial filter for code
10546 developed using SPARK 2005, or in examining legacy code to see how far
10547 it is from meeting SPARK restrictions.
10549 Note that if a unit is compiled in Ada 95 mode with SPARK restriction,
10550 violations will be reported for constructs forbidden in SPARK 95,
10551 instead of SPARK 2005.
10553 @c ------------------------
10554 @node Implementation Advice
10555 @chapter Implementation Advice
10557 The main text of the Ada Reference Manual describes the required
10558 behavior of all Ada compilers, and the GNAT compiler conforms to
10559 these requirements.
10561 In addition, there are sections throughout the Ada Reference Manual headed
10562 by the phrase ``Implementation advice''. These sections are not normative,
10563 i.e., they do not specify requirements that all compilers must
10564 follow. Rather they provide advice on generally desirable behavior. You
10565 may wonder why they are not requirements. The most typical answer is
10566 that they describe behavior that seems generally desirable, but cannot
10567 be provided on all systems, or which may be undesirable on some systems.
10569 As far as practical, GNAT follows the implementation advice sections in
10570 the Ada Reference Manual. This chapter contains a table giving the
10571 reference manual section number, paragraph number and several keywords
10572 for each advice. Each entry consists of the text of the advice followed
10573 by the GNAT interpretation of this advice. Most often, this simply says
10574 ``followed'', which means that GNAT follows the advice. However, in a
10575 number of cases, GNAT deliberately deviates from this advice, in which
10576 case the text describes what GNAT does and why.
10578 @cindex Error detection
10579 @unnumberedsec 1.1.3(20): Error Detection
10582 If an implementation detects the use of an unsupported Specialized Needs
10583 Annex feature at run time, it should raise @code{Program_Error} if
10586 Not relevant. All specialized needs annex features are either supported,
10587 or diagnosed at compile time.
10589 @cindex Child Units
10590 @unnumberedsec 1.1.3(31): Child Units
10593 If an implementation wishes to provide implementation-defined
10594 extensions to the functionality of a language-defined library unit, it
10595 should normally do so by adding children to the library unit.
10599 @cindex Bounded errors
10600 @unnumberedsec 1.1.5(12): Bounded Errors
10603 If an implementation detects a bounded error or erroneous
10604 execution, it should raise @code{Program_Error}.
10606 Followed in all cases in which the implementation detects a bounded
10607 error or erroneous execution. Not all such situations are detected at
10611 @unnumberedsec 2.8(16): Pragmas
10614 Normally, implementation-defined pragmas should have no semantic effect
10615 for error-free programs; that is, if the implementation-defined pragmas
10616 are removed from a working program, the program should still be legal,
10617 and should still have the same semantics.
10619 The following implementation defined pragmas are exceptions to this
10631 @item CPP_Constructor
10635 @item Interface_Name
10637 @item Machine_Attribute
10639 @item Unimplemented_Unit
10641 @item Unchecked_Union
10646 In each of the above cases, it is essential to the purpose of the pragma
10647 that this advice not be followed. For details see the separate section
10648 on implementation defined pragmas.
10650 @unnumberedsec 2.8(17-19): Pragmas
10653 Normally, an implementation should not define pragmas that can
10654 make an illegal program legal, except as follows:
10658 A pragma used to complete a declaration, such as a pragma @code{Import};
10662 A pragma used to configure the environment by adding, removing, or
10663 replacing @code{library_items}.
10665 See response to paragraph 16 of this same section.
10667 @cindex Character Sets
10668 @cindex Alternative Character Sets
10669 @unnumberedsec 3.5.2(5): Alternative Character Sets
10672 If an implementation supports a mode with alternative interpretations
10673 for @code{Character} and @code{Wide_Character}, the set of graphic
10674 characters of @code{Character} should nevertheless remain a proper
10675 subset of the set of graphic characters of @code{Wide_Character}. Any
10676 character set ``localizations'' should be reflected in the results of
10677 the subprograms defined in the language-defined package
10678 @code{Characters.Handling} (see A.3) available in such a mode. In a mode with
10679 an alternative interpretation of @code{Character}, the implementation should
10680 also support a corresponding change in what is a legal
10681 @code{identifier_letter}.
10683 Not all wide character modes follow this advice, in particular the JIS
10684 and IEC modes reflect standard usage in Japan, and in these encoding,
10685 the upper half of the Latin-1 set is not part of the wide-character
10686 subset, since the most significant bit is used for wide character
10687 encoding. However, this only applies to the external forms. Internally
10688 there is no such restriction.
10690 @cindex Integer types
10691 @unnumberedsec 3.5.4(28): Integer Types
10695 An implementation should support @code{Long_Integer} in addition to
10696 @code{Integer} if the target machine supports 32-bit (or longer)
10697 arithmetic. No other named integer subtypes are recommended for package
10698 @code{Standard}. Instead, appropriate named integer subtypes should be
10699 provided in the library package @code{Interfaces} (see B.2).
10701 @code{Long_Integer} is supported. Other standard integer types are supported
10702 so this advice is not fully followed. These types
10703 are supported for convenient interface to C, and so that all hardware
10704 types of the machine are easily available.
10705 @unnumberedsec 3.5.4(29): Integer Types
10709 An implementation for a two's complement machine should support
10710 modular types with a binary modulus up to @code{System.Max_Int*2+2}. An
10711 implementation should support a non-binary modules up to @code{Integer'Last}.
10715 @cindex Enumeration values
10716 @unnumberedsec 3.5.5(8): Enumeration Values
10719 For the evaluation of a call on @code{@var{S}'Pos} for an enumeration
10720 subtype, if the value of the operand does not correspond to the internal
10721 code for any enumeration literal of its type (perhaps due to an
10722 un-initialized variable), then the implementation should raise
10723 @code{Program_Error}. This is particularly important for enumeration
10724 types with noncontiguous internal codes specified by an
10725 enumeration_representation_clause.
10729 @cindex Float types
10730 @unnumberedsec 3.5.7(17): Float Types
10733 An implementation should support @code{Long_Float} in addition to
10734 @code{Float} if the target machine supports 11 or more digits of
10735 precision. No other named floating point subtypes are recommended for
10736 package @code{Standard}. Instead, appropriate named floating point subtypes
10737 should be provided in the library package @code{Interfaces} (see B.2).
10739 @code{Short_Float} and @code{Long_Long_Float} are also provided. The
10740 former provides improved compatibility with other implementations
10741 supporting this type. The latter corresponds to the highest precision
10742 floating-point type supported by the hardware. On most machines, this
10743 will be the same as @code{Long_Float}, but on some machines, it will
10744 correspond to the IEEE extended form. The notable case is all ia32
10745 (x86) implementations, where @code{Long_Long_Float} corresponds to
10746 the 80-bit extended precision format supported in hardware on this
10747 processor. Note that the 128-bit format on SPARC is not supported,
10748 since this is a software rather than a hardware format.
10750 @cindex Multidimensional arrays
10751 @cindex Arrays, multidimensional
10752 @unnumberedsec 3.6.2(11): Multidimensional Arrays
10755 An implementation should normally represent multidimensional arrays in
10756 row-major order, consistent with the notation used for multidimensional
10757 array aggregates (see 4.3.3). However, if a pragma @code{Convention}
10758 (@code{Fortran}, @dots{}) applies to a multidimensional array type, then
10759 column-major order should be used instead (see B.5, ``Interfacing with
10764 @findex Duration'Small
10765 @unnumberedsec 9.6(30-31): Duration'Small
10768 Whenever possible in an implementation, the value of @code{Duration'Small}
10769 should be no greater than 100 microseconds.
10771 Followed. (@code{Duration'Small} = 10**(@minus{}9)).
10775 The time base for @code{delay_relative_statements} should be monotonic;
10776 it need not be the same time base as used for @code{Calendar.Clock}.
10780 @unnumberedsec 10.2.1(12): Consistent Representation
10783 In an implementation, a type declared in a pre-elaborated package should
10784 have the same representation in every elaboration of a given version of
10785 the package, whether the elaborations occur in distinct executions of
10786 the same program, or in executions of distinct programs or partitions
10787 that include the given version.
10789 Followed, except in the case of tagged types. Tagged types involve
10790 implicit pointers to a local copy of a dispatch table, and these pointers
10791 have representations which thus depend on a particular elaboration of the
10792 package. It is not easy to see how it would be possible to follow this
10793 advice without severely impacting efficiency of execution.
10795 @cindex Exception information
10796 @unnumberedsec 11.4.1(19): Exception Information
10799 @code{Exception_Message} by default and @code{Exception_Information}
10800 should produce information useful for
10801 debugging. @code{Exception_Message} should be short, about one
10802 line. @code{Exception_Information} can be long. @code{Exception_Message}
10803 should not include the
10804 @code{Exception_Name}. @code{Exception_Information} should include both
10805 the @code{Exception_Name} and the @code{Exception_Message}.
10807 Followed. For each exception that doesn't have a specified
10808 @code{Exception_Message}, the compiler generates one containing the location
10809 of the raise statement. This location has the form ``file:line'', where
10810 file is the short file name (without path information) and line is the line
10811 number in the file. Note that in the case of the Zero Cost Exception
10812 mechanism, these messages become redundant with the Exception_Information that
10813 contains a full backtrace of the calling sequence, so they are disabled.
10814 To disable explicitly the generation of the source location message, use the
10815 Pragma @code{Discard_Names}.
10817 @cindex Suppression of checks
10818 @cindex Checks, suppression of
10819 @unnumberedsec 11.5(28): Suppression of Checks
10822 The implementation should minimize the code executed for checks that
10823 have been suppressed.
10827 @cindex Representation clauses
10828 @unnumberedsec 13.1 (21-24): Representation Clauses
10831 The recommended level of support for all representation items is
10832 qualified as follows:
10836 An implementation need not support representation items containing
10837 non-static expressions, except that an implementation should support a
10838 representation item for a given entity if each non-static expression in
10839 the representation item is a name that statically denotes a constant
10840 declared before the entity.
10842 Followed. In fact, GNAT goes beyond the recommended level of support
10843 by allowing nonstatic expressions in some representation clauses even
10844 without the need to declare constants initialized with the values of
10848 @smallexample @c ada
10851 for Y'Address use X'Address;>>
10856 An implementation need not support a specification for the @code{Size}
10857 for a given composite subtype, nor the size or storage place for an
10858 object (including a component) of a given composite subtype, unless the
10859 constraints on the subtype and its composite subcomponents (if any) are
10860 all static constraints.
10862 Followed. Size Clauses are not permitted on non-static components, as
10867 An aliased component, or a component whose type is by-reference, should
10868 always be allocated at an addressable location.
10872 @cindex Packed types
10873 @unnumberedsec 13.2(6-8): Packed Types
10876 If a type is packed, then the implementation should try to minimize
10877 storage allocated to objects of the type, possibly at the expense of
10878 speed of accessing components, subject to reasonable complexity in
10879 addressing calculations.
10883 The recommended level of support pragma @code{Pack} is:
10885 For a packed record type, the components should be packed as tightly as
10886 possible subject to the Sizes of the component subtypes, and subject to
10887 any @code{record_representation_clause} that applies to the type; the
10888 implementation may, but need not, reorder components or cross aligned
10889 word boundaries to improve the packing. A component whose @code{Size} is
10890 greater than the word size may be allocated an integral number of words.
10892 Followed. Tight packing of arrays is supported for all component sizes
10893 up to 64-bits. If the array component size is 1 (that is to say, if
10894 the component is a boolean type or an enumeration type with two values)
10895 then values of the type are implicitly initialized to zero. This
10896 happens both for objects of the packed type, and for objects that have a
10897 subcomponent of the packed type.
10901 An implementation should support Address clauses for imported
10905 @cindex @code{Address} clauses
10906 @unnumberedsec 13.3(14-19): Address Clauses
10910 For an array @var{X}, @code{@var{X}'Address} should point at the first
10911 component of the array, and not at the array bounds.
10917 The recommended level of support for the @code{Address} attribute is:
10919 @code{@var{X}'Address} should produce a useful result if @var{X} is an
10920 object that is aliased or of a by-reference type, or is an entity whose
10921 @code{Address} has been specified.
10923 Followed. A valid address will be produced even if none of those
10924 conditions have been met. If necessary, the object is forced into
10925 memory to ensure the address is valid.
10929 An implementation should support @code{Address} clauses for imported
10936 Objects (including subcomponents) that are aliased or of a by-reference
10937 type should be allocated on storage element boundaries.
10943 If the @code{Address} of an object is specified, or it is imported or exported,
10944 then the implementation should not perform optimizations based on
10945 assumptions of no aliases.
10949 @cindex @code{Alignment} clauses
10950 @unnumberedsec 13.3(29-35): Alignment Clauses
10953 The recommended level of support for the @code{Alignment} attribute for
10956 An implementation should support specified Alignments that are factors
10957 and multiples of the number of storage elements per word, subject to the
10964 An implementation need not support specified @code{Alignment}s for
10965 combinations of @code{Size}s and @code{Alignment}s that cannot be easily
10966 loaded and stored by available machine instructions.
10972 An implementation need not support specified @code{Alignment}s that are
10973 greater than the maximum @code{Alignment} the implementation ever returns by
10980 The recommended level of support for the @code{Alignment} attribute for
10983 Same as above, for subtypes, but in addition:
10989 For stand-alone library-level objects of statically constrained
10990 subtypes, the implementation should support all @code{Alignment}s
10991 supported by the target linker. For example, page alignment is likely to
10992 be supported for such objects, but not for subtypes.
10996 @cindex @code{Size} clauses
10997 @unnumberedsec 13.3(42-43): Size Clauses
11000 The recommended level of support for the @code{Size} attribute of
11003 A @code{Size} clause should be supported for an object if the specified
11004 @code{Size} is at least as large as its subtype's @code{Size}, and
11005 corresponds to a size in storage elements that is a multiple of the
11006 object's @code{Alignment} (if the @code{Alignment} is nonzero).
11010 @unnumberedsec 13.3(50-56): Size Clauses
11013 If the @code{Size} of a subtype is specified, and allows for efficient
11014 independent addressability (see 9.10) on the target architecture, then
11015 the @code{Size} of the following objects of the subtype should equal the
11016 @code{Size} of the subtype:
11018 Aliased objects (including components).
11024 @code{Size} clause on a composite subtype should not affect the
11025 internal layout of components.
11027 Followed. But note that this can be overridden by use of the implementation
11028 pragma Implicit_Packing in the case of packed arrays.
11032 The recommended level of support for the @code{Size} attribute of subtypes is:
11036 The @code{Size} (if not specified) of a static discrete or fixed point
11037 subtype should be the number of bits needed to represent each value
11038 belonging to the subtype using an unbiased representation, leaving space
11039 for a sign bit only if the subtype contains negative values. If such a
11040 subtype is a first subtype, then an implementation should support a
11041 specified @code{Size} for it that reflects this representation.
11047 For a subtype implemented with levels of indirection, the @code{Size}
11048 should include the size of the pointers, but not the size of what they
11053 @cindex @code{Component_Size} clauses
11054 @unnumberedsec 13.3(71-73): Component Size Clauses
11057 The recommended level of support for the @code{Component_Size}
11062 An implementation need not support specified @code{Component_Sizes} that are
11063 less than the @code{Size} of the component subtype.
11069 An implementation should support specified @code{Component_Size}s that
11070 are factors and multiples of the word size. For such
11071 @code{Component_Size}s, the array should contain no gaps between
11072 components. For other @code{Component_Size}s (if supported), the array
11073 should contain no gaps between components when packing is also
11074 specified; the implementation should forbid this combination in cases
11075 where it cannot support a no-gaps representation.
11079 @cindex Enumeration representation clauses
11080 @cindex Representation clauses, enumeration
11081 @unnumberedsec 13.4(9-10): Enumeration Representation Clauses
11084 The recommended level of support for enumeration representation clauses
11087 An implementation need not support enumeration representation clauses
11088 for boolean types, but should at minimum support the internal codes in
11089 the range @code{System.Min_Int.System.Max_Int}.
11093 @cindex Record representation clauses
11094 @cindex Representation clauses, records
11095 @unnumberedsec 13.5.1(17-22): Record Representation Clauses
11098 The recommended level of support for
11099 @*@code{record_representation_clauses} is:
11101 An implementation should support storage places that can be extracted
11102 with a load, mask, shift sequence of machine code, and set with a load,
11103 shift, mask, store sequence, given the available machine instructions
11104 and run-time model.
11110 A storage place should be supported if its size is equal to the
11111 @code{Size} of the component subtype, and it starts and ends on a
11112 boundary that obeys the @code{Alignment} of the component subtype.
11118 If the default bit ordering applies to the declaration of a given type,
11119 then for a component whose subtype's @code{Size} is less than the word
11120 size, any storage place that does not cross an aligned word boundary
11121 should be supported.
11127 An implementation may reserve a storage place for the tag field of a
11128 tagged type, and disallow other components from overlapping that place.
11130 Followed. The storage place for the tag field is the beginning of the tagged
11131 record, and its size is Address'Size. GNAT will reject an explicit component
11132 clause for the tag field.
11136 An implementation need not support a @code{component_clause} for a
11137 component of an extension part if the storage place is not after the
11138 storage places of all components of the parent type, whether or not
11139 those storage places had been specified.
11141 Followed. The above advice on record representation clauses is followed,
11142 and all mentioned features are implemented.
11144 @cindex Storage place attributes
11145 @unnumberedsec 13.5.2(5): Storage Place Attributes
11148 If a component is represented using some form of pointer (such as an
11149 offset) to the actual data of the component, and this data is contiguous
11150 with the rest of the object, then the storage place attributes should
11151 reflect the place of the actual data, not the pointer. If a component is
11152 allocated discontinuously from the rest of the object, then a warning
11153 should be generated upon reference to one of its storage place
11156 Followed. There are no such components in GNAT@.
11158 @cindex Bit ordering
11159 @unnumberedsec 13.5.3(7-8): Bit Ordering
11162 The recommended level of support for the non-default bit ordering is:
11166 If @code{Word_Size} = @code{Storage_Unit}, then the implementation
11167 should support the non-default bit ordering in addition to the default
11170 Followed. Word size does not equal storage size in this implementation.
11171 Thus non-default bit ordering is not supported.
11173 @cindex @code{Address}, as private type
11174 @unnumberedsec 13.7(37): Address as Private
11177 @code{Address} should be of a private type.
11181 @cindex Operations, on @code{Address}
11182 @cindex @code{Address}, operations of
11183 @unnumberedsec 13.7.1(16): Address Operations
11186 Operations in @code{System} and its children should reflect the target
11187 environment semantics as closely as is reasonable. For example, on most
11188 machines, it makes sense for address arithmetic to ``wrap around''.
11189 Operations that do not make sense should raise @code{Program_Error}.
11191 Followed. Address arithmetic is modular arithmetic that wraps around. No
11192 operation raises @code{Program_Error}, since all operations make sense.
11194 @cindex Unchecked conversion
11195 @unnumberedsec 13.9(14-17): Unchecked Conversion
11198 The @code{Size} of an array object should not include its bounds; hence,
11199 the bounds should not be part of the converted data.
11205 The implementation should not generate unnecessary run-time checks to
11206 ensure that the representation of @var{S} is a representation of the
11207 target type. It should take advantage of the permission to return by
11208 reference when possible. Restrictions on unchecked conversions should be
11209 avoided unless required by the target environment.
11211 Followed. There are no restrictions on unchecked conversion. A warning is
11212 generated if the source and target types do not have the same size since
11213 the semantics in this case may be target dependent.
11217 The recommended level of support for unchecked conversions is:
11221 Unchecked conversions should be supported and should be reversible in
11222 the cases where this clause defines the result. To enable meaningful use
11223 of unchecked conversion, a contiguous representation should be used for
11224 elementary subtypes, for statically constrained array subtypes whose
11225 component subtype is one of the subtypes described in this paragraph,
11226 and for record subtypes without discriminants whose component subtypes
11227 are described in this paragraph.
11231 @cindex Heap usage, implicit
11232 @unnumberedsec 13.11(23-25): Implicit Heap Usage
11235 An implementation should document any cases in which it dynamically
11236 allocates heap storage for a purpose other than the evaluation of an
11239 Followed, the only other points at which heap storage is dynamically
11240 allocated are as follows:
11244 At initial elaboration time, to allocate dynamically sized global
11248 To allocate space for a task when a task is created.
11251 To extend the secondary stack dynamically when needed. The secondary
11252 stack is used for returning variable length results.
11257 A default (implementation-provided) storage pool for an
11258 access-to-constant type should not have overhead to support deallocation of
11259 individual objects.
11265 A storage pool for an anonymous access type should be created at the
11266 point of an allocator for the type, and be reclaimed when the designated
11267 object becomes inaccessible.
11271 @cindex Unchecked deallocation
11272 @unnumberedsec 13.11.2(17): Unchecked De-allocation
11275 For a standard storage pool, @code{Free} should actually reclaim the
11280 @cindex Stream oriented attributes
11281 @unnumberedsec 13.13.2(17): Stream Oriented Attributes
11284 If a stream element is the same size as a storage element, then the
11285 normal in-memory representation should be used by @code{Read} and
11286 @code{Write} for scalar objects. Otherwise, @code{Read} and @code{Write}
11287 should use the smallest number of stream elements needed to represent
11288 all values in the base range of the scalar type.
11291 Followed. By default, GNAT uses the interpretation suggested by AI-195,
11292 which specifies using the size of the first subtype.
11293 However, such an implementation is based on direct binary
11294 representations and is therefore target- and endianness-dependent.
11295 To address this issue, GNAT also supplies an alternate implementation
11296 of the stream attributes @code{Read} and @code{Write},
11297 which uses the target-independent XDR standard representation
11299 @cindex XDR representation
11300 @cindex @code{Read} attribute
11301 @cindex @code{Write} attribute
11302 @cindex Stream oriented attributes
11303 The XDR implementation is provided as an alternative body of the
11304 @code{System.Stream_Attributes} package, in the file
11305 @file{s-stratt-xdr.adb} in the GNAT library.
11306 There is no @file{s-stratt-xdr.ads} file.
11307 In order to install the XDR implementation, do the following:
11309 @item Replace the default implementation of the
11310 @code{System.Stream_Attributes} package with the XDR implementation.
11311 For example on a Unix platform issue the commands:
11313 $ mv s-stratt.adb s-stratt-default.adb
11314 $ mv s-stratt-xdr.adb s-stratt.adb
11318 Rebuild the GNAT run-time library as documented in
11319 @ref{GNAT and Libraries,,, gnat_ugn, @value{EDITION} User's Guide}.
11322 @unnumberedsec A.1(52): Names of Predefined Numeric Types
11325 If an implementation provides additional named predefined integer types,
11326 then the names should end with @samp{Integer} as in
11327 @samp{Long_Integer}. If an implementation provides additional named
11328 predefined floating point types, then the names should end with
11329 @samp{Float} as in @samp{Long_Float}.
11333 @findex Ada.Characters.Handling
11334 @unnumberedsec A.3.2(49): @code{Ada.Characters.Handling}
11337 If an implementation provides a localized definition of @code{Character}
11338 or @code{Wide_Character}, then the effects of the subprograms in
11339 @code{Characters.Handling} should reflect the localizations. See also
11342 Followed. GNAT provides no such localized definitions.
11344 @cindex Bounded-length strings
11345 @unnumberedsec A.4.4(106): Bounded-Length String Handling
11348 Bounded string objects should not be implemented by implicit pointers
11349 and dynamic allocation.
11351 Followed. No implicit pointers or dynamic allocation are used.
11353 @cindex Random number generation
11354 @unnumberedsec A.5.2(46-47): Random Number Generation
11357 Any storage associated with an object of type @code{Generator} should be
11358 reclaimed on exit from the scope of the object.
11364 If the generator period is sufficiently long in relation to the number
11365 of distinct initiator values, then each possible value of
11366 @code{Initiator} passed to @code{Reset} should initiate a sequence of
11367 random numbers that does not, in a practical sense, overlap the sequence
11368 initiated by any other value. If this is not possible, then the mapping
11369 between initiator values and generator states should be a rapidly
11370 varying function of the initiator value.
11372 Followed. The generator period is sufficiently long for the first
11373 condition here to hold true.
11375 @findex Get_Immediate
11376 @unnumberedsec A.10.7(23): @code{Get_Immediate}
11379 The @code{Get_Immediate} procedures should be implemented with
11380 unbuffered input. For a device such as a keyboard, input should be
11381 @dfn{available} if a key has already been typed, whereas for a disk
11382 file, input should always be available except at end of file. For a file
11383 associated with a keyboard-like device, any line-editing features of the
11384 underlying operating system should be disabled during the execution of
11385 @code{Get_Immediate}.
11387 Followed on all targets except VxWorks. For VxWorks, there is no way to
11388 provide this functionality that does not result in the input buffer being
11389 flushed before the @code{Get_Immediate} call. A special unit
11390 @code{Interfaces.Vxworks.IO} is provided that contains routines to enable
11391 this functionality.
11394 @unnumberedsec B.1(39-41): Pragma @code{Export}
11397 If an implementation supports pragma @code{Export} to a given language,
11398 then it should also allow the main subprogram to be written in that
11399 language. It should support some mechanism for invoking the elaboration
11400 of the Ada library units included in the system, and for invoking the
11401 finalization of the environment task. On typical systems, the
11402 recommended mechanism is to provide two subprograms whose link names are
11403 @code{adainit} and @code{adafinal}. @code{adainit} should contain the
11404 elaboration code for library units. @code{adafinal} should contain the
11405 finalization code. These subprograms should have no effect the second
11406 and subsequent time they are called.
11412 Automatic elaboration of pre-elaborated packages should be
11413 provided when pragma @code{Export} is supported.
11415 Followed when the main program is in Ada. If the main program is in a
11416 foreign language, then
11417 @code{adainit} must be called to elaborate pre-elaborated
11422 For each supported convention @var{L} other than @code{Intrinsic}, an
11423 implementation should support @code{Import} and @code{Export} pragmas
11424 for objects of @var{L}-compatible types and for subprograms, and pragma
11425 @code{Convention} for @var{L}-eligible types and for subprograms,
11426 presuming the other language has corresponding features. Pragma
11427 @code{Convention} need not be supported for scalar types.
11431 @cindex Package @code{Interfaces}
11433 @unnumberedsec B.2(12-13): Package @code{Interfaces}
11436 For each implementation-defined convention identifier, there should be a
11437 child package of package Interfaces with the corresponding name. This
11438 package should contain any declarations that would be useful for
11439 interfacing to the language (implementation) represented by the
11440 convention. Any declarations useful for interfacing to any language on
11441 the given hardware architecture should be provided directly in
11444 Followed. An additional package not defined
11445 in the Ada Reference Manual is @code{Interfaces.CPP}, used
11446 for interfacing to C++.
11450 An implementation supporting an interface to C, COBOL, or Fortran should
11451 provide the corresponding package or packages described in the following
11454 Followed. GNAT provides all the packages described in this section.
11456 @cindex C, interfacing with
11457 @unnumberedsec B.3(63-71): Interfacing with C
11460 An implementation should support the following interface correspondences
11461 between Ada and C@.
11467 An Ada procedure corresponds to a void-returning C function.
11473 An Ada function corresponds to a non-void C function.
11479 An Ada @code{in} scalar parameter is passed as a scalar argument to a C
11486 An Ada @code{in} parameter of an access-to-object type with designated
11487 type @var{T} is passed as a @code{@var{t}*} argument to a C function,
11488 where @var{t} is the C type corresponding to the Ada type @var{T}.
11494 An Ada access @var{T} parameter, or an Ada @code{out} or @code{in out}
11495 parameter of an elementary type @var{T}, is passed as a @code{@var{t}*}
11496 argument to a C function, where @var{t} is the C type corresponding to
11497 the Ada type @var{T}. In the case of an elementary @code{out} or
11498 @code{in out} parameter, a pointer to a temporary copy is used to
11499 preserve by-copy semantics.
11505 An Ada parameter of a record type @var{T}, of any mode, is passed as a
11506 @code{@var{t}*} argument to a C function, where @var{t} is the C
11507 structure corresponding to the Ada type @var{T}.
11509 Followed. This convention may be overridden by the use of the C_Pass_By_Copy
11510 pragma, or Convention, or by explicitly specifying the mechanism for a given
11511 call using an extended import or export pragma.
11515 An Ada parameter of an array type with component type @var{T}, of any
11516 mode, is passed as a @code{@var{t}*} argument to a C function, where
11517 @var{t} is the C type corresponding to the Ada type @var{T}.
11523 An Ada parameter of an access-to-subprogram type is passed as a pointer
11524 to a C function whose prototype corresponds to the designated
11525 subprogram's specification.
11529 @cindex COBOL, interfacing with
11530 @unnumberedsec B.4(95-98): Interfacing with COBOL
11533 An Ada implementation should support the following interface
11534 correspondences between Ada and COBOL@.
11540 An Ada access @var{T} parameter is passed as a @samp{BY REFERENCE} data item of
11541 the COBOL type corresponding to @var{T}.
11547 An Ada in scalar parameter is passed as a @samp{BY CONTENT} data item of
11548 the corresponding COBOL type.
11554 Any other Ada parameter is passed as a @samp{BY REFERENCE} data item of the
11555 COBOL type corresponding to the Ada parameter type; for scalars, a local
11556 copy is used if necessary to ensure by-copy semantics.
11560 @cindex Fortran, interfacing with
11561 @unnumberedsec B.5(22-26): Interfacing with Fortran
11564 An Ada implementation should support the following interface
11565 correspondences between Ada and Fortran:
11571 An Ada procedure corresponds to a Fortran subroutine.
11577 An Ada function corresponds to a Fortran function.
11583 An Ada parameter of an elementary, array, or record type @var{T} is
11584 passed as a @var{T} argument to a Fortran procedure, where @var{T} is
11585 the Fortran type corresponding to the Ada type @var{T}, and where the
11586 INTENT attribute of the corresponding dummy argument matches the Ada
11587 formal parameter mode; the Fortran implementation's parameter passing
11588 conventions are used. For elementary types, a local copy is used if
11589 necessary to ensure by-copy semantics.
11595 An Ada parameter of an access-to-subprogram type is passed as a
11596 reference to a Fortran procedure whose interface corresponds to the
11597 designated subprogram's specification.
11601 @cindex Machine operations
11602 @unnumberedsec C.1(3-5): Access to Machine Operations
11605 The machine code or intrinsic support should allow access to all
11606 operations normally available to assembly language programmers for the
11607 target environment, including privileged instructions, if any.
11613 The interfacing pragmas (see Annex B) should support interface to
11614 assembler; the default assembler should be associated with the
11615 convention identifier @code{Assembler}.
11621 If an entity is exported to assembly language, then the implementation
11622 should allocate it at an addressable location, and should ensure that it
11623 is retained by the linking process, even if not otherwise referenced
11624 from the Ada code. The implementation should assume that any call to a
11625 machine code or assembler subprogram is allowed to read or update every
11626 object that is specified as exported.
11630 @unnumberedsec C.1(10-16): Access to Machine Operations
11633 The implementation should ensure that little or no overhead is
11634 associated with calling intrinsic and machine-code subprograms.
11636 Followed for both intrinsics and machine-code subprograms.
11640 It is recommended that intrinsic subprograms be provided for convenient
11641 access to any machine operations that provide special capabilities or
11642 efficiency and that are not otherwise available through the language
11645 Followed. A full set of machine operation intrinsic subprograms is provided.
11649 Atomic read-modify-write operations---e.g.@:, test and set, compare and
11650 swap, decrement and test, enqueue/dequeue.
11652 Followed on any target supporting such operations.
11656 Standard numeric functions---e.g.@:, sin, log.
11658 Followed on any target supporting such operations.
11662 String manipulation operations---e.g.@:, translate and test.
11664 Followed on any target supporting such operations.
11668 Vector operations---e.g.@:, compare vector against thresholds.
11670 Followed on any target supporting such operations.
11674 Direct operations on I/O ports.
11676 Followed on any target supporting such operations.
11678 @cindex Interrupt support
11679 @unnumberedsec C.3(28): Interrupt Support
11682 If the @code{Ceiling_Locking} policy is not in effect, the
11683 implementation should provide means for the application to specify which
11684 interrupts are to be blocked during protected actions, if the underlying
11685 system allows for a finer-grain control of interrupt blocking.
11687 Followed. The underlying system does not allow for finer-grain control
11688 of interrupt blocking.
11690 @cindex Protected procedure handlers
11691 @unnumberedsec C.3.1(20-21): Protected Procedure Handlers
11694 Whenever possible, the implementation should allow interrupt handlers to
11695 be called directly by the hardware.
11697 Followed on any target where the underlying operating system permits
11702 Whenever practical, violations of any
11703 implementation-defined restrictions should be detected before run time.
11705 Followed. Compile time warnings are given when possible.
11707 @cindex Package @code{Interrupts}
11709 @unnumberedsec C.3.2(25): Package @code{Interrupts}
11713 If implementation-defined forms of interrupt handler procedures are
11714 supported, such as protected procedures with parameters, then for each
11715 such form of a handler, a type analogous to @code{Parameterless_Handler}
11716 should be specified in a child package of @code{Interrupts}, with the
11717 same operations as in the predefined package Interrupts.
11721 @cindex Pre-elaboration requirements
11722 @unnumberedsec C.4(14): Pre-elaboration Requirements
11725 It is recommended that pre-elaborated packages be implemented in such a
11726 way that there should be little or no code executed at run time for the
11727 elaboration of entities not already covered by the Implementation
11730 Followed. Executable code is generated in some cases, e.g.@: loops
11731 to initialize large arrays.
11733 @unnumberedsec C.5(8): Pragma @code{Discard_Names}
11736 If the pragma applies to an entity, then the implementation should
11737 reduce the amount of storage used for storing names associated with that
11742 @cindex Package @code{Task_Attributes}
11743 @findex Task_Attributes
11744 @unnumberedsec C.7.2(30): The Package Task_Attributes
11747 Some implementations are targeted to domains in which memory use at run
11748 time must be completely deterministic. For such implementations, it is
11749 recommended that the storage for task attributes will be pre-allocated
11750 statically and not from the heap. This can be accomplished by either
11751 placing restrictions on the number and the size of the task's
11752 attributes, or by using the pre-allocated storage for the first @var{N}
11753 attribute objects, and the heap for the others. In the latter case,
11754 @var{N} should be documented.
11756 Not followed. This implementation is not targeted to such a domain.
11758 @cindex Locking Policies
11759 @unnumberedsec D.3(17): Locking Policies
11763 The implementation should use names that end with @samp{_Locking} for
11764 locking policies defined by the implementation.
11766 Followed. Two implementation-defined locking policies are defined,
11767 whose names (@code{Inheritance_Locking} and
11768 @code{Concurrent_Readers_Locking}) follow this suggestion.
11770 @cindex Entry queuing policies
11771 @unnumberedsec D.4(16): Entry Queuing Policies
11774 Names that end with @samp{_Queuing} should be used
11775 for all implementation-defined queuing policies.
11777 Followed. No such implementation-defined queuing policies exist.
11779 @cindex Preemptive abort
11780 @unnumberedsec D.6(9-10): Preemptive Abort
11783 Even though the @code{abort_statement} is included in the list of
11784 potentially blocking operations (see 9.5.1), it is recommended that this
11785 statement be implemented in a way that never requires the task executing
11786 the @code{abort_statement} to block.
11792 On a multi-processor, the delay associated with aborting a task on
11793 another processor should be bounded; the implementation should use
11794 periodic polling, if necessary, to achieve this.
11798 @cindex Tasking restrictions
11799 @unnumberedsec D.7(21): Tasking Restrictions
11802 When feasible, the implementation should take advantage of the specified
11803 restrictions to produce a more efficient implementation.
11805 GNAT currently takes advantage of these restrictions by providing an optimized
11806 run time when the Ravenscar profile and the GNAT restricted run time set
11807 of restrictions are specified. See pragma @code{Profile (Ravenscar)} and
11808 pragma @code{Profile (Restricted)} for more details.
11810 @cindex Time, monotonic
11811 @unnumberedsec D.8(47-49): Monotonic Time
11814 When appropriate, implementations should provide configuration
11815 mechanisms to change the value of @code{Tick}.
11817 Such configuration mechanisms are not appropriate to this implementation
11818 and are thus not supported.
11822 It is recommended that @code{Calendar.Clock} and @code{Real_Time.Clock}
11823 be implemented as transformations of the same time base.
11829 It is recommended that the @dfn{best} time base which exists in
11830 the underlying system be available to the application through
11831 @code{Clock}. @dfn{Best} may mean highest accuracy or largest range.
11835 @cindex Partition communication subsystem
11837 @unnumberedsec E.5(28-29): Partition Communication Subsystem
11840 Whenever possible, the PCS on the called partition should allow for
11841 multiple tasks to call the RPC-receiver with different messages and
11842 should allow them to block until the corresponding subprogram body
11845 Followed by GLADE, a separately supplied PCS that can be used with
11850 The @code{Write} operation on a stream of type @code{Params_Stream_Type}
11851 should raise @code{Storage_Error} if it runs out of space trying to
11852 write the @code{Item} into the stream.
11854 Followed by GLADE, a separately supplied PCS that can be used with
11857 @cindex COBOL support
11858 @unnumberedsec F(7): COBOL Support
11861 If COBOL (respectively, C) is widely supported in the target
11862 environment, implementations supporting the Information Systems Annex
11863 should provide the child package @code{Interfaces.COBOL} (respectively,
11864 @code{Interfaces.C}) specified in Annex B and should support a
11865 @code{convention_identifier} of COBOL (respectively, C) in the interfacing
11866 pragmas (see Annex B), thus allowing Ada programs to interface with
11867 programs written in that language.
11871 @cindex Decimal radix support
11872 @unnumberedsec F.1(2): Decimal Radix Support
11875 Packed decimal should be used as the internal representation for objects
11876 of subtype @var{S} when @var{S}'Machine_Radix = 10.
11878 Not followed. GNAT ignores @var{S}'Machine_Radix and always uses binary
11882 @unnumberedsec G: Numerics
11885 If Fortran (respectively, C) is widely supported in the target
11886 environment, implementations supporting the Numerics Annex
11887 should provide the child package @code{Interfaces.Fortran} (respectively,
11888 @code{Interfaces.C}) specified in Annex B and should support a
11889 @code{convention_identifier} of Fortran (respectively, C) in the interfacing
11890 pragmas (see Annex B), thus allowing Ada programs to interface with
11891 programs written in that language.
11895 @cindex Complex types
11896 @unnumberedsec G.1.1(56-58): Complex Types
11899 Because the usual mathematical meaning of multiplication of a complex
11900 operand and a real operand is that of the scaling of both components of
11901 the former by the latter, an implementation should not perform this
11902 operation by first promoting the real operand to complex type and then
11903 performing a full complex multiplication. In systems that, in the
11904 future, support an Ada binding to IEC 559:1989, the latter technique
11905 will not generate the required result when one of the components of the
11906 complex operand is infinite. (Explicit multiplication of the infinite
11907 component by the zero component obtained during promotion yields a NaN
11908 that propagates into the final result.) Analogous advice applies in the
11909 case of multiplication of a complex operand and a pure-imaginary
11910 operand, and in the case of division of a complex operand by a real or
11911 pure-imaginary operand.
11917 Similarly, because the usual mathematical meaning of addition of a
11918 complex operand and a real operand is that the imaginary operand remains
11919 unchanged, an implementation should not perform this operation by first
11920 promoting the real operand to complex type and then performing a full
11921 complex addition. In implementations in which the @code{Signed_Zeros}
11922 attribute of the component type is @code{True} (and which therefore
11923 conform to IEC 559:1989 in regard to the handling of the sign of zero in
11924 predefined arithmetic operations), the latter technique will not
11925 generate the required result when the imaginary component of the complex
11926 operand is a negatively signed zero. (Explicit addition of the negative
11927 zero to the zero obtained during promotion yields a positive zero.)
11928 Analogous advice applies in the case of addition of a complex operand
11929 and a pure-imaginary operand, and in the case of subtraction of a
11930 complex operand and a real or pure-imaginary operand.
11936 Implementations in which @code{Real'Signed_Zeros} is @code{True} should
11937 attempt to provide a rational treatment of the signs of zero results and
11938 result components. As one example, the result of the @code{Argument}
11939 function should have the sign of the imaginary component of the
11940 parameter @code{X} when the point represented by that parameter lies on
11941 the positive real axis; as another, the sign of the imaginary component
11942 of the @code{Compose_From_Polar} function should be the same as
11943 (respectively, the opposite of) that of the @code{Argument} parameter when that
11944 parameter has a value of zero and the @code{Modulus} parameter has a
11945 nonnegative (respectively, negative) value.
11949 @cindex Complex elementary functions
11950 @unnumberedsec G.1.2(49): Complex Elementary Functions
11953 Implementations in which @code{Complex_Types.Real'Signed_Zeros} is
11954 @code{True} should attempt to provide a rational treatment of the signs
11955 of zero results and result components. For example, many of the complex
11956 elementary functions have components that are odd functions of one of
11957 the parameter components; in these cases, the result component should
11958 have the sign of the parameter component at the origin. Other complex
11959 elementary functions have zero components whose sign is opposite that of
11960 a parameter component at the origin, or is always positive or always
11965 @cindex Accuracy requirements
11966 @unnumberedsec G.2.4(19): Accuracy Requirements
11969 The versions of the forward trigonometric functions without a
11970 @code{Cycle} parameter should not be implemented by calling the
11971 corresponding version with a @code{Cycle} parameter of
11972 @code{2.0*Numerics.Pi}, since this will not provide the required
11973 accuracy in some portions of the domain. For the same reason, the
11974 version of @code{Log} without a @code{Base} parameter should not be
11975 implemented by calling the corresponding version with a @code{Base}
11976 parameter of @code{Numerics.e}.
11980 @cindex Complex arithmetic accuracy
11981 @cindex Accuracy, complex arithmetic
11982 @unnumberedsec G.2.6(15): Complex Arithmetic Accuracy
11986 The version of the @code{Compose_From_Polar} function without a
11987 @code{Cycle} parameter should not be implemented by calling the
11988 corresponding version with a @code{Cycle} parameter of
11989 @code{2.0*Numerics.Pi}, since this will not provide the required
11990 accuracy in some portions of the domain.
11994 @cindex Sequential elaboration policy
11995 @unnumberedsec H.6(15/2): Pragma Partition_Elaboration_Policy
11999 If the partition elaboration policy is @code{Sequential} and the
12000 Environment task becomes permanently blocked during elaboration then the
12001 partition is deadlocked and it is recommended that the partition be
12002 immediately terminated.
12006 @c -----------------------------------------
12007 @node Implementation Defined Characteristics
12008 @chapter Implementation Defined Characteristics
12011 In addition to the implementation dependent pragmas and attributes, and the
12012 implementation advice, there are a number of other Ada features that are
12013 potentially implementation dependent and are designated as
12014 implementation-defined. These are mentioned throughout the Ada Reference
12015 Manual, and are summarized in Annex M@.
12017 A requirement for conforming Ada compilers is that they provide
12018 documentation describing how the implementation deals with each of these
12019 issues. In this chapter, you will find each point in Annex M listed
12020 followed by a description in italic font of how GNAT
12021 handles the implementation dependence.
12023 You can use this chapter as a guide to minimizing implementation
12024 dependent features in your programs if portability to other compilers
12025 and other operating systems is an important consideration. The numbers
12026 in each section below correspond to the paragraph number in the Ada
12032 @strong{2}. Whether or not each recommendation given in Implementation
12033 Advice is followed. See 1.1.2(37).
12036 @xref{Implementation Advice}.
12041 @strong{3}. Capacity limitations of the implementation. See 1.1.3(3).
12044 The complexity of programs that can be processed is limited only by the
12045 total amount of available virtual memory, and disk space for the
12046 generated object files.
12051 @strong{4}. Variations from the standard that are impractical to avoid
12052 given the implementation's execution environment. See 1.1.3(6).
12055 There are no variations from the standard.
12060 @strong{5}. Which @code{code_statement}s cause external
12061 interactions. See 1.1.3(10).
12064 Any @code{code_statement} can potentially cause external interactions.
12069 @strong{6}. The coded representation for the text of an Ada
12070 program. See 2.1(4).
12073 See separate section on source representation.
12078 @strong{7}. The control functions allowed in comments. See 2.1(14).
12081 See separate section on source representation.
12086 @strong{8}. The representation for an end of line. See 2.2(2).
12089 See separate section on source representation.
12094 @strong{9}. Maximum supported line length and lexical element
12095 length. See 2.2(15).
12098 The maximum line length is 255 characters and the maximum length of
12099 a lexical element is also 255 characters. This is the default setting
12100 if not overridden by the use of compiler switch @option{-gnaty} (which
12101 sets the maximum to 79) or @option{-gnatyMnn} which allows the maximum
12102 line length to be specified to be any value up to 32767. The maximum
12103 length of a lexical element is the same as the maximum line length.
12108 @strong{10}. Implementation defined pragmas. See 2.8(14).
12112 @xref{Implementation Defined Pragmas}.
12117 @strong{11}. Effect of pragma @code{Optimize}. See 2.8(27).
12120 Pragma @code{Optimize}, if given with a @code{Time} or @code{Space}
12121 parameter, checks that the optimization flag is set, and aborts if it is
12127 @strong{12}. The sequence of characters of the value returned by
12128 @code{@var{S}'Image} when some of the graphic characters of
12129 @code{@var{S}'Wide_Image} are not defined in @code{Character}. See
12133 The sequence of characters is as defined by the wide character encoding
12134 method used for the source. See section on source representation for
12140 @strong{13}. The predefined integer types declared in
12141 @code{Standard}. See 3.5.4(25).
12145 @item Short_Short_Integer
12147 @item Short_Integer
12148 (Short) 16 bit signed
12152 64 bit signed (on most 64 bit targets, depending on the C definition of long).
12153 32 bit signed (all other targets)
12154 @item Long_Long_Integer
12161 @strong{14}. Any nonstandard integer types and the operators defined
12162 for them. See 3.5.4(26).
12165 There are no nonstandard integer types.
12170 @strong{15}. Any nonstandard real types and the operators defined for
12171 them. See 3.5.6(8).
12174 There are no nonstandard real types.
12179 @strong{16}. What combinations of requested decimal precision and range
12180 are supported for floating point types. See 3.5.7(7).
12183 The precision and range is as defined by the IEEE standard.
12188 @strong{17}. The predefined floating point types declared in
12189 @code{Standard}. See 3.5.7(16).
12196 (Short) 32 bit IEEE short
12199 @item Long_Long_Float
12200 64 bit IEEE long (80 bit IEEE long on x86 processors)
12206 @strong{18}. The small of an ordinary fixed point type. See 3.5.9(8).
12209 @code{Fine_Delta} is 2**(@minus{}63)
12214 @strong{19}. What combinations of small, range, and digits are
12215 supported for fixed point types. See 3.5.9(10).
12218 Any combinations are permitted that do not result in a small less than
12219 @code{Fine_Delta} and do not result in a mantissa larger than 63 bits.
12220 If the mantissa is larger than 53 bits on machines where Long_Long_Float
12221 is 64 bits (true of all architectures except ia32), then the output from
12222 Text_IO is accurate to only 53 bits, rather than the full mantissa. This
12223 is because floating-point conversions are used to convert fixed point.
12228 @strong{20}. The result of @code{Tags.Expanded_Name} for types declared
12229 within an unnamed @code{block_statement}. See 3.9(10).
12232 Block numbers of the form @code{B@var{nnn}}, where @var{nnn} is a
12233 decimal integer are allocated.
12238 @strong{21}. Implementation-defined attributes. See 4.1.4(12).
12241 @xref{Implementation Defined Attributes}.
12246 @strong{22}. Any implementation-defined time types. See 9.6(6).
12249 There are no implementation-defined time types.
12254 @strong{23}. The time base associated with relative delays.
12257 See 9.6(20). The time base used is that provided by the C library
12258 function @code{gettimeofday}.
12263 @strong{24}. The time base of the type @code{Calendar.Time}. See
12267 The time base used is that provided by the C library function
12268 @code{gettimeofday}.
12273 @strong{25}. The time zone used for package @code{Calendar}
12274 operations. See 9.6(24).
12277 The time zone used by package @code{Calendar} is the current system time zone
12278 setting for local time, as accessed by the C library function
12284 @strong{26}. Any limit on @code{delay_until_statements} of
12285 @code{select_statements}. See 9.6(29).
12288 There are no such limits.
12293 @strong{27}. Whether or not two non-overlapping parts of a composite
12294 object are independently addressable, in the case where packing, record
12295 layout, or @code{Component_Size} is specified for the object. See
12299 Separate components are independently addressable if they do not share
12300 overlapping storage units.
12305 @strong{28}. The representation for a compilation. See 10.1(2).
12308 A compilation is represented by a sequence of files presented to the
12309 compiler in a single invocation of the @command{gcc} command.
12314 @strong{29}. Any restrictions on compilations that contain multiple
12315 compilation_units. See 10.1(4).
12318 No single file can contain more than one compilation unit, but any
12319 sequence of files can be presented to the compiler as a single
12325 @strong{30}. The mechanisms for creating an environment and for adding
12326 and replacing compilation units. See 10.1.4(3).
12329 See separate section on compilation model.
12334 @strong{31}. The manner of explicitly assigning library units to a
12335 partition. See 10.2(2).
12338 If a unit contains an Ada main program, then the Ada units for the partition
12339 are determined by recursive application of the rules in the Ada Reference
12340 Manual section 10.2(2-6). In other words, the Ada units will be those that
12341 are needed by the main program, and then this definition of need is applied
12342 recursively to those units, and the partition contains the transitive
12343 closure determined by this relationship. In short, all the necessary units
12344 are included, with no need to explicitly specify the list. If additional
12345 units are required, e.g.@: by foreign language units, then all units must be
12346 mentioned in the context clause of one of the needed Ada units.
12348 If the partition contains no main program, or if the main program is in
12349 a language other than Ada, then GNAT
12350 provides the binder options @option{-z} and @option{-n} respectively, and in
12351 this case a list of units can be explicitly supplied to the binder for
12352 inclusion in the partition (all units needed by these units will also
12353 be included automatically). For full details on the use of these
12354 options, refer to @ref{The GNAT Make Program gnatmake,,, gnat_ugn,
12355 @value{EDITION} User's Guide}.
12360 @strong{32}. The implementation-defined means, if any, of specifying
12361 which compilation units are needed by a given compilation unit. See
12365 The units needed by a given compilation unit are as defined in
12366 the Ada Reference Manual section 10.2(2-6). There are no
12367 implementation-defined pragmas or other implementation-defined
12368 means for specifying needed units.
12373 @strong{33}. The manner of designating the main subprogram of a
12374 partition. See 10.2(7).
12377 The main program is designated by providing the name of the
12378 corresponding @file{ALI} file as the input parameter to the binder.
12383 @strong{34}. The order of elaboration of @code{library_items}. See
12387 The first constraint on ordering is that it meets the requirements of
12388 Chapter 10 of the Ada Reference Manual. This still leaves some
12389 implementation dependent choices, which are resolved by first
12390 elaborating bodies as early as possible (i.e., in preference to specs
12391 where there is a choice), and second by evaluating the immediate with
12392 clauses of a unit to determine the probably best choice, and
12393 third by elaborating in alphabetical order of unit names
12394 where a choice still remains.
12399 @strong{35}. Parameter passing and function return for the main
12400 subprogram. See 10.2(21).
12403 The main program has no parameters. It may be a procedure, or a function
12404 returning an integer type. In the latter case, the returned integer
12405 value is the return code of the program (overriding any value that
12406 may have been set by a call to @code{Ada.Command_Line.Set_Exit_Status}).
12411 @strong{36}. The mechanisms for building and running partitions. See
12415 GNAT itself supports programs with only a single partition. The GNATDIST
12416 tool provided with the GLADE package (which also includes an implementation
12417 of the PCS) provides a completely flexible method for building and running
12418 programs consisting of multiple partitions. See the separate GLADE manual
12424 @strong{37}. The details of program execution, including program
12425 termination. See 10.2(25).
12428 See separate section on compilation model.
12433 @strong{38}. The semantics of any non-active partitions supported by the
12434 implementation. See 10.2(28).
12437 Passive partitions are supported on targets where shared memory is
12438 provided by the operating system. See the GLADE reference manual for
12444 @strong{39}. The information returned by @code{Exception_Message}. See
12448 Exception message returns the null string unless a specific message has
12449 been passed by the program.
12454 @strong{40}. The result of @code{Exceptions.Exception_Name} for types
12455 declared within an unnamed @code{block_statement}. See 11.4.1(12).
12458 Blocks have implementation defined names of the form @code{B@var{nnn}}
12459 where @var{nnn} is an integer.
12464 @strong{41}. The information returned by
12465 @code{Exception_Information}. See 11.4.1(13).
12468 @code{Exception_Information} returns a string in the following format:
12471 @emph{Exception_Name:} nnnnn
12472 @emph{Message:} mmmmm
12474 @emph{Load address:} 0xhhhh
12475 @emph{Call stack traceback locations:}
12476 0xhhhh 0xhhhh 0xhhhh ... 0xhhh
12484 @code{nnnn} is the fully qualified name of the exception in all upper
12485 case letters. This line is always present.
12488 @code{mmmm} is the message (this line present only if message is non-null)
12491 @code{ppp} is the Process Id value as a decimal integer (this line is
12492 present only if the Process Id is nonzero). Currently we are
12493 not making use of this field.
12496 The Load address line, the Call stack traceback locations line and the
12497 following values are present only if at least one traceback location was
12498 recorded. The Load address indicates the address at which the main executable
12499 was loaded; this line may not be present if operating system hasn't relocated
12500 the main executable. The values are given in C style format, with lower case
12501 letters for a-f, and only as many digits present as are necessary.
12505 The line terminator sequence at the end of each line, including
12506 the last line is a single @code{LF} character (@code{16#0A#}).
12511 @strong{42}. Implementation-defined check names. See 11.5(27).
12514 The implementation defined check name Alignment_Check controls checking of
12515 address clause values for proper alignment (that is, the address supplied
12516 must be consistent with the alignment of the type).
12518 The implementation defined check name Predicate_Check controls whether
12519 predicate checks are generated.
12521 The implementation defined check name Validity_Check controls whether
12522 validity checks are generated.
12524 In addition, a user program can add implementation-defined check names
12525 by means of the pragma Check_Name.
12530 @strong{43}. The interpretation of each aspect of representation. See
12534 See separate section on data representations.
12539 @strong{44}. Any restrictions placed upon representation items. See
12543 See separate section on data representations.
12548 @strong{45}. The meaning of @code{Size} for indefinite subtypes. See
12552 Size for an indefinite subtype is the maximum possible size, except that
12553 for the case of a subprogram parameter, the size of the parameter object
12554 is the actual size.
12559 @strong{46}. The default external representation for a type tag. See
12563 The default external representation for a type tag is the fully expanded
12564 name of the type in upper case letters.
12569 @strong{47}. What determines whether a compilation unit is the same in
12570 two different partitions. See 13.3(76).
12573 A compilation unit is the same in two different partitions if and only
12574 if it derives from the same source file.
12579 @strong{48}. Implementation-defined components. See 13.5.1(15).
12582 The only implementation defined component is the tag for a tagged type,
12583 which contains a pointer to the dispatching table.
12588 @strong{49}. If @code{Word_Size} = @code{Storage_Unit}, the default bit
12589 ordering. See 13.5.3(5).
12592 @code{Word_Size} (32) is not the same as @code{Storage_Unit} (8) for this
12593 implementation, so no non-default bit ordering is supported. The default
12594 bit ordering corresponds to the natural endianness of the target architecture.
12599 @strong{50}. The contents of the visible part of package @code{System}
12600 and its language-defined children. See 13.7(2).
12603 See the definition of these packages in files @file{system.ads} and
12604 @file{s-stoele.ads}.
12609 @strong{51}. The contents of the visible part of package
12610 @code{System.Machine_Code}, and the meaning of
12611 @code{code_statements}. See 13.8(7).
12614 See the definition and documentation in file @file{s-maccod.ads}.
12619 @strong{52}. The effect of unchecked conversion. See 13.9(11).
12622 Unchecked conversion between types of the same size
12623 results in an uninterpreted transmission of the bits from one type
12624 to the other. If the types are of unequal sizes, then in the case of
12625 discrete types, a shorter source is first zero or sign extended as
12626 necessary, and a shorter target is simply truncated on the left.
12627 For all non-discrete types, the source is first copied if necessary
12628 to ensure that the alignment requirements of the target are met, then
12629 a pointer is constructed to the source value, and the result is obtained
12630 by dereferencing this pointer after converting it to be a pointer to the
12631 target type. Unchecked conversions where the target subtype is an
12632 unconstrained array are not permitted. If the target alignment is
12633 greater than the source alignment, then a copy of the result is
12634 made with appropriate alignment
12639 @strong{53}. The semantics of operations on invalid representations.
12643 For assignments and other operations where the use of invalid values cannot
12644 result in erroneous behavior, the compiler ignores the possibility of invalid
12645 values. An exception is raised at the point where an invalid value would
12646 result in erroneous behavior. For example executing:
12648 @smallexample @c ada
12649 procedure invalidvals is
12651 Y : Natural range 1 .. 10;
12652 for Y'Address use X'Address;
12653 Z : Natural range 1 .. 10;
12654 A : array (Natural range 1 .. 10) of Integer;
12656 Z := Y; -- no exception
12657 A (Z) := 3; -- exception raised;
12662 As indicated, an exception is raised on the array assignment, but not
12663 on the simple assignment of the invalid negative value from Y to Z.
12668 @strong{53}. The manner of choosing a storage pool for an access type
12669 when @code{Storage_Pool} is not specified for the type. See 13.11(17).
12672 There are 3 different standard pools used by the compiler when
12673 @code{Storage_Pool} is not specified depending whether the type is local
12674 to a subprogram or defined at the library level and whether
12675 @code{Storage_Size}is specified or not. See documentation in the runtime
12676 library units @code{System.Pool_Global}, @code{System.Pool_Size} and
12677 @code{System.Pool_Local} in files @file{s-poosiz.ads},
12678 @file{s-pooglo.ads} and @file{s-pooloc.ads} for full details on the
12679 default pools used.
12684 @strong{54}. Whether or not the implementation provides user-accessible
12685 names for the standard pool type(s). See 13.11(17).
12689 See documentation in the sources of the run time mentioned in paragraph
12690 @strong{53} . All these pools are accessible by means of @code{with}'ing
12696 @strong{55}. The meaning of @code{Storage_Size}. See 13.11(18).
12699 @code{Storage_Size} is measured in storage units, and refers to the
12700 total space available for an access type collection, or to the primary
12701 stack space for a task.
12706 @strong{56}. Implementation-defined aspects of storage pools. See
12710 See documentation in the sources of the run time mentioned in paragraph
12711 @strong{53} for details on GNAT-defined aspects of storage pools.
12716 @strong{57}. The set of restrictions allowed in a pragma
12717 @code{Restrictions}. See 13.12(7).
12720 @xref{Standard and Implementation Defined Restrictions}.
12725 @strong{58}. The consequences of violating limitations on
12726 @code{Restrictions} pragmas. See 13.12(9).
12729 Restrictions that can be checked at compile time result in illegalities
12730 if violated. Currently there are no other consequences of violating
12736 @strong{59}. The representation used by the @code{Read} and
12737 @code{Write} attributes of elementary types in terms of stream
12738 elements. See 13.13.2(9).
12741 The representation is the in-memory representation of the base type of
12742 the type, using the number of bits corresponding to the
12743 @code{@var{type}'Size} value, and the natural ordering of the machine.
12748 @strong{60}. The names and characteristics of the numeric subtypes
12749 declared in the visible part of package @code{Standard}. See A.1(3).
12752 See items describing the integer and floating-point types supported.
12757 @strong{61}. The string returned by @code{Character_Set_Version}.
12761 @code{Ada.Wide_Characters.Handling.Character_Set_Version} returns
12762 the string "Unicode 4.0", referring to version 4.0 of the
12763 Unicode specification.
12768 @strong{62}. The accuracy actually achieved by the elementary
12769 functions. See A.5.1(1).
12772 The elementary functions correspond to the functions available in the C
12773 library. Only fast math mode is implemented.
12778 @strong{63}. The sign of a zero result from some of the operators or
12779 functions in @code{Numerics.Generic_Elementary_Functions}, when
12780 @code{Float_Type'Signed_Zeros} is @code{True}. See A.5.1(46).
12783 The sign of zeroes follows the requirements of the IEEE 754 standard on
12789 @strong{64}. The value of
12790 @code{Numerics.Float_Random.Max_Image_Width}. See A.5.2(27).
12793 Maximum image width is 6864, see library file @file{s-rannum.ads}.
12798 @strong{65}. The value of
12799 @code{Numerics.Discrete_Random.Max_Image_Width}. See A.5.2(27).
12802 Maximum image width is 6864, see library file @file{s-rannum.ads}.
12807 @strong{66}. The algorithms for random number generation. See
12811 The algorithm is the Mersenne Twister, as documented in the source file
12812 @file{s-rannum.adb}. This version of the algorithm has a period of
12818 @strong{67}. The string representation of a random number generator's
12819 state. See A.5.2(38).
12822 The value returned by the Image function is the concatenation of
12823 the fixed-width decimal representations of the 624 32-bit integers
12824 of the state vector.
12829 @strong{68}. The minimum time interval between calls to the
12830 time-dependent Reset procedure that are guaranteed to initiate different
12831 random number sequences. See A.5.2(45).
12834 The minimum period between reset calls to guarantee distinct series of
12835 random numbers is one microsecond.
12840 @strong{69}. The values of the @code{Model_Mantissa},
12841 @code{Model_Emin}, @code{Model_Epsilon}, @code{Model},
12842 @code{Safe_First}, and @code{Safe_Last} attributes, if the Numerics
12843 Annex is not supported. See A.5.3(72).
12846 Run the compiler with @option{-gnatS} to produce a listing of package
12847 @code{Standard}, has the values of all numeric attributes.
12852 @strong{70}. Any implementation-defined characteristics of the
12853 input-output packages. See A.7(14).
12856 There are no special implementation defined characteristics for these
12862 @strong{71}. The value of @code{Buffer_Size} in @code{Storage_IO}. See
12866 All type representations are contiguous, and the @code{Buffer_Size} is
12867 the value of @code{@var{type}'Size} rounded up to the next storage unit
12873 @strong{72}. External files for standard input, standard output, and
12874 standard error See A.10(5).
12877 These files are mapped onto the files provided by the C streams
12878 libraries. See source file @file{i-cstrea.ads} for further details.
12883 @strong{73}. The accuracy of the value produced by @code{Put}. See
12887 If more digits are requested in the output than are represented by the
12888 precision of the value, zeroes are output in the corresponding least
12889 significant digit positions.
12894 @strong{74}. The meaning of @code{Argument_Count}, @code{Argument}, and
12895 @code{Command_Name}. See A.15(1).
12898 These are mapped onto the @code{argv} and @code{argc} parameters of the
12899 main program in the natural manner.
12904 @strong{75}. The interpretation of the @code{Form} parameter in procedure
12905 @code{Create_Directory}. See A.16(56).
12908 The @code{Form} parameter is not used.
12913 @strong{76}. The interpretation of the @code{Form} parameter in procedure
12914 @code{Create_Path}. See A.16(60).
12917 The @code{Form} parameter is not used.
12922 @strong{77}. The interpretation of the @code{Form} parameter in procedure
12923 @code{Copy_File}. See A.16(68).
12926 The @code{Form} parameter is case-insensitive.
12928 Two fields are recognized in the @code{Form} parameter:
12932 @item preserve=<value>
12939 <value> starts immediately after the character '=' and ends with the
12940 character immediately preceding the next comma (',') or with the last
12941 character of the parameter.
12943 The only possible values for preserve= are:
12947 @item no_attributes
12948 Do not try to preserve any file attributes. This is the default if no
12949 preserve= is found in Form.
12951 @item all_attributes
12952 Try to preserve all file attributes (timestamps, access rights).
12955 Preserve the timestamp of the copied file, but not the other file attributes.
12960 The only possible values for mode= are:
12965 Only do the copy if the destination file does not already exist. If it already
12966 exists, Copy_File fails.
12969 Copy the file in all cases. Overwrite an already existing destination file.
12972 Append the original file to the destination file. If the destination file does
12973 not exist, the destination file is a copy of the source file. When mode=append,
12974 the field preserve=, if it exists, is not taken into account.
12979 If the Form parameter includes one or both of the fields and the value or
12980 values are incorrect, Copy_file fails with Use_Error.
12982 Examples of correct Forms:
12985 Form => "preserve=no_attributes,mode=overwrite" (the default)
12986 Form => "mode=append"
12987 Form => "mode=copy, preserve=all_attributes"
12991 Examples of incorrect Forms
12994 Form => "preserve=junk"
12995 Form => "mode=internal, preserve=timestamps"
13001 @strong{78}. Implementation-defined convention names. See B.1(11).
13004 The following convention names are supported
13009 @item Ada_Pass_By_Copy
13010 Allowed for any types except by-reference types such as limited
13011 records. Compatible with convention Ada, but causes any parameters
13012 with this convention to be passed by copy.
13013 @item Ada_Pass_By_Reference
13014 Allowed for any types except by-copy types such as scalars.
13015 Compatible with convention Ada, but causes any parameters
13016 with this convention to be passed by reference.
13020 Synonym for Assembler
13022 Synonym for Assembler
13025 @item C_Pass_By_Copy
13026 Allowed only for record types, like C, but also notes that record
13027 is to be passed by copy rather than reference.
13030 @item C_Plus_Plus (or CPP)
13033 Treated the same as C
13035 Treated the same as C
13039 For support of pragma @code{Import} with convention Intrinsic, see
13040 separate section on Intrinsic Subprograms.
13042 Stdcall (used for Windows implementations only). This convention correspond
13043 to the WINAPI (previously called Pascal convention) C/C++ convention under
13044 Windows. A routine with this convention cleans the stack before
13045 exit. This pragma cannot be applied to a dispatching call.
13047 Synonym for Stdcall
13049 Synonym for Stdcall
13051 Stubbed is a special convention used to indicate that the body of the
13052 subprogram will be entirely ignored. Any call to the subprogram
13053 is converted into a raise of the @code{Program_Error} exception. If a
13054 pragma @code{Import} specifies convention @code{stubbed} then no body need
13055 be present at all. This convention is useful during development for the
13056 inclusion of subprograms whose body has not yet been written.
13060 In addition, all otherwise unrecognized convention names are also
13061 treated as being synonymous with convention C@. In all implementations
13062 except for VMS, use of such other names results in a warning. In VMS
13063 implementations, these names are accepted silently.
13068 @strong{79}. The meaning of link names. See B.1(36).
13071 Link names are the actual names used by the linker.
13076 @strong{80}. The manner of choosing link names when neither the link
13077 name nor the address of an imported or exported entity is specified. See
13081 The default linker name is that which would be assigned by the relevant
13082 external language, interpreting the Ada name as being in all lower case
13088 @strong{81}. The effect of pragma @code{Linker_Options}. See B.1(37).
13091 The string passed to @code{Linker_Options} is presented uninterpreted as
13092 an argument to the link command, unless it contains ASCII.NUL characters.
13093 NUL characters if they appear act as argument separators, so for example
13095 @smallexample @c ada
13096 pragma Linker_Options ("-labc" & ASCII.NUL & "-ldef");
13100 causes two separate arguments @code{-labc} and @code{-ldef} to be passed to the
13101 linker. The order of linker options is preserved for a given unit. The final
13102 list of options passed to the linker is in reverse order of the elaboration
13103 order. For example, linker options for a body always appear before the options
13104 from the corresponding package spec.
13109 @strong{82}. The contents of the visible part of package
13110 @code{Interfaces} and its language-defined descendants. See B.2(1).
13113 See files with prefix @file{i-} in the distributed library.
13118 @strong{83}. Implementation-defined children of package
13119 @code{Interfaces}. The contents of the visible part of package
13120 @code{Interfaces}. See B.2(11).
13123 See files with prefix @file{i-} in the distributed library.
13128 @strong{84}. The types @code{Floating}, @code{Long_Floating},
13129 @code{Binary}, @code{Long_Binary}, @code{Decimal_ Element}, and
13130 @code{COBOL_Character}; and the initialization of the variables
13131 @code{Ada_To_COBOL} and @code{COBOL_To_Ada}, in
13132 @code{Interfaces.COBOL}. See B.4(50).
13138 @item Long_Floating
13139 (Floating) Long_Float
13144 @item Decimal_Element
13146 @item COBOL_Character
13151 For initialization, see the file @file{i-cobol.ads} in the distributed library.
13156 @strong{85}. Support for access to machine instructions. See C.1(1).
13159 See documentation in file @file{s-maccod.ads} in the distributed library.
13164 @strong{86}. Implementation-defined aspects of access to machine
13165 operations. See C.1(9).
13168 See documentation in file @file{s-maccod.ads} in the distributed library.
13173 @strong{87}. Implementation-defined aspects of interrupts. See C.3(2).
13176 Interrupts are mapped to signals or conditions as appropriate. See
13178 @code{Ada.Interrupt_Names} in source file @file{a-intnam.ads} for details
13179 on the interrupts supported on a particular target.
13184 @strong{88}. Implementation-defined aspects of pre-elaboration. See
13188 GNAT does not permit a partition to be restarted without reloading,
13189 except under control of the debugger.
13194 @strong{89}. The semantics of pragma @code{Discard_Names}. See C.5(7).
13197 Pragma @code{Discard_Names} causes names of enumeration literals to
13198 be suppressed. In the presence of this pragma, the Image attribute
13199 provides the image of the Pos of the literal, and Value accepts
13205 @strong{90}. The result of the @code{Task_Identification.Image}
13206 attribute. See C.7.1(7).
13209 The result of this attribute is a string that identifies
13210 the object or component that denotes a given task. If a variable @code{Var}
13211 has a task type, the image for this task will have the form @code{Var_@var{XXXXXXXX}},
13213 is the hexadecimal representation of the virtual address of the corresponding
13214 task control block. If the variable is an array of tasks, the image of each
13215 task will have the form of an indexed component indicating the position of a
13216 given task in the array, e.g.@: @code{Group(5)_@var{XXXXXXX}}. If the task is a
13217 component of a record, the image of the task will have the form of a selected
13218 component. These rules are fully recursive, so that the image of a task that
13219 is a subcomponent of a composite object corresponds to the expression that
13220 designates this task.
13222 If a task is created by an allocator, its image depends on the context. If the
13223 allocator is part of an object declaration, the rules described above are used
13224 to construct its image, and this image is not affected by subsequent
13225 assignments. If the allocator appears within an expression, the image
13226 includes only the name of the task type.
13228 If the configuration pragma Discard_Names is present, or if the restriction
13229 No_Implicit_Heap_Allocation is in effect, the image reduces to
13230 the numeric suffix, that is to say the hexadecimal representation of the
13231 virtual address of the control block of the task.
13235 @strong{91}. The value of @code{Current_Task} when in a protected entry
13236 or interrupt handler. See C.7.1(17).
13239 Protected entries or interrupt handlers can be executed by any
13240 convenient thread, so the value of @code{Current_Task} is undefined.
13245 @strong{92}. The effect of calling @code{Current_Task} from an entry
13246 body or interrupt handler. See C.7.1(19).
13249 The effect of calling @code{Current_Task} from an entry body or
13250 interrupt handler is to return the identification of the task currently
13251 executing the code.
13256 @strong{93}. Implementation-defined aspects of
13257 @code{Task_Attributes}. See C.7.2(19).
13260 There are no implementation-defined aspects of @code{Task_Attributes}.
13265 @strong{94}. Values of all @code{Metrics}. See D(2).
13268 The metrics information for GNAT depends on the performance of the
13269 underlying operating system. The sources of the run-time for tasking
13270 implementation, together with the output from @option{-gnatG} can be
13271 used to determine the exact sequence of operating systems calls made
13272 to implement various tasking constructs. Together with appropriate
13273 information on the performance of the underlying operating system,
13274 on the exact target in use, this information can be used to determine
13275 the required metrics.
13280 @strong{95}. The declarations of @code{Any_Priority} and
13281 @code{Priority}. See D.1(11).
13284 See declarations in file @file{system.ads}.
13289 @strong{96}. Implementation-defined execution resources. See D.1(15).
13292 There are no implementation-defined execution resources.
13297 @strong{97}. Whether, on a multiprocessor, a task that is waiting for
13298 access to a protected object keeps its processor busy. See D.2.1(3).
13301 On a multi-processor, a task that is waiting for access to a protected
13302 object does not keep its processor busy.
13307 @strong{98}. The affect of implementation defined execution resources
13308 on task dispatching. See D.2.1(9).
13311 Tasks map to threads in the threads package used by GNAT@. Where possible
13312 and appropriate, these threads correspond to native threads of the
13313 underlying operating system.
13318 @strong{99}. Implementation-defined @code{policy_identifiers} allowed
13319 in a pragma @code{Task_Dispatching_Policy}. See D.2.2(3).
13322 There are no implementation-defined policy-identifiers allowed in this
13328 @strong{100}. Implementation-defined aspects of priority inversion. See
13332 Execution of a task cannot be preempted by the implementation processing
13333 of delay expirations for lower priority tasks.
13338 @strong{101}. Implementation-defined task dispatching. See D.2.2(18).
13341 The policy is the same as that of the underlying threads implementation.
13346 @strong{102}. Implementation-defined @code{policy_identifiers} allowed
13347 in a pragma @code{Locking_Policy}. See D.3(4).
13350 The two implementation defined policies permitted in GNAT are
13351 @code{Inheritance_Locking} and @code{Conccurent_Readers_Locking}. On
13352 targets that support the @code{Inheritance_Locking} policy, locking is
13353 implemented by inheritance, i.e.@: the task owning the lock operates
13354 at a priority equal to the highest priority of any task currently
13355 requesting the lock. On targets that support the
13356 @code{Conccurent_Readers_Locking} policy, locking is implemented with a
13357 read/write lock allowing multiple propected object functions to enter
13363 @strong{103}. Default ceiling priorities. See D.3(10).
13366 The ceiling priority of protected objects of the type
13367 @code{System.Interrupt_Priority'Last} as described in the Ada
13368 Reference Manual D.3(10),
13373 @strong{104}. The ceiling of any protected object used internally by
13374 the implementation. See D.3(16).
13377 The ceiling priority of internal protected objects is
13378 @code{System.Priority'Last}.
13383 @strong{105}. Implementation-defined queuing policies. See D.4(1).
13386 There are no implementation-defined queuing policies.
13391 @strong{106}. On a multiprocessor, any conditions that cause the
13392 completion of an aborted construct to be delayed later than what is
13393 specified for a single processor. See D.6(3).
13396 The semantics for abort on a multi-processor is the same as on a single
13397 processor, there are no further delays.
13402 @strong{107}. Any operations that implicitly require heap storage
13403 allocation. See D.7(8).
13406 The only operation that implicitly requires heap storage allocation is
13412 @strong{108}. Implementation-defined aspects of pragma
13413 @code{Restrictions}. See D.7(20).
13416 There are no such implementation-defined aspects.
13421 @strong{109}. Implementation-defined aspects of package
13422 @code{Real_Time}. See D.8(17).
13425 There are no implementation defined aspects of package @code{Real_Time}.
13430 @strong{110}. Implementation-defined aspects of
13431 @code{delay_statements}. See D.9(8).
13434 Any difference greater than one microsecond will cause the task to be
13435 delayed (see D.9(7)).
13440 @strong{111}. The upper bound on the duration of interrupt blocking
13441 caused by the implementation. See D.12(5).
13444 The upper bound is determined by the underlying operating system. In
13445 no cases is it more than 10 milliseconds.
13450 @strong{112}. The means for creating and executing distributed
13451 programs. See E(5).
13454 The GLADE package provides a utility GNATDIST for creating and executing
13455 distributed programs. See the GLADE reference manual for further details.
13460 @strong{113}. Any events that can result in a partition becoming
13461 inaccessible. See E.1(7).
13464 See the GLADE reference manual for full details on such events.
13469 @strong{114}. The scheduling policies, treatment of priorities, and
13470 management of shared resources between partitions in certain cases. See
13474 See the GLADE reference manual for full details on these aspects of
13475 multi-partition execution.
13480 @strong{115}. Events that cause the version of a compilation unit to
13481 change. See E.3(5).
13484 Editing the source file of a compilation unit, or the source files of
13485 any units on which it is dependent in a significant way cause the version
13486 to change. No other actions cause the version number to change. All changes
13487 are significant except those which affect only layout, capitalization or
13493 @strong{116}. Whether the execution of the remote subprogram is
13494 immediately aborted as a result of cancellation. See E.4(13).
13497 See the GLADE reference manual for details on the effect of abort in
13498 a distributed application.
13503 @strong{117}. Implementation-defined aspects of the PCS@. See E.5(25).
13506 See the GLADE reference manual for a full description of all implementation
13507 defined aspects of the PCS@.
13512 @strong{118}. Implementation-defined interfaces in the PCS@. See
13516 See the GLADE reference manual for a full description of all
13517 implementation defined interfaces.
13522 @strong{119}. The values of named numbers in the package
13523 @code{Decimal}. See F.2(7).
13535 @item Max_Decimal_Digits
13542 @strong{120}. The value of @code{Max_Picture_Length} in the package
13543 @code{Text_IO.Editing}. See F.3.3(16).
13551 @strong{121}. The value of @code{Max_Picture_Length} in the package
13552 @code{Wide_Text_IO.Editing}. See F.3.4(5).
13560 @strong{122}. The accuracy actually achieved by the complex elementary
13561 functions and by other complex arithmetic operations. See G.1(1).
13564 Standard library functions are used for the complex arithmetic
13565 operations. Only fast math mode is currently supported.
13570 @strong{123}. The sign of a zero result (or a component thereof) from
13571 any operator or function in @code{Numerics.Generic_Complex_Types}, when
13572 @code{Real'Signed_Zeros} is True. See G.1.1(53).
13575 The signs of zero values are as recommended by the relevant
13576 implementation advice.
13581 @strong{124}. The sign of a zero result (or a component thereof) from
13582 any operator or function in
13583 @code{Numerics.Generic_Complex_Elementary_Functions}, when
13584 @code{Real'Signed_Zeros} is @code{True}. See G.1.2(45).
13587 The signs of zero values are as recommended by the relevant
13588 implementation advice.
13593 @strong{125}. Whether the strict mode or the relaxed mode is the
13594 default. See G.2(2).
13597 The strict mode is the default. There is no separate relaxed mode. GNAT
13598 provides a highly efficient implementation of strict mode.
13603 @strong{126}. The result interval in certain cases of fixed-to-float
13604 conversion. See G.2.1(10).
13607 For cases where the result interval is implementation dependent, the
13608 accuracy is that provided by performing all operations in 64-bit IEEE
13609 floating-point format.
13614 @strong{127}. The result of a floating point arithmetic operation in
13615 overflow situations, when the @code{Machine_Overflows} attribute of the
13616 result type is @code{False}. See G.2.1(13).
13619 Infinite and NaN values are produced as dictated by the IEEE
13620 floating-point standard.
13622 Note that on machines that are not fully compliant with the IEEE
13623 floating-point standard, such as Alpha, the @option{-mieee} compiler flag
13624 must be used for achieving IEEE conforming behavior (although at the cost
13625 of a significant performance penalty), so infinite and NaN values are
13626 properly generated.
13631 @strong{128}. The result interval for division (or exponentiation by a
13632 negative exponent), when the floating point hardware implements division
13633 as multiplication by a reciprocal. See G.2.1(16).
13636 Not relevant, division is IEEE exact.
13641 @strong{129}. The definition of close result set, which determines the
13642 accuracy of certain fixed point multiplications and divisions. See
13646 Operations in the close result set are performed using IEEE long format
13647 floating-point arithmetic. The input operands are converted to
13648 floating-point, the operation is done in floating-point, and the result
13649 is converted to the target type.
13654 @strong{130}. Conditions on a @code{universal_real} operand of a fixed
13655 point multiplication or division for which the result shall be in the
13656 perfect result set. See G.2.3(22).
13659 The result is only defined to be in the perfect result set if the result
13660 can be computed by a single scaling operation involving a scale factor
13661 representable in 64-bits.
13666 @strong{131}. The result of a fixed point arithmetic operation in
13667 overflow situations, when the @code{Machine_Overflows} attribute of the
13668 result type is @code{False}. See G.2.3(27).
13671 Not relevant, @code{Machine_Overflows} is @code{True} for fixed-point
13677 @strong{132}. The result of an elementary function reference in
13678 overflow situations, when the @code{Machine_Overflows} attribute of the
13679 result type is @code{False}. See G.2.4(4).
13682 IEEE infinite and Nan values are produced as appropriate.
13687 @strong{133}. The value of the angle threshold, within which certain
13688 elementary functions, complex arithmetic operations, and complex
13689 elementary functions yield results conforming to a maximum relative
13690 error bound. See G.2.4(10).
13693 Information on this subject is not yet available.
13698 @strong{134}. The accuracy of certain elementary functions for
13699 parameters beyond the angle threshold. See G.2.4(10).
13702 Information on this subject is not yet available.
13707 @strong{135}. The result of a complex arithmetic operation or complex
13708 elementary function reference in overflow situations, when the
13709 @code{Machine_Overflows} attribute of the corresponding real type is
13710 @code{False}. See G.2.6(5).
13713 IEEE infinite and Nan values are produced as appropriate.
13718 @strong{136}. The accuracy of certain complex arithmetic operations and
13719 certain complex elementary functions for parameters (or components
13720 thereof) beyond the angle threshold. See G.2.6(8).
13723 Information on those subjects is not yet available.
13728 @strong{137}. Information regarding bounded errors and erroneous
13729 execution. See H.2(1).
13732 Information on this subject is not yet available.
13737 @strong{138}. Implementation-defined aspects of pragma
13738 @code{Inspection_Point}. See H.3.2(8).
13741 Pragma @code{Inspection_Point} ensures that the variable is live and can
13742 be examined by the debugger at the inspection point.
13747 @strong{139}. Implementation-defined aspects of pragma
13748 @code{Restrictions}. See H.4(25).
13751 There are no implementation-defined aspects of pragma @code{Restrictions}. The
13752 use of pragma @code{Restrictions [No_Exceptions]} has no effect on the
13753 generated code. Checks must suppressed by use of pragma @code{Suppress}.
13758 @strong{140}. Any restrictions on pragma @code{Restrictions}. See
13762 There are no restrictions on pragma @code{Restrictions}.
13764 @node Intrinsic Subprograms
13765 @chapter Intrinsic Subprograms
13766 @cindex Intrinsic Subprograms
13769 * Intrinsic Operators::
13770 * Enclosing_Entity::
13771 * Exception_Information::
13772 * Exception_Message::
13776 * Shifts and Rotates::
13777 * Source_Location::
13781 GNAT allows a user application program to write the declaration:
13783 @smallexample @c ada
13784 pragma Import (Intrinsic, name);
13788 providing that the name corresponds to one of the implemented intrinsic
13789 subprograms in GNAT, and that the parameter profile of the referenced
13790 subprogram meets the requirements. This chapter describes the set of
13791 implemented intrinsic subprograms, and the requirements on parameter profiles.
13792 Note that no body is supplied; as with other uses of pragma Import, the
13793 body is supplied elsewhere (in this case by the compiler itself). Note
13794 that any use of this feature is potentially non-portable, since the
13795 Ada standard does not require Ada compilers to implement this feature.
13797 @node Intrinsic Operators
13798 @section Intrinsic Operators
13799 @cindex Intrinsic operator
13802 All the predefined numeric operators in package Standard
13803 in @code{pragma Import (Intrinsic,..)}
13804 declarations. In the binary operator case, the operands must have the same
13805 size. The operand or operands must also be appropriate for
13806 the operator. For example, for addition, the operands must
13807 both be floating-point or both be fixed-point, and the
13808 right operand for @code{"**"} must have a root type of
13809 @code{Standard.Integer'Base}.
13810 You can use an intrinsic operator declaration as in the following example:
13812 @smallexample @c ada
13813 type Int1 is new Integer;
13814 type Int2 is new Integer;
13816 function "+" (X1 : Int1; X2 : Int2) return Int1;
13817 function "+" (X1 : Int1; X2 : Int2) return Int2;
13818 pragma Import (Intrinsic, "+");
13822 This declaration would permit ``mixed mode'' arithmetic on items
13823 of the differing types @code{Int1} and @code{Int2}.
13824 It is also possible to specify such operators for private types, if the
13825 full views are appropriate arithmetic types.
13827 @node Enclosing_Entity
13828 @section Enclosing_Entity
13829 @cindex Enclosing_Entity
13831 This intrinsic subprogram is used in the implementation of the
13832 library routine @code{GNAT.Source_Info}. The only useful use of the
13833 intrinsic import in this case is the one in this unit, so an
13834 application program should simply call the function
13835 @code{GNAT.Source_Info.Enclosing_Entity} to obtain the name of
13836 the current subprogram, package, task, entry, or protected subprogram.
13838 @node Exception_Information
13839 @section Exception_Information
13840 @cindex Exception_Information'
13842 This intrinsic subprogram is used in the implementation of the
13843 library routine @code{GNAT.Current_Exception}. The only useful
13844 use of the intrinsic import in this case is the one in this unit,
13845 so an application program should simply call the function
13846 @code{GNAT.Current_Exception.Exception_Information} to obtain
13847 the exception information associated with the current exception.
13849 @node Exception_Message
13850 @section Exception_Message
13851 @cindex Exception_Message
13853 This intrinsic subprogram is used in the implementation of the
13854 library routine @code{GNAT.Current_Exception}. The only useful
13855 use of the intrinsic import in this case is the one in this unit,
13856 so an application program should simply call the function
13857 @code{GNAT.Current_Exception.Exception_Message} to obtain
13858 the message associated with the current exception.
13860 @node Exception_Name
13861 @section Exception_Name
13862 @cindex Exception_Name
13864 This intrinsic subprogram is used in the implementation of the
13865 library routine @code{GNAT.Current_Exception}. The only useful
13866 use of the intrinsic import in this case is the one in this unit,
13867 so an application program should simply call the function
13868 @code{GNAT.Current_Exception.Exception_Name} to obtain
13869 the name of the current exception.
13875 This intrinsic subprogram is used in the implementation of the
13876 library routine @code{GNAT.Source_Info}. The only useful use of the
13877 intrinsic import in this case is the one in this unit, so an
13878 application program should simply call the function
13879 @code{GNAT.Source_Info.File} to obtain the name of the current
13886 This intrinsic subprogram is used in the implementation of the
13887 library routine @code{GNAT.Source_Info}. The only useful use of the
13888 intrinsic import in this case is the one in this unit, so an
13889 application program should simply call the function
13890 @code{GNAT.Source_Info.Line} to obtain the number of the current
13893 @node Shifts and Rotates
13894 @section Shifts and Rotates
13896 @cindex Shift_Right
13897 @cindex Shift_Right_Arithmetic
13898 @cindex Rotate_Left
13899 @cindex Rotate_Right
13901 In standard Ada, the shift and rotate functions are available only
13902 for the predefined modular types in package @code{Interfaces}. However, in
13903 GNAT it is possible to define these functions for any integer
13904 type (signed or modular), as in this example:
13906 @smallexample @c ada
13907 function Shift_Left
13909 Amount : Natural) return T;
13913 The function name must be one of
13914 Shift_Left, Shift_Right, Shift_Right_Arithmetic, Rotate_Left, or
13915 Rotate_Right. T must be an integer type. T'Size must be
13916 8, 16, 32 or 64 bits; if T is modular, the modulus
13917 must be 2**8, 2**16, 2**32 or 2**64.
13918 The result type must be the same as the type of @code{Value}.
13919 The shift amount must be Natural.
13920 The formal parameter names can be anything.
13922 A more convenient way of providing these shift operators is to use
13923 the Provide_Shift_Operators pragma, which provides the function declarations
13924 and corresponding pragma Import's for all five shift functions.
13926 @node Source_Location
13927 @section Source_Location
13928 @cindex Source_Location
13930 This intrinsic subprogram is used in the implementation of the
13931 library routine @code{GNAT.Source_Info}. The only useful use of the
13932 intrinsic import in this case is the one in this unit, so an
13933 application program should simply call the function
13934 @code{GNAT.Source_Info.Source_Location} to obtain the current
13935 source file location.
13937 @node Representation Clauses and Pragmas
13938 @chapter Representation Clauses and Pragmas
13939 @cindex Representation Clauses
13942 * Alignment Clauses::
13944 * Storage_Size Clauses::
13945 * Size of Variant Record Objects::
13946 * Biased Representation ::
13947 * Value_Size and Object_Size Clauses::
13948 * Component_Size Clauses::
13949 * Bit_Order Clauses::
13950 * Effect of Bit_Order on Byte Ordering::
13951 * Pragma Pack for Arrays::
13952 * Pragma Pack for Records::
13953 * Record Representation Clauses::
13954 * Handling of Records with Holes::
13955 * Enumeration Clauses::
13956 * Address Clauses::
13957 * Effect of Convention on Representation::
13958 * Conventions and Anonymous Access Types::
13959 * Determining the Representations chosen by GNAT::
13963 @cindex Representation Clause
13964 @cindex Representation Pragma
13965 @cindex Pragma, representation
13966 This section describes the representation clauses accepted by GNAT, and
13967 their effect on the representation of corresponding data objects.
13969 GNAT fully implements Annex C (Systems Programming). This means that all
13970 the implementation advice sections in chapter 13 are fully implemented.
13971 However, these sections only require a minimal level of support for
13972 representation clauses. GNAT provides much more extensive capabilities,
13973 and this section describes the additional capabilities provided.
13975 @node Alignment Clauses
13976 @section Alignment Clauses
13977 @cindex Alignment Clause
13980 GNAT requires that all alignment clauses specify a power of 2, and all
13981 default alignments are always a power of 2. The default alignment
13982 values are as follows:
13985 @item @emph{Primitive Types}.
13986 For primitive types, the alignment is the minimum of the actual size of
13987 objects of the type divided by @code{Storage_Unit},
13988 and the maximum alignment supported by the target.
13989 (This maximum alignment is given by the GNAT-specific attribute
13990 @code{Standard'Maximum_Alignment}; see @ref{Attribute Maximum_Alignment}.)
13991 @cindex @code{Maximum_Alignment} attribute
13992 For example, for type @code{Long_Float}, the object size is 8 bytes, and the
13993 default alignment will be 8 on any target that supports alignments
13994 this large, but on some targets, the maximum alignment may be smaller
13995 than 8, in which case objects of type @code{Long_Float} will be maximally
13998 @item @emph{Arrays}.
13999 For arrays, the alignment is equal to the alignment of the component type
14000 for the normal case where no packing or component size is given. If the
14001 array is packed, and the packing is effective (see separate section on
14002 packed arrays), then the alignment will be one for long packed arrays,
14003 or arrays whose length is not known at compile time. For short packed
14004 arrays, which are handled internally as modular types, the alignment
14005 will be as described for primitive types, e.g.@: a packed array of length
14006 31 bits will have an object size of four bytes, and an alignment of 4.
14008 @item @emph{Records}.
14009 For the normal non-packed case, the alignment of a record is equal to
14010 the maximum alignment of any of its components. For tagged records, this
14011 includes the implicit access type used for the tag. If a pragma @code{Pack}
14012 is used and all components are packable (see separate section on pragma
14013 @code{Pack}), then the resulting alignment is 1, unless the layout of the
14014 record makes it profitable to increase it.
14016 A special case is when:
14019 the size of the record is given explicitly, or a
14020 full record representation clause is given, and
14022 the size of the record is 2, 4, or 8 bytes.
14025 In this case, an alignment is chosen to match the
14026 size of the record. For example, if we have:
14028 @smallexample @c ada
14029 type Small is record
14032 for Small'Size use 16;
14036 then the default alignment of the record type @code{Small} is 2, not 1. This
14037 leads to more efficient code when the record is treated as a unit, and also
14038 allows the type to specified as @code{Atomic} on architectures requiring
14044 An alignment clause may specify a larger alignment than the default value
14045 up to some maximum value dependent on the target (obtainable by using the
14046 attribute reference @code{Standard'Maximum_Alignment}). It may also specify
14047 a smaller alignment than the default value for enumeration, integer and
14048 fixed point types, as well as for record types, for example
14050 @smallexample @c ada
14055 for V'alignment use 1;
14059 @cindex Alignment, default
14060 The default alignment for the type @code{V} is 4, as a result of the
14061 Integer field in the record, but it is permissible, as shown, to
14062 override the default alignment of the record with a smaller value.
14064 @cindex Alignment, subtypes
14065 Note that according to the Ada standard, an alignment clause applies only
14066 to the first named subtype. If additional subtypes are declared, then the
14067 compiler is allowed to choose any alignment it likes, and there is no way
14068 to control this choice. Consider:
14070 @smallexample @c ada
14071 type R is range 1 .. 10_000;
14072 for R'Alignment use 1;
14073 subtype RS is R range 1 .. 1000;
14077 The alignment clause specifies an alignment of 1 for the first named subtype
14078 @code{R} but this does not necessarily apply to @code{RS}. When writing
14079 portable Ada code, you should avoid writing code that explicitly or
14080 implicitly relies on the alignment of such subtypes.
14082 For the GNAT compiler, if an explicit alignment clause is given, this
14083 value is also used for any subsequent subtypes. So for GNAT, in the
14084 above example, you can count on the alignment of @code{RS} being 1. But this
14085 assumption is non-portable, and other compilers may choose different
14086 alignments for the subtype @code{RS}.
14089 @section Size Clauses
14090 @cindex Size Clause
14093 The default size for a type @code{T} is obtainable through the
14094 language-defined attribute @code{T'Size} and also through the
14095 equivalent GNAT-defined attribute @code{T'Value_Size}.
14096 For objects of type @code{T}, GNAT will generally increase the type size
14097 so that the object size (obtainable through the GNAT-defined attribute
14098 @code{T'Object_Size})
14099 is a multiple of @code{T'Alignment * Storage_Unit}.
14102 @smallexample @c ada
14103 type Smallint is range 1 .. 6;
14112 In this example, @code{Smallint'Size} = @code{Smallint'Value_Size} = 3,
14113 as specified by the RM rules,
14114 but objects of this type will have a size of 8
14115 (@code{Smallint'Object_Size} = 8),
14116 since objects by default occupy an integral number
14117 of storage units. On some targets, notably older
14118 versions of the Digital Alpha, the size of stand
14119 alone objects of this type may be 32, reflecting
14120 the inability of the hardware to do byte load/stores.
14122 Similarly, the size of type @code{Rec} is 40 bits
14123 (@code{Rec'Size} = @code{Rec'Value_Size} = 40), but
14124 the alignment is 4, so objects of this type will have
14125 their size increased to 64 bits so that it is a multiple
14126 of the alignment (in bits). This decision is
14127 in accordance with the specific Implementation Advice in RM 13.3(43):
14130 A @code{Size} clause should be supported for an object if the specified
14131 @code{Size} is at least as large as its subtype's @code{Size}, and corresponds
14132 to a size in storage elements that is a multiple of the object's
14133 @code{Alignment} (if the @code{Alignment} is nonzero).
14137 An explicit size clause may be used to override the default size by
14138 increasing it. For example, if we have:
14140 @smallexample @c ada
14141 type My_Boolean is new Boolean;
14142 for My_Boolean'Size use 32;
14146 then values of this type will always be 32 bits long. In the case of
14147 discrete types, the size can be increased up to 64 bits, with the effect
14148 that the entire specified field is used to hold the value, sign- or
14149 zero-extended as appropriate. If more than 64 bits is specified, then
14150 padding space is allocated after the value, and a warning is issued that
14151 there are unused bits.
14153 Similarly the size of records and arrays may be increased, and the effect
14154 is to add padding bits after the value. This also causes a warning message
14157 The largest Size value permitted in GNAT is 2**31@minus{}1. Since this is a
14158 Size in bits, this corresponds to an object of size 256 megabytes (minus
14159 one). This limitation is true on all targets. The reason for this
14160 limitation is that it improves the quality of the code in many cases
14161 if it is known that a Size value can be accommodated in an object of
14164 @node Storage_Size Clauses
14165 @section Storage_Size Clauses
14166 @cindex Storage_Size Clause
14169 For tasks, the @code{Storage_Size} clause specifies the amount of space
14170 to be allocated for the task stack. This cannot be extended, and if the
14171 stack is exhausted, then @code{Storage_Error} will be raised (if stack
14172 checking is enabled). Use a @code{Storage_Size} attribute definition clause,
14173 or a @code{Storage_Size} pragma in the task definition to set the
14174 appropriate required size. A useful technique is to include in every
14175 task definition a pragma of the form:
14177 @smallexample @c ada
14178 pragma Storage_Size (Default_Stack_Size);
14182 Then @code{Default_Stack_Size} can be defined in a global package, and
14183 modified as required. Any tasks requiring stack sizes different from the
14184 default can have an appropriate alternative reference in the pragma.
14186 You can also use the @option{-d} binder switch to modify the default stack
14189 For access types, the @code{Storage_Size} clause specifies the maximum
14190 space available for allocation of objects of the type. If this space is
14191 exceeded then @code{Storage_Error} will be raised by an allocation attempt.
14192 In the case where the access type is declared local to a subprogram, the
14193 use of a @code{Storage_Size} clause triggers automatic use of a special
14194 predefined storage pool (@code{System.Pool_Size}) that ensures that all
14195 space for the pool is automatically reclaimed on exit from the scope in
14196 which the type is declared.
14198 A special case recognized by the compiler is the specification of a
14199 @code{Storage_Size} of zero for an access type. This means that no
14200 items can be allocated from the pool, and this is recognized at compile
14201 time, and all the overhead normally associated with maintaining a fixed
14202 size storage pool is eliminated. Consider the following example:
14204 @smallexample @c ada
14206 type R is array (Natural) of Character;
14207 type P is access all R;
14208 for P'Storage_Size use 0;
14209 -- Above access type intended only for interfacing purposes
14213 procedure g (m : P);
14214 pragma Import (C, g);
14225 As indicated in this example, these dummy storage pools are often useful in
14226 connection with interfacing where no object will ever be allocated. If you
14227 compile the above example, you get the warning:
14230 p.adb:16:09: warning: allocation from empty storage pool
14231 p.adb:16:09: warning: Storage_Error will be raised at run time
14235 Of course in practice, there will not be any explicit allocators in the
14236 case of such an access declaration.
14238 @node Size of Variant Record Objects
14239 @section Size of Variant Record Objects
14240 @cindex Size, variant record objects
14241 @cindex Variant record objects, size
14244 In the case of variant record objects, there is a question whether Size gives
14245 information about a particular variant, or the maximum size required
14246 for any variant. Consider the following program
14248 @smallexample @c ada
14249 with Text_IO; use Text_IO;
14251 type R1 (A : Boolean := False) is record
14253 when True => X : Character;
14254 when False => null;
14262 Put_Line (Integer'Image (V1'Size));
14263 Put_Line (Integer'Image (V2'Size));
14268 Here we are dealing with a variant record, where the True variant
14269 requires 16 bits, and the False variant requires 8 bits.
14270 In the above example, both V1 and V2 contain the False variant,
14271 which is only 8 bits long. However, the result of running the
14280 The reason for the difference here is that the discriminant value of
14281 V1 is fixed, and will always be False. It is not possible to assign
14282 a True variant value to V1, therefore 8 bits is sufficient. On the
14283 other hand, in the case of V2, the initial discriminant value is
14284 False (from the default), but it is possible to assign a True
14285 variant value to V2, therefore 16 bits must be allocated for V2
14286 in the general case, even fewer bits may be needed at any particular
14287 point during the program execution.
14289 As can be seen from the output of this program, the @code{'Size}
14290 attribute applied to such an object in GNAT gives the actual allocated
14291 size of the variable, which is the largest size of any of the variants.
14292 The Ada Reference Manual is not completely clear on what choice should
14293 be made here, but the GNAT behavior seems most consistent with the
14294 language in the RM@.
14296 In some cases, it may be desirable to obtain the size of the current
14297 variant, rather than the size of the largest variant. This can be
14298 achieved in GNAT by making use of the fact that in the case of a
14299 subprogram parameter, GNAT does indeed return the size of the current
14300 variant (because a subprogram has no way of knowing how much space
14301 is actually allocated for the actual).
14303 Consider the following modified version of the above program:
14305 @smallexample @c ada
14306 with Text_IO; use Text_IO;
14308 type R1 (A : Boolean := False) is record
14310 when True => X : Character;
14311 when False => null;
14317 function Size (V : R1) return Integer is
14323 Put_Line (Integer'Image (V2'Size));
14324 Put_Line (Integer'IMage (Size (V2)));
14326 Put_Line (Integer'Image (V2'Size));
14327 Put_Line (Integer'IMage (Size (V2)));
14332 The output from this program is
14342 Here we see that while the @code{'Size} attribute always returns
14343 the maximum size, regardless of the current variant value, the
14344 @code{Size} function does indeed return the size of the current
14347 @node Biased Representation
14348 @section Biased Representation
14349 @cindex Size for biased representation
14350 @cindex Biased representation
14353 In the case of scalars with a range starting at other than zero, it is
14354 possible in some cases to specify a size smaller than the default minimum
14355 value, and in such cases, GNAT uses an unsigned biased representation,
14356 in which zero is used to represent the lower bound, and successive values
14357 represent successive values of the type.
14359 For example, suppose we have the declaration:
14361 @smallexample @c ada
14362 type Small is range -7 .. -4;
14363 for Small'Size use 2;
14367 Although the default size of type @code{Small} is 4, the @code{Size}
14368 clause is accepted by GNAT and results in the following representation
14372 -7 is represented as 2#00#
14373 -6 is represented as 2#01#
14374 -5 is represented as 2#10#
14375 -4 is represented as 2#11#
14379 Biased representation is only used if the specified @code{Size} clause
14380 cannot be accepted in any other manner. These reduced sizes that force
14381 biased representation can be used for all discrete types except for
14382 enumeration types for which a representation clause is given.
14384 @node Value_Size and Object_Size Clauses
14385 @section Value_Size and Object_Size Clauses
14387 @findex Object_Size
14388 @cindex Size, of objects
14391 In Ada 95 and Ada 2005, @code{T'Size} for a type @code{T} is the minimum
14392 number of bits required to hold values of type @code{T}.
14393 Although this interpretation was allowed in Ada 83, it was not required,
14394 and this requirement in practice can cause some significant difficulties.
14395 For example, in most Ada 83 compilers, @code{Natural'Size} was 32.
14396 However, in Ada 95 and Ada 2005,
14397 @code{Natural'Size} is
14398 typically 31. This means that code may change in behavior when moving
14399 from Ada 83 to Ada 95 or Ada 2005. For example, consider:
14401 @smallexample @c ada
14402 type Rec is record;
14408 at 0 range 0 .. Natural'Size - 1;
14409 at 0 range Natural'Size .. 2 * Natural'Size - 1;
14414 In the above code, since the typical size of @code{Natural} objects
14415 is 32 bits and @code{Natural'Size} is 31, the above code can cause
14416 unexpected inefficient packing in Ada 95 and Ada 2005, and in general
14417 there are cases where the fact that the object size can exceed the
14418 size of the type causes surprises.
14420 To help get around this problem GNAT provides two implementation
14421 defined attributes, @code{Value_Size} and @code{Object_Size}. When
14422 applied to a type, these attributes yield the size of the type
14423 (corresponding to the RM defined size attribute), and the size of
14424 objects of the type respectively.
14426 The @code{Object_Size} is used for determining the default size of
14427 objects and components. This size value can be referred to using the
14428 @code{Object_Size} attribute. The phrase ``is used'' here means that it is
14429 the basis of the determination of the size. The backend is free to
14430 pad this up if necessary for efficiency, e.g.@: an 8-bit stand-alone
14431 character might be stored in 32 bits on a machine with no efficient
14432 byte access instructions such as the Alpha.
14434 The default rules for the value of @code{Object_Size} for
14435 discrete types are as follows:
14439 The @code{Object_Size} for base subtypes reflect the natural hardware
14440 size in bits (run the compiler with @option{-gnatS} to find those values
14441 for numeric types). Enumeration types and fixed-point base subtypes have
14442 8, 16, 32 or 64 bits for this size, depending on the range of values
14446 The @code{Object_Size} of a subtype is the same as the
14447 @code{Object_Size} of
14448 the type from which it is obtained.
14451 The @code{Object_Size} of a derived base type is copied from the parent
14452 base type, and the @code{Object_Size} of a derived first subtype is copied
14453 from the parent first subtype.
14457 The @code{Value_Size} attribute
14458 is the (minimum) number of bits required to store a value
14460 This value is used to determine how tightly to pack
14461 records or arrays with components of this type, and also affects
14462 the semantics of unchecked conversion (unchecked conversions where
14463 the @code{Value_Size} values differ generate a warning, and are potentially
14466 The default rules for the value of @code{Value_Size} are as follows:
14470 The @code{Value_Size} for a base subtype is the minimum number of bits
14471 required to store all values of the type (including the sign bit
14472 only if negative values are possible).
14475 If a subtype statically matches the first subtype of a given type, then it has
14476 by default the same @code{Value_Size} as the first subtype. This is a
14477 consequence of RM 13.1(14) (``if two subtypes statically match,
14478 then their subtype-specific aspects are the same''.)
14481 All other subtypes have a @code{Value_Size} corresponding to the minimum
14482 number of bits required to store all values of the subtype. For
14483 dynamic bounds, it is assumed that the value can range down or up
14484 to the corresponding bound of the ancestor
14488 The RM defined attribute @code{Size} corresponds to the
14489 @code{Value_Size} attribute.
14491 The @code{Size} attribute may be defined for a first-named subtype. This sets
14492 the @code{Value_Size} of
14493 the first-named subtype to the given value, and the
14494 @code{Object_Size} of this first-named subtype to the given value padded up
14495 to an appropriate boundary. It is a consequence of the default rules
14496 above that this @code{Object_Size} will apply to all further subtypes. On the
14497 other hand, @code{Value_Size} is affected only for the first subtype, any
14498 dynamic subtypes obtained from it directly, and any statically matching
14499 subtypes. The @code{Value_Size} of any other static subtypes is not affected.
14501 @code{Value_Size} and
14502 @code{Object_Size} may be explicitly set for any subtype using
14503 an attribute definition clause. Note that the use of these attributes
14504 can cause the RM 13.1(14) rule to be violated. If two access types
14505 reference aliased objects whose subtypes have differing @code{Object_Size}
14506 values as a result of explicit attribute definition clauses, then it
14507 is illegal to convert from one access subtype to the other. For a more
14508 complete description of this additional legality rule, see the
14509 description of the @code{Object_Size} attribute.
14511 At the implementation level, Esize stores the Object_Size and the
14512 RM_Size field stores the @code{Value_Size} (and hence the value of the
14513 @code{Size} attribute,
14514 which, as noted above, is equivalent to @code{Value_Size}).
14516 To get a feel for the difference, consider the following examples (note
14517 that in each case the base is @code{Short_Short_Integer} with a size of 8):
14520 Object_Size Value_Size
14522 type x1 is range 0 .. 5; 8 3
14524 type x2 is range 0 .. 5;
14525 for x2'size use 12; 16 12
14527 subtype x3 is x2 range 0 .. 3; 16 2
14529 subtype x4 is x2'base range 0 .. 10; 8 4
14531 subtype x5 is x2 range 0 .. dynamic; 16 3*
14533 subtype x6 is x2'base range 0 .. dynamic; 8 3*
14538 Note: the entries marked ``3*'' are not actually specified by the Ada
14539 Reference Manual, but it seems in the spirit of the RM rules to allocate
14540 the minimum number of bits (here 3, given the range for @code{x2})
14541 known to be large enough to hold the given range of values.
14543 So far, so good, but GNAT has to obey the RM rules, so the question is
14544 under what conditions must the RM @code{Size} be used.
14545 The following is a list
14546 of the occasions on which the RM @code{Size} must be used:
14550 Component size for packed arrays or records
14553 Value of the attribute @code{Size} for a type
14556 Warning about sizes not matching for unchecked conversion
14560 For record types, the @code{Object_Size} is always a multiple of the
14561 alignment of the type (this is true for all types). In some cases the
14562 @code{Value_Size} can be smaller. Consider:
14572 On a typical 32-bit architecture, the X component will be four bytes, and
14573 require four-byte alignment, and the Y component will be one byte. In this
14574 case @code{R'Value_Size} will be 40 (bits) since this is the minimum size
14575 required to store a value of this type, and for example, it is permissible
14576 to have a component of type R in an outer array whose component size is
14577 specified to be 48 bits. However, @code{R'Object_Size} will be 64 (bits),
14578 since it must be rounded up so that this value is a multiple of the
14579 alignment (4 bytes = 32 bits).
14582 For all other types, the @code{Object_Size}
14583 and Value_Size are the same (and equivalent to the RM attribute @code{Size}).
14584 Only @code{Size} may be specified for such types.
14586 Note that @code{Value_Size} can be used to force biased representation
14587 for a particular subtype. Consider this example:
14590 type R is (A, B, C, D, E, F);
14591 subtype RAB is R range A .. B;
14592 subtype REF is R range E .. F;
14596 By default, @code{RAB}
14597 has a size of 1 (sufficient to accommodate the representation
14598 of @code{A} and @code{B}, 0 and 1), and @code{REF}
14599 has a size of 3 (sufficient to accommodate the representation
14600 of @code{E} and @code{F}, 4 and 5). But if we add the
14601 following @code{Value_Size} attribute definition clause:
14604 for REF'Value_Size use 1;
14608 then biased representation is forced for @code{REF},
14609 and 0 will represent @code{E} and 1 will represent @code{F}.
14610 A warning is issued when a @code{Value_Size} attribute
14611 definition clause forces biased representation. This
14612 warning can be turned off using @code{-gnatw.B}.
14614 @node Component_Size Clauses
14615 @section Component_Size Clauses
14616 @cindex Component_Size Clause
14619 Normally, the value specified in a component size clause must be consistent
14620 with the subtype of the array component with regard to size and alignment.
14621 In other words, the value specified must be at least equal to the size
14622 of this subtype, and must be a multiple of the alignment value.
14624 In addition, component size clauses are allowed which cause the array
14625 to be packed, by specifying a smaller value. A first case is for
14626 component size values in the range 1 through 63. The value specified
14627 must not be smaller than the Size of the subtype. GNAT will accurately
14628 honor all packing requests in this range. For example, if we have:
14630 @smallexample @c ada
14631 type r is array (1 .. 8) of Natural;
14632 for r'Component_Size use 31;
14636 then the resulting array has a length of 31 bytes (248 bits = 8 * 31).
14637 Of course access to the components of such an array is considerably
14638 less efficient than if the natural component size of 32 is used.
14639 A second case is when the subtype of the component is a record type
14640 padded because of its default alignment. For example, if we have:
14642 @smallexample @c ada
14649 type a is array (1 .. 8) of r;
14650 for a'Component_Size use 72;
14654 then the resulting array has a length of 72 bytes, instead of 96 bytes
14655 if the alignment of the record (4) was obeyed.
14657 Note that there is no point in giving both a component size clause
14658 and a pragma Pack for the same array type. if such duplicate
14659 clauses are given, the pragma Pack will be ignored.
14661 @node Bit_Order Clauses
14662 @section Bit_Order Clauses
14663 @cindex Bit_Order Clause
14664 @cindex bit ordering
14665 @cindex ordering, of bits
14668 For record subtypes, GNAT permits the specification of the @code{Bit_Order}
14669 attribute. The specification may either correspond to the default bit
14670 order for the target, in which case the specification has no effect and
14671 places no additional restrictions, or it may be for the non-standard
14672 setting (that is the opposite of the default).
14674 In the case where the non-standard value is specified, the effect is
14675 to renumber bits within each byte, but the ordering of bytes is not
14676 affected. There are certain
14677 restrictions placed on component clauses as follows:
14681 @item Components fitting within a single storage unit.
14683 These are unrestricted, and the effect is merely to renumber bits. For
14684 example if we are on a little-endian machine with @code{Low_Order_First}
14685 being the default, then the following two declarations have exactly
14688 @smallexample @c ada
14691 B : Integer range 1 .. 120;
14695 A at 0 range 0 .. 0;
14696 B at 0 range 1 .. 7;
14701 B : Integer range 1 .. 120;
14704 for R2'Bit_Order use High_Order_First;
14707 A at 0 range 7 .. 7;
14708 B at 0 range 0 .. 6;
14713 The useful application here is to write the second declaration with the
14714 @code{Bit_Order} attribute definition clause, and know that it will be treated
14715 the same, regardless of whether the target is little-endian or big-endian.
14717 @item Components occupying an integral number of bytes.
14719 These are components that exactly fit in two or more bytes. Such component
14720 declarations are allowed, but have no effect, since it is important to realize
14721 that the @code{Bit_Order} specification does not affect the ordering of bytes.
14722 In particular, the following attempt at getting an endian-independent integer
14725 @smallexample @c ada
14730 for R2'Bit_Order use High_Order_First;
14733 A at 0 range 0 .. 31;
14738 This declaration will result in a little-endian integer on a
14739 little-endian machine, and a big-endian integer on a big-endian machine.
14740 If byte flipping is required for interoperability between big- and
14741 little-endian machines, this must be explicitly programmed. This capability
14742 is not provided by @code{Bit_Order}.
14744 @item Components that are positioned across byte boundaries
14746 but do not occupy an integral number of bytes. Given that bytes are not
14747 reordered, such fields would occupy a non-contiguous sequence of bits
14748 in memory, requiring non-trivial code to reassemble. They are for this
14749 reason not permitted, and any component clause specifying such a layout
14750 will be flagged as illegal by GNAT@.
14755 Since the misconception that Bit_Order automatically deals with all
14756 endian-related incompatibilities is a common one, the specification of
14757 a component field that is an integral number of bytes will always
14758 generate a warning. This warning may be suppressed using @code{pragma
14759 Warnings (Off)} if desired. The following section contains additional
14760 details regarding the issue of byte ordering.
14762 @node Effect of Bit_Order on Byte Ordering
14763 @section Effect of Bit_Order on Byte Ordering
14764 @cindex byte ordering
14765 @cindex ordering, of bytes
14768 In this section we will review the effect of the @code{Bit_Order} attribute
14769 definition clause on byte ordering. Briefly, it has no effect at all, but
14770 a detailed example will be helpful. Before giving this
14771 example, let us review the precise
14772 definition of the effect of defining @code{Bit_Order}. The effect of a
14773 non-standard bit order is described in section 15.5.3 of the Ada
14777 2 A bit ordering is a method of interpreting the meaning of
14778 the storage place attributes.
14782 To understand the precise definition of storage place attributes in
14783 this context, we visit section 13.5.1 of the manual:
14786 13 A record_representation_clause (without the mod_clause)
14787 specifies the layout. The storage place attributes (see 13.5.2)
14788 are taken from the values of the position, first_bit, and last_bit
14789 expressions after normalizing those values so that first_bit is
14790 less than Storage_Unit.
14794 The critical point here is that storage places are taken from
14795 the values after normalization, not before. So the @code{Bit_Order}
14796 interpretation applies to normalized values. The interpretation
14797 is described in the later part of the 15.5.3 paragraph:
14800 2 A bit ordering is a method of interpreting the meaning of
14801 the storage place attributes. High_Order_First (known in the
14802 vernacular as ``big endian'') means that the first bit of a
14803 storage element (bit 0) is the most significant bit (interpreting
14804 the sequence of bits that represent a component as an unsigned
14805 integer value). Low_Order_First (known in the vernacular as
14806 ``little endian'') means the opposite: the first bit is the
14811 Note that the numbering is with respect to the bits of a storage
14812 unit. In other words, the specification affects only the numbering
14813 of bits within a single storage unit.
14815 We can make the effect clearer by giving an example.
14817 Suppose that we have an external device which presents two bytes, the first
14818 byte presented, which is the first (low addressed byte) of the two byte
14819 record is called Master, and the second byte is called Slave.
14821 The left most (most significant bit is called Control for each byte, and
14822 the remaining 7 bits are called V1, V2, @dots{} V7, where V7 is the rightmost
14823 (least significant) bit.
14825 On a big-endian machine, we can write the following representation clause
14827 @smallexample @c ada
14828 type Data is record
14829 Master_Control : Bit;
14837 Slave_Control : Bit;
14847 for Data use record
14848 Master_Control at 0 range 0 .. 0;
14849 Master_V1 at 0 range 1 .. 1;
14850 Master_V2 at 0 range 2 .. 2;
14851 Master_V3 at 0 range 3 .. 3;
14852 Master_V4 at 0 range 4 .. 4;
14853 Master_V5 at 0 range 5 .. 5;
14854 Master_V6 at 0 range 6 .. 6;
14855 Master_V7 at 0 range 7 .. 7;
14856 Slave_Control at 1 range 0 .. 0;
14857 Slave_V1 at 1 range 1 .. 1;
14858 Slave_V2 at 1 range 2 .. 2;
14859 Slave_V3 at 1 range 3 .. 3;
14860 Slave_V4 at 1 range 4 .. 4;
14861 Slave_V5 at 1 range 5 .. 5;
14862 Slave_V6 at 1 range 6 .. 6;
14863 Slave_V7 at 1 range 7 .. 7;
14868 Now if we move this to a little endian machine, then the bit ordering within
14869 the byte is backwards, so we have to rewrite the record rep clause as:
14871 @smallexample @c ada
14872 for Data use record
14873 Master_Control at 0 range 7 .. 7;
14874 Master_V1 at 0 range 6 .. 6;
14875 Master_V2 at 0 range 5 .. 5;
14876 Master_V3 at 0 range 4 .. 4;
14877 Master_V4 at 0 range 3 .. 3;
14878 Master_V5 at 0 range 2 .. 2;
14879 Master_V6 at 0 range 1 .. 1;
14880 Master_V7 at 0 range 0 .. 0;
14881 Slave_Control at 1 range 7 .. 7;
14882 Slave_V1 at 1 range 6 .. 6;
14883 Slave_V2 at 1 range 5 .. 5;
14884 Slave_V3 at 1 range 4 .. 4;
14885 Slave_V4 at 1 range 3 .. 3;
14886 Slave_V5 at 1 range 2 .. 2;
14887 Slave_V6 at 1 range 1 .. 1;
14888 Slave_V7 at 1 range 0 .. 0;
14893 It is a nuisance to have to rewrite the clause, especially if
14894 the code has to be maintained on both machines. However,
14895 this is a case that we can handle with the
14896 @code{Bit_Order} attribute if it is implemented.
14897 Note that the implementation is not required on byte addressed
14898 machines, but it is indeed implemented in GNAT.
14899 This means that we can simply use the
14900 first record clause, together with the declaration
14902 @smallexample @c ada
14903 for Data'Bit_Order use High_Order_First;
14907 and the effect is what is desired, namely the layout is exactly the same,
14908 independent of whether the code is compiled on a big-endian or little-endian
14911 The important point to understand is that byte ordering is not affected.
14912 A @code{Bit_Order} attribute definition never affects which byte a field
14913 ends up in, only where it ends up in that byte.
14914 To make this clear, let us rewrite the record rep clause of the previous
14917 @smallexample @c ada
14918 for Data'Bit_Order use High_Order_First;
14919 for Data use record
14920 Master_Control at 0 range 0 .. 0;
14921 Master_V1 at 0 range 1 .. 1;
14922 Master_V2 at 0 range 2 .. 2;
14923 Master_V3 at 0 range 3 .. 3;
14924 Master_V4 at 0 range 4 .. 4;
14925 Master_V5 at 0 range 5 .. 5;
14926 Master_V6 at 0 range 6 .. 6;
14927 Master_V7 at 0 range 7 .. 7;
14928 Slave_Control at 0 range 8 .. 8;
14929 Slave_V1 at 0 range 9 .. 9;
14930 Slave_V2 at 0 range 10 .. 10;
14931 Slave_V3 at 0 range 11 .. 11;
14932 Slave_V4 at 0 range 12 .. 12;
14933 Slave_V5 at 0 range 13 .. 13;
14934 Slave_V6 at 0 range 14 .. 14;
14935 Slave_V7 at 0 range 15 .. 15;
14940 This is exactly equivalent to saying (a repeat of the first example):
14942 @smallexample @c ada
14943 for Data'Bit_Order use High_Order_First;
14944 for Data use record
14945 Master_Control at 0 range 0 .. 0;
14946 Master_V1 at 0 range 1 .. 1;
14947 Master_V2 at 0 range 2 .. 2;
14948 Master_V3 at 0 range 3 .. 3;
14949 Master_V4 at 0 range 4 .. 4;
14950 Master_V5 at 0 range 5 .. 5;
14951 Master_V6 at 0 range 6 .. 6;
14952 Master_V7 at 0 range 7 .. 7;
14953 Slave_Control at 1 range 0 .. 0;
14954 Slave_V1 at 1 range 1 .. 1;
14955 Slave_V2 at 1 range 2 .. 2;
14956 Slave_V3 at 1 range 3 .. 3;
14957 Slave_V4 at 1 range 4 .. 4;
14958 Slave_V5 at 1 range 5 .. 5;
14959 Slave_V6 at 1 range 6 .. 6;
14960 Slave_V7 at 1 range 7 .. 7;
14965 Why are they equivalent? Well take a specific field, the @code{Slave_V2}
14966 field. The storage place attributes are obtained by normalizing the
14967 values given so that the @code{First_Bit} value is less than 8. After
14968 normalizing the values (0,10,10) we get (1,2,2) which is exactly what
14969 we specified in the other case.
14971 Now one might expect that the @code{Bit_Order} attribute might affect
14972 bit numbering within the entire record component (two bytes in this
14973 case, thus affecting which byte fields end up in), but that is not
14974 the way this feature is defined, it only affects numbering of bits,
14975 not which byte they end up in.
14977 Consequently it never makes sense to specify a starting bit number
14978 greater than 7 (for a byte addressable field) if an attribute
14979 definition for @code{Bit_Order} has been given, and indeed it
14980 may be actively confusing to specify such a value, so the compiler
14981 generates a warning for such usage.
14983 If you do need to control byte ordering then appropriate conditional
14984 values must be used. If in our example, the slave byte came first on
14985 some machines we might write:
14987 @smallexample @c ada
14988 Master_Byte_First constant Boolean := @dots{};
14990 Master_Byte : constant Natural :=
14991 1 - Boolean'Pos (Master_Byte_First);
14992 Slave_Byte : constant Natural :=
14993 Boolean'Pos (Master_Byte_First);
14995 for Data'Bit_Order use High_Order_First;
14996 for Data use record
14997 Master_Control at Master_Byte range 0 .. 0;
14998 Master_V1 at Master_Byte range 1 .. 1;
14999 Master_V2 at Master_Byte range 2 .. 2;
15000 Master_V3 at Master_Byte range 3 .. 3;
15001 Master_V4 at Master_Byte range 4 .. 4;
15002 Master_V5 at Master_Byte range 5 .. 5;
15003 Master_V6 at Master_Byte range 6 .. 6;
15004 Master_V7 at Master_Byte range 7 .. 7;
15005 Slave_Control at Slave_Byte range 0 .. 0;
15006 Slave_V1 at Slave_Byte range 1 .. 1;
15007 Slave_V2 at Slave_Byte range 2 .. 2;
15008 Slave_V3 at Slave_Byte range 3 .. 3;
15009 Slave_V4 at Slave_Byte range 4 .. 4;
15010 Slave_V5 at Slave_Byte range 5 .. 5;
15011 Slave_V6 at Slave_Byte range 6 .. 6;
15012 Slave_V7 at Slave_Byte range 7 .. 7;
15017 Now to switch between machines, all that is necessary is
15018 to set the boolean constant @code{Master_Byte_First} in
15019 an appropriate manner.
15021 @node Pragma Pack for Arrays
15022 @section Pragma Pack for Arrays
15023 @cindex Pragma Pack (for arrays)
15026 Pragma @code{Pack} applied to an array has no effect unless the component type
15027 is packable. For a component type to be packable, it must be one of the
15034 Any type whose size is specified with a size clause
15036 Any packed array type with a static size
15038 Any record type padded because of its default alignment
15042 For all these cases, if the component subtype size is in the range
15043 1 through 63, then the effect of the pragma @code{Pack} is exactly as though a
15044 component size were specified giving the component subtype size.
15045 For example if we have:
15047 @smallexample @c ada
15048 type r is range 0 .. 17;
15050 type ar is array (1 .. 8) of r;
15055 Then the component size of @code{ar} will be set to 5 (i.e.@: to @code{r'size},
15056 and the size of the array @code{ar} will be exactly 40 bits.
15058 Note that in some cases this rather fierce approach to packing can produce
15059 unexpected effects. For example, in Ada 95 and Ada 2005,
15060 subtype @code{Natural} typically has a size of 31, meaning that if you
15061 pack an array of @code{Natural}, you get 31-bit
15062 close packing, which saves a few bits, but results in far less efficient
15063 access. Since many other Ada compilers will ignore such a packing request,
15064 GNAT will generate a warning on some uses of pragma @code{Pack} that it guesses
15065 might not be what is intended. You can easily remove this warning by
15066 using an explicit @code{Component_Size} setting instead, which never generates
15067 a warning, since the intention of the programmer is clear in this case.
15069 GNAT treats packed arrays in one of two ways. If the size of the array is
15070 known at compile time and is less than 64 bits, then internally the array
15071 is represented as a single modular type, of exactly the appropriate number
15072 of bits. If the length is greater than 63 bits, or is not known at compile
15073 time, then the packed array is represented as an array of bytes, and the
15074 length is always a multiple of 8 bits.
15076 Note that to represent a packed array as a modular type, the alignment must
15077 be suitable for the modular type involved. For example, on typical machines
15078 a 32-bit packed array will be represented by a 32-bit modular integer with
15079 an alignment of four bytes. If you explicitly override the default alignment
15080 with an alignment clause that is too small, the modular representation
15081 cannot be used. For example, consider the following set of declarations:
15083 @smallexample @c ada
15084 type R is range 1 .. 3;
15085 type S is array (1 .. 31) of R;
15086 for S'Component_Size use 2;
15088 for S'Alignment use 1;
15092 If the alignment clause were not present, then a 62-bit modular
15093 representation would be chosen (typically with an alignment of 4 or 8
15094 bytes depending on the target). But the default alignment is overridden
15095 with the explicit alignment clause. This means that the modular
15096 representation cannot be used, and instead the array of bytes
15097 representation must be used, meaning that the length must be a multiple
15098 of 8. Thus the above set of declarations will result in a diagnostic
15099 rejecting the size clause and noting that the minimum size allowed is 64.
15101 @cindex Pragma Pack (for type Natural)
15102 @cindex Pragma Pack warning
15104 One special case that is worth noting occurs when the base type of the
15105 component size is 8/16/32 and the subtype is one bit less. Notably this
15106 occurs with subtype @code{Natural}. Consider:
15108 @smallexample @c ada
15109 type Arr is array (1 .. 32) of Natural;
15114 In all commonly used Ada 83 compilers, this pragma Pack would be ignored,
15115 since typically @code{Natural'Size} is 32 in Ada 83, and in any case most
15116 Ada 83 compilers did not attempt 31 bit packing.
15118 In Ada 95 and Ada 2005, @code{Natural'Size} is required to be 31. Furthermore,
15119 GNAT really does pack 31-bit subtype to 31 bits. This may result in a
15120 substantial unintended performance penalty when porting legacy Ada 83 code.
15121 To help prevent this, GNAT generates a warning in such cases. If you really
15122 want 31 bit packing in a case like this, you can set the component size
15125 @smallexample @c ada
15126 type Arr is array (1 .. 32) of Natural;
15127 for Arr'Component_Size use 31;
15131 Here 31-bit packing is achieved as required, and no warning is generated,
15132 since in this case the programmer intention is clear.
15134 @node Pragma Pack for Records
15135 @section Pragma Pack for Records
15136 @cindex Pragma Pack (for records)
15139 Pragma @code{Pack} applied to a record will pack the components to reduce
15140 wasted space from alignment gaps and by reducing the amount of space
15141 taken by components. We distinguish between @emph{packable} components and
15142 @emph{non-packable} components.
15143 Components of the following types are considered packable:
15146 All primitive types are packable.
15149 Small packed arrays, whose size does not exceed 64 bits, and where the
15150 size is statically known at compile time, are represented internally
15151 as modular integers, and so they are also packable.
15156 All packable components occupy the exact number of bits corresponding to
15157 their @code{Size} value, and are packed with no padding bits, i.e.@: they
15158 can start on an arbitrary bit boundary.
15160 All other types are non-packable, they occupy an integral number of
15162 are placed at a boundary corresponding to their alignment requirements.
15164 For example, consider the record
15166 @smallexample @c ada
15167 type Rb1 is array (1 .. 13) of Boolean;
15170 type Rb2 is array (1 .. 65) of Boolean;
15185 The representation for the record x2 is as follows:
15187 @smallexample @c ada
15188 for x2'Size use 224;
15190 l1 at 0 range 0 .. 0;
15191 l2 at 0 range 1 .. 64;
15192 l3 at 12 range 0 .. 31;
15193 l4 at 16 range 0 .. 0;
15194 l5 at 16 range 1 .. 13;
15195 l6 at 18 range 0 .. 71;
15200 Studying this example, we see that the packable fields @code{l1}
15202 of length equal to their sizes, and placed at specific bit boundaries (and
15203 not byte boundaries) to
15204 eliminate padding. But @code{l3} is of a non-packable float type, so
15205 it is on the next appropriate alignment boundary.
15207 The next two fields are fully packable, so @code{l4} and @code{l5} are
15208 minimally packed with no gaps. However, type @code{Rb2} is a packed
15209 array that is longer than 64 bits, so it is itself non-packable. Thus
15210 the @code{l6} field is aligned to the next byte boundary, and takes an
15211 integral number of bytes, i.e.@: 72 bits.
15213 @node Record Representation Clauses
15214 @section Record Representation Clauses
15215 @cindex Record Representation Clause
15218 Record representation clauses may be given for all record types, including
15219 types obtained by record extension. Component clauses are allowed for any
15220 static component. The restrictions on component clauses depend on the type
15223 @cindex Component Clause
15224 For all components of an elementary type, the only restriction on component
15225 clauses is that the size must be at least the 'Size value of the type
15226 (actually the Value_Size). There are no restrictions due to alignment,
15227 and such components may freely cross storage boundaries.
15229 Packed arrays with a size up to and including 64 bits are represented
15230 internally using a modular type with the appropriate number of bits, and
15231 thus the same lack of restriction applies. For example, if you declare:
15233 @smallexample @c ada
15234 type R is array (1 .. 49) of Boolean;
15240 then a component clause for a component of type R may start on any
15241 specified bit boundary, and may specify a value of 49 bits or greater.
15243 For packed bit arrays that are longer than 64 bits, there are two
15244 cases. If the component size is a power of 2 (1,2,4,8,16,32 bits),
15245 including the important case of single bits or boolean values, then
15246 there are no limitations on placement of such components, and they
15247 may start and end at arbitrary bit boundaries.
15249 If the component size is not a power of 2 (e.g.@: 3 or 5), then
15250 an array of this type longer than 64 bits must always be placed on
15251 on a storage unit (byte) boundary and occupy an integral number
15252 of storage units (bytes). Any component clause that does not
15253 meet this requirement will be rejected.
15255 Any aliased component, or component of an aliased type, must
15256 have its normal alignment and size. A component clause that
15257 does not meet this requirement will be rejected.
15259 The tag field of a tagged type always occupies an address sized field at
15260 the start of the record. No component clause may attempt to overlay this
15261 tag. When a tagged type appears as a component, the tag field must have
15264 In the case of a record extension T1, of a type T, no component clause applied
15265 to the type T1 can specify a storage location that would overlap the first
15266 T'Size bytes of the record.
15268 For all other component types, including non-bit-packed arrays,
15269 the component can be placed at an arbitrary bit boundary,
15270 so for example, the following is permitted:
15272 @smallexample @c ada
15273 type R is array (1 .. 10) of Boolean;
15282 G at 0 range 0 .. 0;
15283 H at 0 range 1 .. 1;
15284 L at 0 range 2 .. 81;
15285 R at 0 range 82 .. 161;
15290 Note: the above rules apply to recent releases of GNAT 5.
15291 In GNAT 3, there are more severe restrictions on larger components.
15292 For non-primitive types, including packed arrays with a size greater than
15293 64 bits, component clauses must respect the alignment requirement of the
15294 type, in particular, always starting on a byte boundary, and the length
15295 must be a multiple of the storage unit.
15297 @node Handling of Records with Holes
15298 @section Handling of Records with Holes
15299 @cindex Handling of Records with Holes
15301 As a result of alignment considerations, records may contain "holes"
15303 which do not correspond to the data bits of any of the components.
15304 Record representation clauses can also result in holes in records.
15306 GNAT does not attempt to clear these holes, so in record objects,
15307 they should be considered to hold undefined rubbish. The generated
15308 equality routine just tests components so does not access these
15309 undefined bits, and assignment and copy operations may or may not
15310 preserve the contents of these holes (for assignments, the holes
15311 in the target will in practice contain either the bits that are
15312 present in the holes in the source, or the bits that were present
15313 in the target before the assignment).
15315 If it is necessary to ensure that holes in records have all zero
15316 bits, then record objects for which this initialization is desired
15317 should be explicitly set to all zero values using Unchecked_Conversion
15318 or address overlays. For example
15320 @smallexample @c ada
15321 type HRec is record
15328 On typical machines, integers need to be aligned on a four-byte
15329 boundary, resulting in three bytes of undefined rubbish following
15330 the 8-bit field for C. To ensure that the hole in a variable of
15331 type HRec is set to all zero bits,
15332 you could for example do:
15334 @smallexample @c ada
15335 type Base is record
15336 Dummy1, Dummy2 : Integer := 0;
15341 for RealVar'Address use BaseVar'Address;
15345 Now the 8-bytes of the value of RealVar start out containing all zero
15346 bits. A safer approach is to just define dummy fields, avoiding the
15349 @smallexample @c ada
15350 type HRec is record
15352 Dummy1 : Short_Short_Integer := 0;
15353 Dummy2 : Short_Short_Integer := 0;
15354 Dummy3 : Short_Short_Integer := 0;
15360 And to make absolutely sure that the intent of this is followed, you
15361 can use representation clauses:
15363 @smallexample @c ada
15364 for Hrec use record
15365 C at 0 range 0 .. 7;
15366 Dummy1 at 1 range 0 .. 7;
15367 Dummy2 at 2 range 0 .. 7;
15368 Dummy3 at 3 range 0 .. 7;
15369 I at 4 range 0 .. 31;
15371 for Hrec'Size use 64;
15374 @node Enumeration Clauses
15375 @section Enumeration Clauses
15377 The only restriction on enumeration clauses is that the range of values
15378 must be representable. For the signed case, if one or more of the
15379 representation values are negative, all values must be in the range:
15381 @smallexample @c ada
15382 System.Min_Int .. System.Max_Int
15386 For the unsigned case, where all values are nonnegative, the values must
15389 @smallexample @c ada
15390 0 .. System.Max_Binary_Modulus;
15394 A @emph{confirming} representation clause is one in which the values range
15395 from 0 in sequence, i.e.@: a clause that confirms the default representation
15396 for an enumeration type.
15397 Such a confirming representation
15398 is permitted by these rules, and is specially recognized by the compiler so
15399 that no extra overhead results from the use of such a clause.
15401 If an array has an index type which is an enumeration type to which an
15402 enumeration clause has been applied, then the array is stored in a compact
15403 manner. Consider the declarations:
15405 @smallexample @c ada
15406 type r is (A, B, C);
15407 for r use (A => 1, B => 5, C => 10);
15408 type t is array (r) of Character;
15412 The array type t corresponds to a vector with exactly three elements and
15413 has a default size equal to @code{3*Character'Size}. This ensures efficient
15414 use of space, but means that accesses to elements of the array will incur
15415 the overhead of converting representation values to the corresponding
15416 positional values, (i.e.@: the value delivered by the @code{Pos} attribute).
15418 @node Address Clauses
15419 @section Address Clauses
15420 @cindex Address Clause
15422 The reference manual allows a general restriction on representation clauses,
15423 as found in RM 13.1(22):
15426 An implementation need not support representation
15427 items containing nonstatic expressions, except that
15428 an implementation should support a representation item
15429 for a given entity if each nonstatic expression in the
15430 representation item is a name that statically denotes
15431 a constant declared before the entity.
15435 In practice this is applicable only to address clauses, since this is the
15436 only case in which a non-static expression is permitted by the syntax. As
15437 the AARM notes in sections 13.1 (22.a-22.h):
15440 22.a Reason: This is to avoid the following sort of thing:
15442 22.b X : Integer := F(@dots{});
15443 Y : Address := G(@dots{});
15444 for X'Address use Y;
15446 22.c In the above, we have to evaluate the
15447 initialization expression for X before we
15448 know where to put the result. This seems
15449 like an unreasonable implementation burden.
15451 22.d The above code should instead be written
15454 22.e Y : constant Address := G(@dots{});
15455 X : Integer := F(@dots{});
15456 for X'Address use Y;
15458 22.f This allows the expression ``Y'' to be safely
15459 evaluated before X is created.
15461 22.g The constant could be a formal parameter of mode in.
15463 22.h An implementation can support other nonstatic
15464 expressions if it wants to. Expressions of type
15465 Address are hardly ever static, but their value
15466 might be known at compile time anyway in many
15471 GNAT does indeed permit many additional cases of non-static expressions. In
15472 particular, if the type involved is elementary there are no restrictions
15473 (since in this case, holding a temporary copy of the initialization value,
15474 if one is present, is inexpensive). In addition, if there is no implicit or
15475 explicit initialization, then there are no restrictions. GNAT will reject
15476 only the case where all three of these conditions hold:
15481 The type of the item is non-elementary (e.g.@: a record or array).
15484 There is explicit or implicit initialization required for the object.
15485 Note that access values are always implicitly initialized.
15488 The address value is non-static. Here GNAT is more permissive than the
15489 RM, and allows the address value to be the address of a previously declared
15490 stand-alone variable, as long as it does not itself have an address clause.
15492 @smallexample @c ada
15493 Anchor : Some_Initialized_Type;
15494 Overlay : Some_Initialized_Type;
15495 for Overlay'Address use Anchor'Address;
15499 However, the prefix of the address clause cannot be an array component, or
15500 a component of a discriminated record.
15505 As noted above in section 22.h, address values are typically non-static. In
15506 particular the To_Address function, even if applied to a literal value, is
15507 a non-static function call. To avoid this minor annoyance, GNAT provides
15508 the implementation defined attribute 'To_Address. The following two
15509 expressions have identical values:
15513 @smallexample @c ada
15514 To_Address (16#1234_0000#)
15515 System'To_Address (16#1234_0000#);
15519 except that the second form is considered to be a static expression, and
15520 thus when used as an address clause value is always permitted.
15523 Additionally, GNAT treats as static an address clause that is an
15524 unchecked_conversion of a static integer value. This simplifies the porting
15525 of legacy code, and provides a portable equivalent to the GNAT attribute
15528 Another issue with address clauses is the interaction with alignment
15529 requirements. When an address clause is given for an object, the address
15530 value must be consistent with the alignment of the object (which is usually
15531 the same as the alignment of the type of the object). If an address clause
15532 is given that specifies an inappropriately aligned address value, then the
15533 program execution is erroneous.
15535 Since this source of erroneous behavior can have unfortunate effects, GNAT
15536 checks (at compile time if possible, generating a warning, or at execution
15537 time with a run-time check) that the alignment is appropriate. If the
15538 run-time check fails, then @code{Program_Error} is raised. This run-time
15539 check is suppressed if range checks are suppressed, or if the special GNAT
15540 check Alignment_Check is suppressed, or if
15541 @code{pragma Restrictions (No_Elaboration_Code)} is in effect.
15543 Finally, GNAT does not permit overlaying of objects of controlled types or
15544 composite types containing a controlled component. In most cases, the compiler
15545 can detect an attempt at such overlays and will generate a warning at compile
15546 time and a Program_Error exception at run time.
15549 An address clause cannot be given for an exported object. More
15550 understandably the real restriction is that objects with an address
15551 clause cannot be exported. This is because such variables are not
15552 defined by the Ada program, so there is no external object to export.
15555 It is permissible to give an address clause and a pragma Import for the
15556 same object. In this case, the variable is not really defined by the
15557 Ada program, so there is no external symbol to be linked. The link name
15558 and the external name are ignored in this case. The reason that we allow this
15559 combination is that it provides a useful idiom to avoid unwanted
15560 initializations on objects with address clauses.
15562 When an address clause is given for an object that has implicit or
15563 explicit initialization, then by default initialization takes place. This
15564 means that the effect of the object declaration is to overwrite the
15565 memory at the specified address. This is almost always not what the
15566 programmer wants, so GNAT will output a warning:
15576 for Ext'Address use System'To_Address (16#1234_1234#);
15578 >>> warning: implicit initialization of "Ext" may
15579 modify overlaid storage
15580 >>> warning: use pragma Import for "Ext" to suppress
15581 initialization (RM B(24))
15587 As indicated by the warning message, the solution is to use a (dummy) pragma
15588 Import to suppress this initialization. The pragma tell the compiler that the
15589 object is declared and initialized elsewhere. The following package compiles
15590 without warnings (and the initialization is suppressed):
15592 @smallexample @c ada
15600 for Ext'Address use System'To_Address (16#1234_1234#);
15601 pragma Import (Ada, Ext);
15606 A final issue with address clauses involves their use for overlaying
15607 variables, as in the following example:
15608 @cindex Overlaying of objects
15610 @smallexample @c ada
15613 for B'Address use A'Address;
15617 or alternatively, using the form recommended by the RM:
15619 @smallexample @c ada
15621 Addr : constant Address := A'Address;
15623 for B'Address use Addr;
15627 In both of these cases, @code{A}
15628 and @code{B} become aliased to one another via the
15629 address clause. This use of address clauses to overlay
15630 variables, achieving an effect similar to unchecked
15631 conversion was erroneous in Ada 83, but in Ada 95 and Ada 2005
15632 the effect is implementation defined. Furthermore, the
15633 Ada RM specifically recommends that in a situation
15634 like this, @code{B} should be subject to the following
15635 implementation advice (RM 13.3(19)):
15638 19 If the Address of an object is specified, or it is imported
15639 or exported, then the implementation should not perform
15640 optimizations based on assumptions of no aliases.
15644 GNAT follows this recommendation, and goes further by also applying
15645 this recommendation to the overlaid variable (@code{A}
15646 in the above example) in this case. This means that the overlay
15647 works "as expected", in that a modification to one of the variables
15648 will affect the value of the other.
15650 Note that when address clause overlays are used in this way, there is an
15651 issue of unintentional initialization, as shown by this example:
15653 @smallexample @c ada
15654 package Overwrite_Record is
15656 A : Character := 'C';
15657 B : Character := 'A';
15659 X : Short_Integer := 3;
15661 for Y'Address use X'Address;
15663 >>> warning: default initialization of "Y" may
15664 modify "X", use pragma Import for "Y" to
15665 suppress initialization (RM B.1(24))
15667 end Overwrite_Record;
15671 Here the default initialization of @code{Y} will clobber the value
15672 of @code{X}, which justifies the warning. The warning notes that
15673 this effect can be eliminated by adding a @code{pragma Import}
15674 which suppresses the initialization:
15676 @smallexample @c ada
15677 package Overwrite_Record is
15679 A : Character := 'C';
15680 B : Character := 'A';
15682 X : Short_Integer := 3;
15684 for Y'Address use X'Address;
15685 pragma Import (Ada, Y);
15686 end Overwrite_Record;
15690 Note that the use of @code{pragma Initialize_Scalars} may cause variables to
15691 be initialized when they would not otherwise have been in the absence
15692 of the use of this pragma. This may cause an overlay to have this
15693 unintended clobbering effect. The compiler avoids this for scalar
15694 types, but not for composite objects (where in general the effect
15695 of @code{Initialize_Scalars} is part of the initialization routine
15696 for the composite object:
15698 @smallexample @c ada
15699 pragma Initialize_Scalars;
15700 with Ada.Text_IO; use Ada.Text_IO;
15701 procedure Overwrite_Array is
15702 type Arr is array (1 .. 5) of Integer;
15703 X : Arr := (others => 1);
15705 for A'Address use X'Address;
15707 >>> warning: default initialization of "A" may
15708 modify "X", use pragma Import for "A" to
15709 suppress initialization (RM B.1(24))
15712 if X /= Arr'(others => 1) then
15713 Put_Line ("X was clobbered");
15715 Put_Line ("X was not clobbered");
15717 end Overwrite_Array;
15721 The above program generates the warning as shown, and at execution
15722 time, prints @code{X was clobbered}. If the @code{pragma Import} is
15723 added as suggested:
15725 @smallexample @c ada
15726 pragma Initialize_Scalars;
15727 with Ada.Text_IO; use Ada.Text_IO;
15728 procedure Overwrite_Array is
15729 type Arr is array (1 .. 5) of Integer;
15730 X : Arr := (others => 1);
15732 for A'Address use X'Address;
15733 pragma Import (Ada, A);
15735 if X /= Arr'(others => 1) then
15736 Put_Line ("X was clobbered");
15738 Put_Line ("X was not clobbered");
15740 end Overwrite_Array;
15744 then the program compiles without the waraning and when run will generate
15745 the output @code{X was not clobbered}.
15747 @node Effect of Convention on Representation
15748 @section Effect of Convention on Representation
15749 @cindex Convention, effect on representation
15752 Normally the specification of a foreign language convention for a type or
15753 an object has no effect on the chosen representation. In particular, the
15754 representation chosen for data in GNAT generally meets the standard system
15755 conventions, and for example records are laid out in a manner that is
15756 consistent with C@. This means that specifying convention C (for example)
15759 There are four exceptions to this general rule:
15763 @item Convention Fortran and array subtypes
15764 If pragma Convention Fortran is specified for an array subtype, then in
15765 accordance with the implementation advice in section 3.6.2(11) of the
15766 Ada Reference Manual, the array will be stored in a Fortran-compatible
15767 column-major manner, instead of the normal default row-major order.
15769 @item Convention C and enumeration types
15770 GNAT normally stores enumeration types in 8, 16, or 32 bits as required
15771 to accommodate all values of the type. For example, for the enumeration
15774 @smallexample @c ada
15775 type Color is (Red, Green, Blue);
15779 8 bits is sufficient to store all values of the type, so by default, objects
15780 of type @code{Color} will be represented using 8 bits. However, normal C
15781 convention is to use 32 bits for all enum values in C, since enum values
15782 are essentially of type int. If pragma @code{Convention C} is specified for an
15783 Ada enumeration type, then the size is modified as necessary (usually to
15784 32 bits) to be consistent with the C convention for enum values.
15786 Note that this treatment applies only to types. If Convention C is given for
15787 an enumeration object, where the enumeration type is not Convention C, then
15788 Object_Size bits are allocated. For example, for a normal enumeration type,
15789 with less than 256 elements, only 8 bits will be allocated for the object.
15790 Since this may be a surprise in terms of what C expects, GNAT will issue a
15791 warning in this situation. The warning can be suppressed by giving an explicit
15792 size clause specifying the desired size.
15794 @item Convention C/Fortran and Boolean types
15795 In C, the usual convention for boolean values, that is values used for
15796 conditions, is that zero represents false, and nonzero values represent
15797 true. In Ada, the normal convention is that two specific values, typically
15798 0/1, are used to represent false/true respectively.
15800 Fortran has a similar convention for @code{LOGICAL} values (any nonzero
15801 value represents true).
15803 To accommodate the Fortran and C conventions, if a pragma Convention specifies
15804 C or Fortran convention for a derived Boolean, as in the following example:
15806 @smallexample @c ada
15807 type C_Switch is new Boolean;
15808 pragma Convention (C, C_Switch);
15812 then the GNAT generated code will treat any nonzero value as true. For truth
15813 values generated by GNAT, the conventional value 1 will be used for True, but
15814 when one of these values is read, any nonzero value is treated as True.
15816 @item Access types on OpenVMS
15817 For 64-bit OpenVMS systems, access types (other than those for unconstrained
15818 arrays) are 64-bits long. An exception to this rule is for the case of
15819 C-convention access types where there is no explicit size clause present (or
15820 inherited for derived types). In this case, GNAT chooses to make these
15821 pointers 32-bits, which provides an easier path for migration of 32-bit legacy
15822 code. size clause specifying 64-bits must be used to obtain a 64-bit pointer.
15826 @node Conventions and Anonymous Access Types
15827 @section Conventions and Anonymous Access Types
15828 @cindex Anonymous access types
15829 @cindex Convention for anonymous access types
15831 The RM is not entirely clear on convention handling in a number of cases,
15832 and in particular, it is not clear on the convention to be given to
15833 anonymous access types in general, and in particular what is to be
15834 done for the case of anonymous access-to-subprogram.
15836 In GNAT, we decide that if an explicit Convention is applied
15837 to an object or component, and its type is such an anonymous type,
15838 then the convention will apply to this anonymous type as well. This
15839 seems to make sense since it is anomolous in any case to have a
15840 different convention for an object and its type, and there is clearly
15841 no way to explicitly specify a convention for an anonymous type, since
15842 it doesn't have a name to specify!
15844 Furthermore, we decide that if a convention is applied to a record type,
15845 then this convention is inherited by any of its components that are of an
15846 anonymous access type which do not have an explicitly specified convention.
15848 The following program shows these conventions in action:
15850 @smallexample @c ada
15851 package ConvComp is
15852 type Foo is range 1 .. 10;
15854 A : access function (X : Foo) return Integer;
15857 pragma Convention (C, T1);
15860 A : access function (X : Foo) return Integer;
15861 pragma Convention (C, A);
15864 pragma Convention (COBOL, T2);
15867 A : access function (X : Foo) return Integer;
15868 pragma Convention (COBOL, A);
15871 pragma Convention (C, T3);
15874 A : access function (X : Foo) return Integer;
15877 pragma Convention (COBOL, T4);
15879 function F (X : Foo) return Integer;
15880 pragma Convention (C, F);
15882 function F (X : Foo) return Integer is (13);
15884 TV1 : T1 := (F'Access, 12); -- OK
15885 TV2 : T2 := (F'Access, 13); -- OK
15887 TV3 : T3 := (F'Access, 13); -- ERROR
15889 >>> subprogram "F" has wrong convention
15890 >>> does not match access to subprogram declared at line 17
15891 38. TV4 : T4 := (F'Access, 13); -- ERROR
15893 >>> subprogram "F" has wrong convention
15894 >>> does not match access to subprogram declared at line 24
15898 @node Determining the Representations chosen by GNAT
15899 @section Determining the Representations chosen by GNAT
15900 @cindex Representation, determination of
15901 @cindex @option{-gnatR} switch
15904 Although the descriptions in this section are intended to be complete, it is
15905 often easier to simply experiment to see what GNAT accepts and what the
15906 effect is on the layout of types and objects.
15908 As required by the Ada RM, if a representation clause is not accepted, then
15909 it must be rejected as illegal by the compiler. However, when a
15910 representation clause or pragma is accepted, there can still be questions
15911 of what the compiler actually does. For example, if a partial record
15912 representation clause specifies the location of some components and not
15913 others, then where are the non-specified components placed? Or if pragma
15914 @code{Pack} is used on a record, then exactly where are the resulting
15915 fields placed? The section on pragma @code{Pack} in this chapter can be
15916 used to answer the second question, but it is often easier to just see
15917 what the compiler does.
15919 For this purpose, GNAT provides the option @option{-gnatR}. If you compile
15920 with this option, then the compiler will output information on the actual
15921 representations chosen, in a format similar to source representation
15922 clauses. For example, if we compile the package:
15924 @smallexample @c ada
15926 type r (x : boolean) is tagged record
15928 when True => S : String (1 .. 100);
15929 when False => null;
15933 type r2 is new r (false) with record
15938 y2 at 16 range 0 .. 31;
15945 type x1 is array (1 .. 10) of x;
15946 for x1'component_size use 11;
15948 type ia is access integer;
15950 type Rb1 is array (1 .. 13) of Boolean;
15953 type Rb2 is array (1 .. 65) of Boolean;
15969 using the switch @option{-gnatR} we obtain the following output:
15972 Representation information for unit q
15973 -------------------------------------
15976 for r'Alignment use 4;
15978 x at 4 range 0 .. 7;
15979 _tag at 0 range 0 .. 31;
15980 s at 5 range 0 .. 799;
15983 for r2'Size use 160;
15984 for r2'Alignment use 4;
15986 x at 4 range 0 .. 7;
15987 _tag at 0 range 0 .. 31;
15988 _parent at 0 range 0 .. 63;
15989 y2 at 16 range 0 .. 31;
15993 for x'Alignment use 1;
15995 y at 0 range 0 .. 7;
15998 for x1'Size use 112;
15999 for x1'Alignment use 1;
16000 for x1'Component_Size use 11;
16002 for rb1'Size use 13;
16003 for rb1'Alignment use 2;
16004 for rb1'Component_Size use 1;
16006 for rb2'Size use 72;
16007 for rb2'Alignment use 1;
16008 for rb2'Component_Size use 1;
16010 for x2'Size use 224;
16011 for x2'Alignment use 4;
16013 l1 at 0 range 0 .. 0;
16014 l2 at 0 range 1 .. 64;
16015 l3 at 12 range 0 .. 31;
16016 l4 at 16 range 0 .. 0;
16017 l5 at 16 range 1 .. 13;
16018 l6 at 18 range 0 .. 71;
16023 The Size values are actually the Object_Size, i.e.@: the default size that
16024 will be allocated for objects of the type.
16025 The ?? size for type r indicates that we have a variant record, and the
16026 actual size of objects will depend on the discriminant value.
16028 The Alignment values show the actual alignment chosen by the compiler
16029 for each record or array type.
16031 The record representation clause for type r shows where all fields
16032 are placed, including the compiler generated tag field (whose location
16033 cannot be controlled by the programmer).
16035 The record representation clause for the type extension r2 shows all the
16036 fields present, including the parent field, which is a copy of the fields
16037 of the parent type of r2, i.e.@: r1.
16039 The component size and size clauses for types rb1 and rb2 show
16040 the exact effect of pragma @code{Pack} on these arrays, and the record
16041 representation clause for type x2 shows how pragma @code{Pack} affects
16044 In some cases, it may be useful to cut and paste the representation clauses
16045 generated by the compiler into the original source to fix and guarantee
16046 the actual representation to be used.
16048 @node Standard Library Routines
16049 @chapter Standard Library Routines
16052 The Ada Reference Manual contains in Annex A a full description of an
16053 extensive set of standard library routines that can be used in any Ada
16054 program, and which must be provided by all Ada compilers. They are
16055 analogous to the standard C library used by C programs.
16057 GNAT implements all of the facilities described in annex A, and for most
16058 purposes the description in the Ada Reference Manual, or appropriate Ada
16059 text book, will be sufficient for making use of these facilities.
16061 In the case of the input-output facilities,
16062 @xref{The Implementation of Standard I/O},
16063 gives details on exactly how GNAT interfaces to the
16064 file system. For the remaining packages, the Ada Reference Manual
16065 should be sufficient. The following is a list of the packages included,
16066 together with a brief description of the functionality that is provided.
16068 For completeness, references are included to other predefined library
16069 routines defined in other sections of the Ada Reference Manual (these are
16070 cross-indexed from Annex A). For further details see the relevant
16071 package declarations in the run-time library. In particular, a few units
16072 are not implemented, as marked by the presence of pragma Unimplemented_Unit,
16073 and in this case the package declaration contains comments explaining why
16074 the unit is not implemented.
16078 This is a parent package for all the standard library packages. It is
16079 usually included implicitly in your program, and itself contains no
16080 useful data or routines.
16082 @item Ada.Assertions (11.4.2)
16083 @code{Assertions} provides the @code{Assert} subprograms, and also
16084 the declaration of the @code{Assertion_Error} exception.
16086 @item Ada.Asynchronous_Task_Control (D.11)
16087 @code{Asynchronous_Task_Control} provides low level facilities for task
16088 synchronization. It is typically not implemented. See package spec for details.
16090 @item Ada.Calendar (9.6)
16091 @code{Calendar} provides time of day access, and routines for
16092 manipulating times and durations.
16094 @item Ada.Calendar.Arithmetic (9.6.1)
16095 This package provides additional arithmetic
16096 operations for @code{Calendar}.
16098 @item Ada.Calendar.Formatting (9.6.1)
16099 This package provides formatting operations for @code{Calendar}.
16101 @item Ada.Calendar.Time_Zones (9.6.1)
16102 This package provides additional @code{Calendar} facilities
16103 for handling time zones.
16105 @item Ada.Characters (A.3.1)
16106 This is a dummy parent package that contains no useful entities
16108 @item Ada.Characters.Conversions (A.3.2)
16109 This package provides character conversion functions.
16111 @item Ada.Characters.Handling (A.3.2)
16112 This package provides some basic character handling capabilities,
16113 including classification functions for classes of characters (e.g.@: test
16114 for letters, or digits).
16116 @item Ada.Characters.Latin_1 (A.3.3)
16117 This package includes a complete set of definitions of the characters
16118 that appear in type CHARACTER@. It is useful for writing programs that
16119 will run in international environments. For example, if you want an
16120 upper case E with an acute accent in a string, it is often better to use
16121 the definition of @code{UC_E_Acute} in this package. Then your program
16122 will print in an understandable manner even if your environment does not
16123 support these extended characters.
16125 @item Ada.Command_Line (A.15)
16126 This package provides access to the command line parameters and the name
16127 of the current program (analogous to the use of @code{argc} and @code{argv}
16128 in C), and also allows the exit status for the program to be set in a
16129 system-independent manner.
16131 @item Ada.Complex_Text_IO (G.1.3)
16132 This package provides text input and output of complex numbers.
16134 @item Ada.Containers (A.18.1)
16135 A top level package providing a few basic definitions used by all the
16136 following specific child packages that provide specific kinds of
16139 @item Ada.Containers.Bounded_Priority_Queues (A.18.31)
16141 @item Ada.Containers.Bounded_Synchronized_Queues (A.18.29)
16143 @item Ada.Containers.Doubly_Linked_Lists (A.18.3)
16145 @item Ada.Containers.Generic_Array_Sort (A.18.26)
16147 @item Ada.Containers.Generic_Constrained_Array_Sort (A.18.26)
16149 @item Ada.Containers.Generic_Sort (A.18.26)
16151 @item Ada.Containers.Hashed_Maps (A.18.5)
16153 @item Ada.Containers.Hashed_Sets (A.18.8)
16155 @item Ada.Containers.Indefinite_Doubly_Linked_Lists (A.18.12)
16157 @item Ada.Containers.Indefinite_Hashed_Maps (A.18.13)
16159 @item Ada.Containers.Indefinite_Hashed_Sets (A.18.15)
16161 @item Ada.Containers.Indefinite_Holders (A.18.18)
16163 @item Ada.Containers.Indefinite_Multiway_Trees (A.18.17)
16165 @item Ada.Containers.Indefinite_Ordered_Maps (A.18.14)
16167 @item Ada.Containers.Indefinite_Ordered_Sets (A.18.16)
16169 @item Ada.Containers.Indefinite_Vectors (A.18.11)
16171 @item Ada.Containers.Multiway_Trees (A.18.10)
16173 @item Ada.Containers.Ordered_Maps (A.18.6)
16175 @item Ada.Containers.Ordered_Sets (A.18.9)
16177 @item Ada.Containers.Synchronized_Queue_Interfaces (A.18.27)
16179 @item Ada.Containers.Unbounded_Priority_Queues (A.18.30)
16181 @item Ada.Containers.Unbounded_Synchronized_Queues (A.18.28)
16183 @item Ada.Containers.Vectors (A.18.2)
16185 @item Ada.Directories (A.16)
16186 This package provides operations on directories.
16188 @item Ada.Directories.Hierarchical_File_Names (A.16.1)
16189 This package provides additional directory operations handling
16190 hiearchical file names.
16192 @item Ada.Directories.Information (A.16)
16193 This is an implementation defined package for additional directory
16194 operations, which is not implemented in GNAT.
16196 @item Ada.Decimal (F.2)
16197 This package provides constants describing the range of decimal numbers
16198 implemented, and also a decimal divide routine (analogous to the COBOL
16199 verb DIVIDE @dots{} GIVING @dots{} REMAINDER @dots{})
16201 @item Ada.Direct_IO (A.8.4)
16202 This package provides input-output using a model of a set of records of
16203 fixed-length, containing an arbitrary definite Ada type, indexed by an
16204 integer record number.
16206 @item Ada.Dispatching (D.2.1)
16207 A parent package containing definitions for task dispatching operations.
16209 @item Ada.Dispatching.EDF (D.2.6)
16210 Not implemented in GNAT.
16212 @item Ada.Dispatching.Non_Preemptive (D.2.4)
16213 Not implemented in GNAT.
16215 @item Ada.Dispatching.Round_Robin (D.2.5)
16216 Not implemented in GNAT.
16218 @item Ada.Dynamic_Priorities (D.5)
16219 This package allows the priorities of a task to be adjusted dynamically
16220 as the task is running.
16222 @item Ada.Environment_Variables (A.17)
16223 This package provides facilities for accessing environment variables.
16225 @item Ada.Exceptions (11.4.1)
16226 This package provides additional information on exceptions, and also
16227 contains facilities for treating exceptions as data objects, and raising
16228 exceptions with associated messages.
16230 @item Ada.Execution_Time (D.14)
16231 Not implemented in GNAT.
16233 @item Ada.Execution_Time.Group_Budgets (D.14.2)
16234 Not implemented in GNAT.
16236 @item Ada.Execution_Time.Timers (D.14.1)'
16237 Not implemented in GNAT.
16239 @item Ada.Finalization (7.6)
16240 This package contains the declarations and subprograms to support the
16241 use of controlled types, providing for automatic initialization and
16242 finalization (analogous to the constructors and destructors of C++).
16244 @item Ada.Float_Text_IO (A.10.9)
16245 A library level instantiation of Text_IO.Float_IO for type Float.
16247 @item Ada.Float_Wide_Text_IO (A.10.9)
16248 A library level instantiation of Wide_Text_IO.Float_IO for type Float.
16250 @item Ada.Float_Wide_Wide_Text_IO (A.10.9)
16251 A library level instantiation of Wide_Wide_Text_IO.Float_IO for type Float.
16253 @item Ada.Integer_Text_IO (A.10.9)
16254 A library level instantiation of Text_IO.Integer_IO for type Integer.
16256 @item Ada.Integer_Wide_Text_IO (A.10.9)
16257 A library level instantiation of Wide_Text_IO.Integer_IO for type Integer.
16259 @item Ada.Integer_Wide_Wide_Text_IO (A.10.9)
16260 A library level instantiation of Wide_Wide_Text_IO.Integer_IO for type Integer.
16262 @item Ada.Interrupts (C.3.2)
16263 This package provides facilities for interfacing to interrupts, which
16264 includes the set of signals or conditions that can be raised and
16265 recognized as interrupts.
16267 @item Ada.Interrupts.Names (C.3.2)
16268 This package provides the set of interrupt names (actually signal
16269 or condition names) that can be handled by GNAT@.
16271 @item Ada.IO_Exceptions (A.13)
16272 This package defines the set of exceptions that can be raised by use of
16273 the standard IO packages.
16275 @item Ada.Iterator_Interfaces (5.5.1)
16276 This package provides a generic interface to generalized iterators.
16278 @item Ada.Locales (A.19)
16279 This package provides declarations providing information (Language
16280 and Country) about the current locale.
16283 This package contains some standard constants and exceptions used
16284 throughout the numerics packages. Note that the constants pi and e are
16285 defined here, and it is better to use these definitions than rolling
16288 @item Ada.Numerics.Complex_Arrays (G.3.2)
16289 Provides operations on arrays of complex numbers.
16291 @item Ada.Numerics.Complex_Elementary_Functions
16292 Provides the implementation of standard elementary functions (such as
16293 log and trigonometric functions) operating on complex numbers using the
16294 standard @code{Float} and the @code{Complex} and @code{Imaginary} types
16295 created by the package @code{Numerics.Complex_Types}.
16297 @item Ada.Numerics.Complex_Types
16298 This is a predefined instantiation of
16299 @code{Numerics.Generic_Complex_Types} using @code{Standard.Float} to
16300 build the type @code{Complex} and @code{Imaginary}.
16302 @item Ada.Numerics.Discrete_Random
16303 This generic package provides a random number generator suitable for generating
16304 uniformly distributed values of a specified discrete subtype.
16306 @item Ada.Numerics.Float_Random
16307 This package provides a random number generator suitable for generating
16308 uniformly distributed floating point values in the unit interval.
16310 @item Ada.Numerics.Generic_Complex_Elementary_Functions
16311 This is a generic version of the package that provides the
16312 implementation of standard elementary functions (such as log and
16313 trigonometric functions) for an arbitrary complex type.
16315 The following predefined instantiations of this package are provided:
16319 @code{Ada.Numerics.Short_Complex_Elementary_Functions}
16321 @code{Ada.Numerics.Complex_Elementary_Functions}
16323 @code{Ada.Numerics.Long_Complex_Elementary_Functions}
16326 @item Ada.Numerics.Generic_Complex_Types
16327 This is a generic package that allows the creation of complex types,
16328 with associated complex arithmetic operations.
16330 The following predefined instantiations of this package exist
16333 @code{Ada.Numerics.Short_Complex_Complex_Types}
16335 @code{Ada.Numerics.Complex_Complex_Types}
16337 @code{Ada.Numerics.Long_Complex_Complex_Types}
16340 @item Ada.Numerics.Generic_Elementary_Functions
16341 This is a generic package that provides the implementation of standard
16342 elementary functions (such as log an trigonometric functions) for an
16343 arbitrary float type.
16345 The following predefined instantiations of this package exist
16349 @code{Ada.Numerics.Short_Elementary_Functions}
16351 @code{Ada.Numerics.Elementary_Functions}
16353 @code{Ada.Numerics.Long_Elementary_Functions}
16356 @item Ada.Numerics.Generic_Real_Arrays (G.3.1)
16357 Generic operations on arrays of reals
16359 @item Ada.Numerics.Real_Arrays (G.3.1)
16360 Preinstantiation of Ada.Numerics.Generic_Real_Arrays (Float).
16362 @item Ada.Real_Time (D.8)
16363 This package provides facilities similar to those of @code{Calendar}, but
16364 operating with a finer clock suitable for real time control. Note that
16365 annex D requires that there be no backward clock jumps, and GNAT generally
16366 guarantees this behavior, but of course if the external clock on which
16367 the GNAT runtime depends is deliberately reset by some external event,
16368 then such a backward jump may occur.
16370 @item Ada.Real_Time.Timing_Events (D.15)
16371 Not implemented in GNAT.
16373 @item Ada.Sequential_IO (A.8.1)
16374 This package provides input-output facilities for sequential files,
16375 which can contain a sequence of values of a single type, which can be
16376 any Ada type, including indefinite (unconstrained) types.
16378 @item Ada.Storage_IO (A.9)
16379 This package provides a facility for mapping arbitrary Ada types to and
16380 from a storage buffer. It is primarily intended for the creation of new
16383 @item Ada.Streams (13.13.1)
16384 This is a generic package that provides the basic support for the
16385 concept of streams as used by the stream attributes (@code{Input},
16386 @code{Output}, @code{Read} and @code{Write}).
16388 @item Ada.Streams.Stream_IO (A.12.1)
16389 This package is a specialization of the type @code{Streams} defined in
16390 package @code{Streams} together with a set of operations providing
16391 Stream_IO capability. The Stream_IO model permits both random and
16392 sequential access to a file which can contain an arbitrary set of values
16393 of one or more Ada types.
16395 @item Ada.Strings (A.4.1)
16396 This package provides some basic constants used by the string handling
16399 @item Ada.Strings.Bounded (A.4.4)
16400 This package provides facilities for handling variable length
16401 strings. The bounded model requires a maximum length. It is thus
16402 somewhat more limited than the unbounded model, but avoids the use of
16403 dynamic allocation or finalization.
16405 @item Ada.Strings.Bounded.Equal_Case_Insensitive (A.4.10)
16406 Provides case-insensitive comparisons of bounded strings
16408 @item Ada.Strings.Bounded.Hash (A.4.9)
16409 This package provides a generic hash function for bounded strings
16411 @item Ada.Strings.Bounded.Hash_Case_Insensitive (A.4.9)
16412 This package provides a generic hash function for bounded strings that
16413 converts the string to be hashed to lower case.
16415 @item Ada.Strings.Bounded.Less_Case_Insensitive (A.4.10)
16416 This package provides a comparison function for bounded strings that works
16417 in a case insensitive manner by converting to lower case before the comparison.
16419 @item Ada.Strings.Fixed (A.4.3)
16420 This package provides facilities for handling fixed length strings.
16422 @item Ada.Strings.Fixed.Equal_Case_Insensitive (A.4.10)
16423 This package provides an equality function for fixed strings that compares
16424 the strings after converting both to lower case.
16426 @item Ada.Strings.Fixed.Hash_Case_Insensitive (A.4.9)
16427 This package provides a case insensitive hash function for fixed strings that
16428 converts the string to lower case before computing the hash.
16430 @item Ada.Strings.Fixed.Less_Case_Insensitive (A.4.10)
16431 This package provides a comparison function for fixed strings that works
16432 in a case insensitive manner by converting to lower case before the comparison.
16434 Ada.Strings.Hash (A.4.9)
16435 This package provides a hash function for strings.
16437 Ada.Strings.Hash_Case_Insensitive (A.4.9)
16438 This package provides a hash function for strings that is case insensitive.
16439 The string is converted to lower case before computing the hash.
16441 @item Ada.Strings.Less_Case_Insensitive (A.4.10)
16442 This package provides a comparison function for\strings that works
16443 in a case insensitive manner by converting to lower case before the comparison.
16445 @item Ada.Strings.Maps (A.4.2)
16446 This package provides facilities for handling character mappings and
16447 arbitrarily defined subsets of characters. For instance it is useful in
16448 defining specialized translation tables.
16450 @item Ada.Strings.Maps.Constants (A.4.6)
16451 This package provides a standard set of predefined mappings and
16452 predefined character sets. For example, the standard upper to lower case
16453 conversion table is found in this package. Note that upper to lower case
16454 conversion is non-trivial if you want to take the entire set of
16455 characters, including extended characters like E with an acute accent,
16456 into account. You should use the mappings in this package (rather than
16457 adding 32 yourself) to do case mappings.
16459 @item Ada.Strings.Unbounded (A.4.5)
16460 This package provides facilities for handling variable length
16461 strings. The unbounded model allows arbitrary length strings, but
16462 requires the use of dynamic allocation and finalization.
16464 @item Ada.Strings.Unbounded.Equal_Case_Insensitive (A.4.10)
16465 Provides case-insensitive comparisons of unbounded strings
16467 @item Ada.Strings.Unbounded.Hash (A.4.9)
16468 This package provides a generic hash function for unbounded strings
16470 @item Ada.Strings.Unbounded.Hash_Case_Insensitive (A.4.9)
16471 This package provides a generic hash function for unbounded strings that
16472 converts the string to be hashed to lower case.
16474 @item Ada.Strings.Unbounded.Less_Case_Insensitive (A.4.10)
16475 This package provides a comparison function for unbounded strings that works
16476 in a case insensitive manner by converting to lower case before the comparison.
16478 @item Ada.Strings.UTF_Encoding (A.4.11)
16479 This package provides basic definitions for dealing with UTF-encoded strings.
16481 @item Ada.Strings.UTF_Encoding.Conversions (A.4.11)
16482 This package provides conversion functions for UTF-encoded strings.
16484 @item Ada.Strings.UTF_Encoding.Strings (A.4.11)
16485 @itemx Ada.Strings.UTF_Encoding.Wide_Strings (A.4.11)
16486 @itemx Ada.Strings.UTF_Encoding.Wide_Wide_Strings (A.4.11)
16487 These packages provide facilities for handling UTF encodings for
16488 Strings, Wide_Strings and Wide_Wide_Strings.
16490 @item Ada.Strings.Wide_Bounded (A.4.7)
16491 @itemx Ada.Strings.Wide_Fixed (A.4.7)
16492 @itemx Ada.Strings.Wide_Maps (A.4.7)
16493 @itemx Ada.Strings.Wide_Unbounded (A.4.7)
16494 These packages provide analogous capabilities to the corresponding
16495 packages without @samp{Wide_} in the name, but operate with the types
16496 @code{Wide_String} and @code{Wide_Character} instead of @code{String}
16497 and @code{Character}. Versions of all the child packages are available.
16499 @item Ada.Strings.Wide_Wide_Bounded (A.4.7)
16500 @itemx Ada.Strings.Wide_Wide_Fixed (A.4.7)
16501 @itemx Ada.Strings.Wide_Wide_Maps (A.4.7)
16502 @itemx Ada.Strings.Wide_Wide_Unbounded (A.4.7)
16503 These packages provide analogous capabilities to the corresponding
16504 packages without @samp{Wide_} in the name, but operate with the types
16505 @code{Wide_Wide_String} and @code{Wide_Wide_Character} instead
16506 of @code{String} and @code{Character}.
16508 @item Ada.Synchronous_Barriers (D.10.1)
16509 This package provides facilities for synchronizing tasks at a low level
16512 @item Ada.Synchronous_Task_Control (D.10)
16513 This package provides some standard facilities for controlling task
16514 communication in a synchronous manner.
16516 @item Ada.Synchronous_Task_Control.EDF (D.10)
16517 Not implemented in GNAT.
16520 This package contains definitions for manipulation of the tags of tagged
16523 @item Ada.Tags.Generic_Dispatching_Constructor (3.9)
16524 This package provides a way of constructing tagged class-wide values given
16525 only the tag value.
16527 @item Ada.Task_Attributes (C.7.2)
16528 This package provides the capability of associating arbitrary
16529 task-specific data with separate tasks.
16531 @item Ada.Task_Identifification (C.7.1)
16532 This package provides capabilities for task identification.
16534 @item Ada.Task_Termination (C.7.3)
16535 This package provides control over task termination.
16538 This package provides basic text input-output capabilities for
16539 character, string and numeric data. The subpackages of this
16540 package are listed next. Note that although these are defined
16541 as subpackages in the RM, they are actually transparently
16542 implemented as child packages in GNAT, meaning that they
16543 are only loaded if needed.
16545 @item Ada.Text_IO.Decimal_IO
16546 Provides input-output facilities for decimal fixed-point types
16548 @item Ada.Text_IO.Enumeration_IO
16549 Provides input-output facilities for enumeration types.
16551 @item Ada.Text_IO.Fixed_IO
16552 Provides input-output facilities for ordinary fixed-point types.
16554 @item Ada.Text_IO.Float_IO
16555 Provides input-output facilities for float types. The following
16556 predefined instantiations of this generic package are available:
16560 @code{Short_Float_Text_IO}
16562 @code{Float_Text_IO}
16564 @code{Long_Float_Text_IO}
16567 @item Ada.Text_IO.Integer_IO
16568 Provides input-output facilities for integer types. The following
16569 predefined instantiations of this generic package are available:
16572 @item Short_Short_Integer
16573 @code{Ada.Short_Short_Integer_Text_IO}
16574 @item Short_Integer
16575 @code{Ada.Short_Integer_Text_IO}
16577 @code{Ada.Integer_Text_IO}
16579 @code{Ada.Long_Integer_Text_IO}
16580 @item Long_Long_Integer
16581 @code{Ada.Long_Long_Integer_Text_IO}
16584 @item Ada.Text_IO.Modular_IO
16585 Provides input-output facilities for modular (unsigned) types.
16587 @item Ada.Text_IO.Bounded_IO (A.10.11)
16588 Provides input-output facilities for bounded strings.
16590 @item Ada.Text_IO.Complex_IO (G.1.3)
16591 This package provides basic text input-output capabilities for complex
16594 @item Ada.Text_IO.Editing (F.3.3)
16595 This package contains routines for edited output, analogous to the use
16596 of pictures in COBOL@. The picture formats used by this package are a
16597 close copy of the facility in COBOL@.
16599 @item Ada.Text_IO.Text_Streams (A.12.2)
16600 This package provides a facility that allows Text_IO files to be treated
16601 as streams, so that the stream attributes can be used for writing
16602 arbitrary data, including binary data, to Text_IO files.
16604 @item Ada.Text_IO.Unbounded_IO (A.10.12)
16605 This package provides input-output facilities for unbounded strings.
16607 @item Ada.Unchecked_Conversion (13.9)
16608 This generic package allows arbitrary conversion from one type to
16609 another of the same size, providing for breaking the type safety in
16610 special circumstances.
16612 If the types have the same Size (more accurately the same Value_Size),
16613 then the effect is simply to transfer the bits from the source to the
16614 target type without any modification. This usage is well defined, and
16615 for simple types whose representation is typically the same across
16616 all implementations, gives a portable method of performing such
16619 If the types do not have the same size, then the result is implementation
16620 defined, and thus may be non-portable. The following describes how GNAT
16621 handles such unchecked conversion cases.
16623 If the types are of different sizes, and are both discrete types, then
16624 the effect is of a normal type conversion without any constraint checking.
16625 In particular if the result type has a larger size, the result will be
16626 zero or sign extended. If the result type has a smaller size, the result
16627 will be truncated by ignoring high order bits.
16629 If the types are of different sizes, and are not both discrete types,
16630 then the conversion works as though pointers were created to the source
16631 and target, and the pointer value is converted. The effect is that bits
16632 are copied from successive low order storage units and bits of the source
16633 up to the length of the target type.
16635 A warning is issued if the lengths differ, since the effect in this
16636 case is implementation dependent, and the above behavior may not match
16637 that of some other compiler.
16639 A pointer to one type may be converted to a pointer to another type using
16640 unchecked conversion. The only case in which the effect is undefined is
16641 when one or both pointers are pointers to unconstrained array types. In
16642 this case, the bounds information may get incorrectly transferred, and in
16643 particular, GNAT uses double size pointers for such types, and it is
16644 meaningless to convert between such pointer types. GNAT will issue a
16645 warning if the alignment of the target designated type is more strict
16646 than the alignment of the source designated type (since the result may
16647 be unaligned in this case).
16649 A pointer other than a pointer to an unconstrained array type may be
16650 converted to and from System.Address. Such usage is common in Ada 83
16651 programs, but note that Ada.Address_To_Access_Conversions is the
16652 preferred method of performing such conversions in Ada 95 and Ada 2005.
16654 unchecked conversion nor Ada.Address_To_Access_Conversions should be
16655 used in conjunction with pointers to unconstrained objects, since
16656 the bounds information cannot be handled correctly in this case.
16658 @item Ada.Unchecked_Deallocation (13.11.2)
16659 This generic package allows explicit freeing of storage previously
16660 allocated by use of an allocator.
16662 @item Ada.Wide_Text_IO (A.11)
16663 This package is similar to @code{Ada.Text_IO}, except that the external
16664 file supports wide character representations, and the internal types are
16665 @code{Wide_Character} and @code{Wide_String} instead of @code{Character}
16666 and @code{String}. The corresponding set of nested packages and child
16667 packages are defined.
16669 @item Ada.Wide_Wide_Text_IO (A.11)
16670 This package is similar to @code{Ada.Text_IO}, except that the external
16671 file supports wide character representations, and the internal types are
16672 @code{Wide_Character} and @code{Wide_String} instead of @code{Character}
16673 and @code{String}. The corresponding set of nested packages and child
16674 packages are defined.
16678 For packages in Interfaces and System, all the RM defined packages are
16679 available in GNAT, see the Ada 2012 RM for full details.
16681 @node The Implementation of Standard I/O
16682 @chapter The Implementation of Standard I/O
16685 GNAT implements all the required input-output facilities described in
16686 A.6 through A.14. These sections of the Ada Reference Manual describe the
16687 required behavior of these packages from the Ada point of view, and if
16688 you are writing a portable Ada program that does not need to know the
16689 exact manner in which Ada maps to the outside world when it comes to
16690 reading or writing external files, then you do not need to read this
16691 chapter. As long as your files are all regular files (not pipes or
16692 devices), and as long as you write and read the files only from Ada, the
16693 description in the Ada Reference Manual is sufficient.
16695 However, if you want to do input-output to pipes or other devices, such
16696 as the keyboard or screen, or if the files you are dealing with are
16697 either generated by some other language, or to be read by some other
16698 language, then you need to know more about the details of how the GNAT
16699 implementation of these input-output facilities behaves.
16701 In this chapter we give a detailed description of exactly how GNAT
16702 interfaces to the file system. As always, the sources of the system are
16703 available to you for answering questions at an even more detailed level,
16704 but for most purposes the information in this chapter will suffice.
16706 Another reason that you may need to know more about how input-output is
16707 implemented arises when you have a program written in mixed languages
16708 where, for example, files are shared between the C and Ada sections of
16709 the same program. GNAT provides some additional facilities, in the form
16710 of additional child library packages, that facilitate this sharing, and
16711 these additional facilities are also described in this chapter.
16714 * Standard I/O Packages::
16720 * Wide_Wide_Text_IO::
16722 * Text Translation::
16724 * Filenames encoding::
16726 * Operations on C Streams::
16727 * Interfacing to C Streams::
16730 @node Standard I/O Packages
16731 @section Standard I/O Packages
16734 The Standard I/O packages described in Annex A for
16740 Ada.Text_IO.Complex_IO
16742 Ada.Text_IO.Text_Streams
16746 Ada.Wide_Text_IO.Complex_IO
16748 Ada.Wide_Text_IO.Text_Streams
16750 Ada.Wide_Wide_Text_IO
16752 Ada.Wide_Wide_Text_IO.Complex_IO
16754 Ada.Wide_Wide_Text_IO.Text_Streams
16764 are implemented using the C
16765 library streams facility; where
16769 All files are opened using @code{fopen}.
16771 All input/output operations use @code{fread}/@code{fwrite}.
16775 There is no internal buffering of any kind at the Ada library level. The only
16776 buffering is that provided at the system level in the implementation of the
16777 library routines that support streams. This facilitates shared use of these
16778 streams by mixed language programs. Note though that system level buffering is
16779 explicitly enabled at elaboration of the standard I/O packages and that can
16780 have an impact on mixed language programs, in particular those using I/O before
16781 calling the Ada elaboration routine (e.g.@: adainit). It is recommended to call
16782 the Ada elaboration routine before performing any I/O or when impractical,
16783 flush the common I/O streams and in particular Standard_Output before
16784 elaborating the Ada code.
16787 @section FORM Strings
16790 The format of a FORM string in GNAT is:
16793 "keyword=value,keyword=value,@dots{},keyword=value"
16797 where letters may be in upper or lower case, and there are no spaces
16798 between values. The order of the entries is not important. Currently
16799 the following keywords defined.
16802 TEXT_TRANSLATION=[YES|NO]
16804 WCEM=[n|h|u|s|e|8|b]
16805 ENCODING=[UTF8|8BITS]
16809 The use of these parameters is described later in this section. If an
16810 unrecognized keyword appears in a form string, it is silently ignored
16811 and not considered invalid.
16814 For OpenVMS additional FORM string keywords are available for use with
16815 RMS services. The syntax is:
16818 VMS_RMS_Keys=(keyword=value,@dots{},keyword=value)
16822 The following RMS keywords and values are currently defined:
16825 Context=Force_Stream_Mode|Force_Record_Mode
16829 VMS RMS keys are silently ignored on non-VMS systems. On OpenVMS
16830 unimplented RMS keywords, values, or invalid syntax will raise Use_Error.
16836 Direct_IO can only be instantiated for definite types. This is a
16837 restriction of the Ada language, which means that the records are fixed
16838 length (the length being determined by @code{@var{type}'Size}, rounded
16839 up to the next storage unit boundary if necessary).
16841 The records of a Direct_IO file are simply written to the file in index
16842 sequence, with the first record starting at offset zero, and subsequent
16843 records following. There is no control information of any kind. For
16844 example, if 32-bit integers are being written, each record takes
16845 4-bytes, so the record at index @var{K} starts at offset
16846 (@var{K}@minus{}1)*4.
16848 There is no limit on the size of Direct_IO files, they are expanded as
16849 necessary to accommodate whatever records are written to the file.
16851 @node Sequential_IO
16852 @section Sequential_IO
16855 Sequential_IO may be instantiated with either a definite (constrained)
16856 or indefinite (unconstrained) type.
16858 For the definite type case, the elements written to the file are simply
16859 the memory images of the data values with no control information of any
16860 kind. The resulting file should be read using the same type, no validity
16861 checking is performed on input.
16863 For the indefinite type case, the elements written consist of two
16864 parts. First is the size of the data item, written as the memory image
16865 of a @code{Interfaces.C.size_t} value, followed by the memory image of
16866 the data value. The resulting file can only be read using the same
16867 (unconstrained) type. Normal assignment checks are performed on these
16868 read operations, and if these checks fail, @code{Data_Error} is
16869 raised. In particular, in the array case, the lengths must match, and in
16870 the variant record case, if the variable for a particular read operation
16871 is constrained, the discriminants must match.
16873 Note that it is not possible to use Sequential_IO to write variable
16874 length array items, and then read the data back into different length
16875 arrays. For example, the following will raise @code{Data_Error}:
16877 @smallexample @c ada
16878 package IO is new Sequential_IO (String);
16883 IO.Write (F, "hello!")
16884 IO.Reset (F, Mode=>In_File);
16891 On some Ada implementations, this will print @code{hell}, but the program is
16892 clearly incorrect, since there is only one element in the file, and that
16893 element is the string @code{hello!}.
16895 In Ada 95 and Ada 2005, this kind of behavior can be legitimately achieved
16896 using Stream_IO, and this is the preferred mechanism. In particular, the
16897 above program fragment rewritten to use Stream_IO will work correctly.
16903 Text_IO files consist of a stream of characters containing the following
16904 special control characters:
16907 LF (line feed, 16#0A#) Line Mark
16908 FF (form feed, 16#0C#) Page Mark
16912 A canonical Text_IO file is defined as one in which the following
16913 conditions are met:
16917 The character @code{LF} is used only as a line mark, i.e.@: to mark the end
16921 The character @code{FF} is used only as a page mark, i.e.@: to mark the
16922 end of a page and consequently can appear only immediately following a
16923 @code{LF} (line mark) character.
16926 The file ends with either @code{LF} (line mark) or @code{LF}-@code{FF}
16927 (line mark, page mark). In the former case, the page mark is implicitly
16928 assumed to be present.
16932 A file written using Text_IO will be in canonical form provided that no
16933 explicit @code{LF} or @code{FF} characters are written using @code{Put}
16934 or @code{Put_Line}. There will be no @code{FF} character at the end of
16935 the file unless an explicit @code{New_Page} operation was performed
16936 before closing the file.
16938 A canonical Text_IO file that is a regular file (i.e., not a device or a
16939 pipe) can be read using any of the routines in Text_IO@. The
16940 semantics in this case will be exactly as defined in the Ada Reference
16941 Manual, and all the routines in Text_IO are fully implemented.
16943 A text file that does not meet the requirements for a canonical Text_IO
16944 file has one of the following:
16948 The file contains @code{FF} characters not immediately following a
16949 @code{LF} character.
16952 The file contains @code{LF} or @code{FF} characters written by
16953 @code{Put} or @code{Put_Line}, which are not logically considered to be
16954 line marks or page marks.
16957 The file ends in a character other than @code{LF} or @code{FF},
16958 i.e.@: there is no explicit line mark or page mark at the end of the file.
16962 Text_IO can be used to read such non-standard text files but subprograms
16963 to do with line or page numbers do not have defined meanings. In
16964 particular, a @code{FF} character that does not follow a @code{LF}
16965 character may or may not be treated as a page mark from the point of
16966 view of page and line numbering. Every @code{LF} character is considered
16967 to end a line, and there is an implied @code{LF} character at the end of
16971 * Text_IO Stream Pointer Positioning::
16972 * Text_IO Reading and Writing Non-Regular Files::
16974 * Treating Text_IO Files as Streams::
16975 * Text_IO Extensions::
16976 * Text_IO Facilities for Unbounded Strings::
16979 @node Text_IO Stream Pointer Positioning
16980 @subsection Stream Pointer Positioning
16983 @code{Ada.Text_IO} has a definition of current position for a file that
16984 is being read. No internal buffering occurs in Text_IO, and usually the
16985 physical position in the stream used to implement the file corresponds
16986 to this logical position defined by Text_IO@. There are two exceptions:
16990 After a call to @code{End_Of_Page} that returns @code{True}, the stream
16991 is positioned past the @code{LF} (line mark) that precedes the page
16992 mark. Text_IO maintains an internal flag so that subsequent read
16993 operations properly handle the logical position which is unchanged by
16994 the @code{End_Of_Page} call.
16997 After a call to @code{End_Of_File} that returns @code{True}, if the
16998 Text_IO file was positioned before the line mark at the end of file
16999 before the call, then the logical position is unchanged, but the stream
17000 is physically positioned right at the end of file (past the line mark,
17001 and past a possible page mark following the line mark. Again Text_IO
17002 maintains internal flags so that subsequent read operations properly
17003 handle the logical position.
17007 These discrepancies have no effect on the observable behavior of
17008 Text_IO, but if a single Ada stream is shared between a C program and
17009 Ada program, or shared (using @samp{shared=yes} in the form string)
17010 between two Ada files, then the difference may be observable in some
17013 @node Text_IO Reading and Writing Non-Regular Files
17014 @subsection Reading and Writing Non-Regular Files
17017 A non-regular file is a device (such as a keyboard), or a pipe. Text_IO
17018 can be used for reading and writing. Writing is not affected and the
17019 sequence of characters output is identical to the normal file case, but
17020 for reading, the behavior of Text_IO is modified to avoid undesirable
17021 look-ahead as follows:
17023 An input file that is not a regular file is considered to have no page
17024 marks. Any @code{Ascii.FF} characters (the character normally used for a
17025 page mark) appearing in the file are considered to be data
17026 characters. In particular:
17030 @code{Get_Line} and @code{Skip_Line} do not test for a page mark
17031 following a line mark. If a page mark appears, it will be treated as a
17035 This avoids the need to wait for an extra character to be typed or
17036 entered from the pipe to complete one of these operations.
17039 @code{End_Of_Page} always returns @code{False}
17042 @code{End_Of_File} will return @code{False} if there is a page mark at
17043 the end of the file.
17047 Output to non-regular files is the same as for regular files. Page marks
17048 may be written to non-regular files using @code{New_Page}, but as noted
17049 above they will not be treated as page marks on input if the output is
17050 piped to another Ada program.
17052 Another important discrepancy when reading non-regular files is that the end
17053 of file indication is not ``sticky''. If an end of file is entered, e.g.@: by
17054 pressing the @key{EOT} key,
17056 is signaled once (i.e.@: the test @code{End_Of_File}
17057 will yield @code{True}, or a read will
17058 raise @code{End_Error}), but then reading can resume
17059 to read data past that end of
17060 file indication, until another end of file indication is entered.
17062 @node Get_Immediate
17063 @subsection Get_Immediate
17064 @cindex Get_Immediate
17067 Get_Immediate returns the next character (including control characters)
17068 from the input file. In particular, Get_Immediate will return LF or FF
17069 characters used as line marks or page marks. Such operations leave the
17070 file positioned past the control character, and it is thus not treated
17071 as having its normal function. This means that page, line and column
17072 counts after this kind of Get_Immediate call are set as though the mark
17073 did not occur. In the case where a Get_Immediate leaves the file
17074 positioned between the line mark and page mark (which is not normally
17075 possible), it is undefined whether the FF character will be treated as a
17078 @node Treating Text_IO Files as Streams
17079 @subsection Treating Text_IO Files as Streams
17080 @cindex Stream files
17083 The package @code{Text_IO.Streams} allows a Text_IO file to be treated
17084 as a stream. Data written to a Text_IO file in this stream mode is
17085 binary data. If this binary data contains bytes 16#0A# (@code{LF}) or
17086 16#0C# (@code{FF}), the resulting file may have non-standard
17087 format. Similarly if read operations are used to read from a Text_IO
17088 file treated as a stream, then @code{LF} and @code{FF} characters may be
17089 skipped and the effect is similar to that described above for
17090 @code{Get_Immediate}.
17092 @node Text_IO Extensions
17093 @subsection Text_IO Extensions
17094 @cindex Text_IO extensions
17097 A package GNAT.IO_Aux in the GNAT library provides some useful extensions
17098 to the standard @code{Text_IO} package:
17101 @item function File_Exists (Name : String) return Boolean;
17102 Determines if a file of the given name exists.
17104 @item function Get_Line return String;
17105 Reads a string from the standard input file. The value returned is exactly
17106 the length of the line that was read.
17108 @item function Get_Line (File : Ada.Text_IO.File_Type) return String;
17109 Similar, except that the parameter File specifies the file from which
17110 the string is to be read.
17114 @node Text_IO Facilities for Unbounded Strings
17115 @subsection Text_IO Facilities for Unbounded Strings
17116 @cindex Text_IO for unbounded strings
17117 @cindex Unbounded_String, Text_IO operations
17120 The package @code{Ada.Strings.Unbounded.Text_IO}
17121 in library files @code{a-suteio.ads/adb} contains some GNAT-specific
17122 subprograms useful for Text_IO operations on unbounded strings:
17126 @item function Get_Line (File : File_Type) return Unbounded_String;
17127 Reads a line from the specified file
17128 and returns the result as an unbounded string.
17130 @item procedure Put (File : File_Type; U : Unbounded_String);
17131 Writes the value of the given unbounded string to the specified file
17132 Similar to the effect of
17133 @code{Put (To_String (U))} except that an extra copy is avoided.
17135 @item procedure Put_Line (File : File_Type; U : Unbounded_String);
17136 Writes the value of the given unbounded string to the specified file,
17137 followed by a @code{New_Line}.
17138 Similar to the effect of @code{Put_Line (To_String (U))} except
17139 that an extra copy is avoided.
17143 In the above procedures, @code{File} is of type @code{Ada.Text_IO.File_Type}
17144 and is optional. If the parameter is omitted, then the standard input or
17145 output file is referenced as appropriate.
17147 The package @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} in library
17148 files @file{a-swuwti.ads} and @file{a-swuwti.adb} provides similar extended
17149 @code{Wide_Text_IO} functionality for unbounded wide strings.
17151 The package @code{Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO} in library
17152 files @file{a-szuzti.ads} and @file{a-szuzti.adb} provides similar extended
17153 @code{Wide_Wide_Text_IO} functionality for unbounded wide wide strings.
17156 @section Wide_Text_IO
17159 @code{Wide_Text_IO} is similar in most respects to Text_IO, except that
17160 both input and output files may contain special sequences that represent
17161 wide character values. The encoding scheme for a given file may be
17162 specified using a FORM parameter:
17169 as part of the FORM string (WCEM = wide character encoding method),
17170 where @var{x} is one of the following characters
17176 Upper half encoding
17188 The encoding methods match those that
17189 can be used in a source
17190 program, but there is no requirement that the encoding method used for
17191 the source program be the same as the encoding method used for files,
17192 and different files may use different encoding methods.
17194 The default encoding method for the standard files, and for opened files
17195 for which no WCEM parameter is given in the FORM string matches the
17196 wide character encoding specified for the main program (the default
17197 being brackets encoding if no coding method was specified with -gnatW).
17201 In this encoding, a wide character is represented by a five character
17209 where @var{a}, @var{b}, @var{c}, @var{d} are the four hexadecimal
17210 characters (using upper case letters) of the wide character code. For
17211 example, ESC A345 is used to represent the wide character with code
17212 16#A345#. This scheme is compatible with use of the full
17213 @code{Wide_Character} set.
17215 @item Upper Half Coding
17216 The wide character with encoding 16#abcd#, where the upper bit is on
17217 (i.e.@: a is in the range 8-F) is represented as two bytes 16#ab# and
17218 16#cd#. The second byte may never be a format control character, but is
17219 not required to be in the upper half. This method can be also used for
17220 shift-JIS or EUC where the internal coding matches the external coding.
17222 @item Shift JIS Coding
17223 A wide character is represented by a two character sequence 16#ab# and
17224 16#cd#, with the restrictions described for upper half encoding as
17225 described above. The internal character code is the corresponding JIS
17226 character according to the standard algorithm for Shift-JIS
17227 conversion. Only characters defined in the JIS code set table can be
17228 used with this encoding method.
17231 A wide character is represented by a two character sequence 16#ab# and
17232 16#cd#, with both characters being in the upper half. The internal
17233 character code is the corresponding JIS character according to the EUC
17234 encoding algorithm. Only characters defined in the JIS code set table
17235 can be used with this encoding method.
17238 A wide character is represented using
17239 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
17240 10646-1/Am.2. Depending on the character value, the representation
17241 is a one, two, or three byte sequence:
17244 16#0000#-16#007f#: 2#0xxxxxxx#
17245 16#0080#-16#07ff#: 2#110xxxxx# 2#10xxxxxx#
17246 16#0800#-16#ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
17250 where the @var{xxx} bits correspond to the left-padded bits of the
17251 16-bit character value. Note that all lower half ASCII characters
17252 are represented as ASCII bytes and all upper half characters and
17253 other wide characters are represented as sequences of upper-half
17254 (The full UTF-8 scheme allows for encoding 31-bit characters as
17255 6-byte sequences, but in this implementation, all UTF-8 sequences
17256 of four or more bytes length will raise a Constraint_Error, as
17257 will all invalid UTF-8 sequences.)
17259 @item Brackets Coding
17260 In this encoding, a wide character is represented by the following eight
17261 character sequence:
17268 where @code{a}, @code{b}, @code{c}, @code{d} are the four hexadecimal
17269 characters (using uppercase letters) of the wide character code. For
17270 example, @code{["A345"]} is used to represent the wide character with code
17272 This scheme is compatible with use of the full Wide_Character set.
17273 On input, brackets coding can also be used for upper half characters,
17274 e.g.@: @code{["C1"]} for lower case a. However, on output, brackets notation
17275 is only used for wide characters with a code greater than @code{16#FF#}.
17277 Note that brackets coding is not normally used in the context of
17278 Wide_Text_IO or Wide_Wide_Text_IO, since it is really just designed as
17279 a portable way of encoding source files. In the context of Wide_Text_IO
17280 or Wide_Wide_Text_IO, it can only be used if the file does not contain
17281 any instance of the left bracket character other than to encode wide
17282 character values using the brackets encoding method. In practice it is
17283 expected that some standard wide character encoding method such
17284 as UTF-8 will be used for text input output.
17286 If brackets notation is used, then any occurrence of a left bracket
17287 in the input file which is not the start of a valid wide character
17288 sequence will cause Constraint_Error to be raised. It is possible to
17289 encode a left bracket as ["5B"] and Wide_Text_IO and Wide_Wide_Text_IO
17290 input will interpret this as a left bracket.
17292 However, when a left bracket is output, it will be output as a left bracket
17293 and not as ["5B"]. We make this decision because for normal use of
17294 Wide_Text_IO for outputting messages, it is unpleasant to clobber left
17295 brackets. For example, if we write:
17298 Put_Line ("Start of output [first run]");
17302 we really do not want to have the left bracket in this message clobbered so
17303 that the output reads:
17306 Start of output ["5B"]first run]
17310 In practice brackets encoding is reasonably useful for normal Put_Line use
17311 since we won't get confused between left brackets and wide character
17312 sequences in the output. But for input, or when files are written out
17313 and read back in, it really makes better sense to use one of the standard
17314 encoding methods such as UTF-8.
17319 For the coding schemes other than UTF-8, Hex, or Brackets encoding,
17320 not all wide character
17321 values can be represented. An attempt to output a character that cannot
17322 be represented using the encoding scheme for the file causes
17323 Constraint_Error to be raised. An invalid wide character sequence on
17324 input also causes Constraint_Error to be raised.
17327 * Wide_Text_IO Stream Pointer Positioning::
17328 * Wide_Text_IO Reading and Writing Non-Regular Files::
17331 @node Wide_Text_IO Stream Pointer Positioning
17332 @subsection Stream Pointer Positioning
17335 @code{Ada.Wide_Text_IO} is similar to @code{Ada.Text_IO} in its handling
17336 of stream pointer positioning (@pxref{Text_IO}). There is one additional
17339 If @code{Ada.Wide_Text_IO.Look_Ahead} reads a character outside the
17340 normal lower ASCII set (i.e.@: a character in the range:
17342 @smallexample @c ada
17343 Wide_Character'Val (16#0080#) .. Wide_Character'Val (16#FFFF#)
17347 then although the logical position of the file pointer is unchanged by
17348 the @code{Look_Ahead} call, the stream is physically positioned past the
17349 wide character sequence. Again this is to avoid the need for buffering
17350 or backup, and all @code{Wide_Text_IO} routines check the internal
17351 indication that this situation has occurred so that this is not visible
17352 to a normal program using @code{Wide_Text_IO}. However, this discrepancy
17353 can be observed if the wide text file shares a stream with another file.
17355 @node Wide_Text_IO Reading and Writing Non-Regular Files
17356 @subsection Reading and Writing Non-Regular Files
17359 As in the case of Text_IO, when a non-regular file is read, it is
17360 assumed that the file contains no page marks (any form characters are
17361 treated as data characters), and @code{End_Of_Page} always returns
17362 @code{False}. Similarly, the end of file indication is not sticky, so
17363 it is possible to read beyond an end of file.
17365 @node Wide_Wide_Text_IO
17366 @section Wide_Wide_Text_IO
17369 @code{Wide_Wide_Text_IO} is similar in most respects to Text_IO, except that
17370 both input and output files may contain special sequences that represent
17371 wide wide character values. The encoding scheme for a given file may be
17372 specified using a FORM parameter:
17379 as part of the FORM string (WCEM = wide character encoding method),
17380 where @var{x} is one of the following characters
17386 Upper half encoding
17398 The encoding methods match those that
17399 can be used in a source
17400 program, but there is no requirement that the encoding method used for
17401 the source program be the same as the encoding method used for files,
17402 and different files may use different encoding methods.
17404 The default encoding method for the standard files, and for opened files
17405 for which no WCEM parameter is given in the FORM string matches the
17406 wide character encoding specified for the main program (the default
17407 being brackets encoding if no coding method was specified with -gnatW).
17412 A wide character is represented using
17413 UCS Transformation Format 8 (UTF-8) as defined in Annex R of ISO
17414 10646-1/Am.2. Depending on the character value, the representation
17415 is a one, two, three, or four byte sequence:
17418 16#000000#-16#00007f#: 2#0xxxxxxx#
17419 16#000080#-16#0007ff#: 2#110xxxxx# 2#10xxxxxx#
17420 16#000800#-16#00ffff#: 2#1110xxxx# 2#10xxxxxx# 2#10xxxxxx#
17421 16#010000#-16#10ffff#: 2#11110xxx# 2#10xxxxxx# 2#10xxxxxx# 2#10xxxxxx#
17425 where the @var{xxx} bits correspond to the left-padded bits of the
17426 21-bit character value. Note that all lower half ASCII characters
17427 are represented as ASCII bytes and all upper half characters and
17428 other wide characters are represented as sequences of upper-half
17431 @item Brackets Coding
17432 In this encoding, a wide wide character is represented by the following eight
17433 character sequence if is in wide character range
17439 and by the following ten character sequence if not
17442 [ " a b c d e f " ]
17446 where @code{a}, @code{b}, @code{c}, @code{d}, @code{e}, and @code{f}
17447 are the four or six hexadecimal
17448 characters (using uppercase letters) of the wide wide character code. For
17449 example, @code{["01A345"]} is used to represent the wide wide character
17450 with code @code{16#01A345#}.
17452 This scheme is compatible with use of the full Wide_Wide_Character set.
17453 On input, brackets coding can also be used for upper half characters,
17454 e.g.@: @code{["C1"]} for lower case a. However, on output, brackets notation
17455 is only used for wide characters with a code greater than @code{16#FF#}.
17460 If is also possible to use the other Wide_Character encoding methods,
17461 such as Shift-JIS, but the other schemes cannot support the full range
17462 of wide wide characters.
17463 An attempt to output a character that cannot
17464 be represented using the encoding scheme for the file causes
17465 Constraint_Error to be raised. An invalid wide character sequence on
17466 input also causes Constraint_Error to be raised.
17469 * Wide_Wide_Text_IO Stream Pointer Positioning::
17470 * Wide_Wide_Text_IO Reading and Writing Non-Regular Files::
17473 @node Wide_Wide_Text_IO Stream Pointer Positioning
17474 @subsection Stream Pointer Positioning
17477 @code{Ada.Wide_Wide_Text_IO} is similar to @code{Ada.Text_IO} in its handling
17478 of stream pointer positioning (@pxref{Text_IO}). There is one additional
17481 If @code{Ada.Wide_Wide_Text_IO.Look_Ahead} reads a character outside the
17482 normal lower ASCII set (i.e.@: a character in the range:
17484 @smallexample @c ada
17485 Wide_Wide_Character'Val (16#0080#) .. Wide_Wide_Character'Val (16#10FFFF#)
17489 then although the logical position of the file pointer is unchanged by
17490 the @code{Look_Ahead} call, the stream is physically positioned past the
17491 wide character sequence. Again this is to avoid the need for buffering
17492 or backup, and all @code{Wide_Wide_Text_IO} routines check the internal
17493 indication that this situation has occurred so that this is not visible
17494 to a normal program using @code{Wide_Wide_Text_IO}. However, this discrepancy
17495 can be observed if the wide text file shares a stream with another file.
17497 @node Wide_Wide_Text_IO Reading and Writing Non-Regular Files
17498 @subsection Reading and Writing Non-Regular Files
17501 As in the case of Text_IO, when a non-regular file is read, it is
17502 assumed that the file contains no page marks (any form characters are
17503 treated as data characters), and @code{End_Of_Page} always returns
17504 @code{False}. Similarly, the end of file indication is not sticky, so
17505 it is possible to read beyond an end of file.
17511 A stream file is a sequence of bytes, where individual elements are
17512 written to the file as described in the Ada Reference Manual. The type
17513 @code{Stream_Element} is simply a byte. There are two ways to read or
17514 write a stream file.
17518 The operations @code{Read} and @code{Write} directly read or write a
17519 sequence of stream elements with no control information.
17522 The stream attributes applied to a stream file transfer data in the
17523 manner described for stream attributes.
17526 @node Text Translation
17527 @section Text Translation
17530 @samp{Text_Translation=@var{xxx}} may be used as the Form parameter
17531 passed to Text_IO.Create and Text_IO.Open:
17532 @samp{Text_Translation=@var{Yes}} is the default, which means to
17533 translate LF to/from CR/LF on Windows systems.
17534 @samp{Text_Translation=@var{No}} disables this translation; i.e. it
17535 uses binary mode. For output files, @samp{Text_Translation=@var{No}}
17536 may be used to create Unix-style files on
17537 Windows. @samp{Text_Translation=@var{xxx}} has no effect on Unix
17541 @section Shared Files
17544 Section A.14 of the Ada Reference Manual allows implementations to
17545 provide a wide variety of behavior if an attempt is made to access the
17546 same external file with two or more internal files.
17548 To provide a full range of functionality, while at the same time
17549 minimizing the problems of portability caused by this implementation
17550 dependence, GNAT handles file sharing as follows:
17554 In the absence of a @samp{shared=@var{xxx}} form parameter, an attempt
17555 to open two or more files with the same full name is considered an error
17556 and is not supported. The exception @code{Use_Error} will be
17557 raised. Note that a file that is not explicitly closed by the program
17558 remains open until the program terminates.
17561 If the form parameter @samp{shared=no} appears in the form string, the
17562 file can be opened or created with its own separate stream identifier,
17563 regardless of whether other files sharing the same external file are
17564 opened. The exact effect depends on how the C stream routines handle
17565 multiple accesses to the same external files using separate streams.
17568 If the form parameter @samp{shared=yes} appears in the form string for
17569 each of two or more files opened using the same full name, the same
17570 stream is shared between these files, and the semantics are as described
17571 in Ada Reference Manual, Section A.14.
17575 When a program that opens multiple files with the same name is ported
17576 from another Ada compiler to GNAT, the effect will be that
17577 @code{Use_Error} is raised.
17579 The documentation of the original compiler and the documentation of the
17580 program should then be examined to determine if file sharing was
17581 expected, and @samp{shared=@var{xxx}} parameters added to @code{Open}
17582 and @code{Create} calls as required.
17584 When a program is ported from GNAT to some other Ada compiler, no
17585 special attention is required unless the @samp{shared=@var{xxx}} form
17586 parameter is used in the program. In this case, you must examine the
17587 documentation of the new compiler to see if it supports the required
17588 file sharing semantics, and form strings modified appropriately. Of
17589 course it may be the case that the program cannot be ported if the
17590 target compiler does not support the required functionality. The best
17591 approach in writing portable code is to avoid file sharing (and hence
17592 the use of the @samp{shared=@var{xxx}} parameter in the form string)
17595 One common use of file sharing in Ada 83 is the use of instantiations of
17596 Sequential_IO on the same file with different types, to achieve
17597 heterogeneous input-output. Although this approach will work in GNAT if
17598 @samp{shared=yes} is specified, it is preferable in Ada to use Stream_IO
17599 for this purpose (using the stream attributes)
17601 @node Filenames encoding
17602 @section Filenames encoding
17605 An encoding form parameter can be used to specify the filename
17606 encoding @samp{encoding=@var{xxx}}.
17610 If the form parameter @samp{encoding=utf8} appears in the form string, the
17611 filename must be encoded in UTF-8.
17614 If the form parameter @samp{encoding=8bits} appears in the form
17615 string, the filename must be a standard 8bits string.
17618 In the absence of a @samp{encoding=@var{xxx}} form parameter, the
17619 encoding is controlled by the @samp{GNAT_CODE_PAGE} environment
17620 variable. And if not set @samp{utf8} is assumed.
17624 The current system Windows ANSI code page.
17629 This encoding form parameter is only supported on the Windows
17630 platform. On the other Operating Systems the run-time is supporting
17634 @section Open Modes
17637 @code{Open} and @code{Create} calls result in a call to @code{fopen}
17638 using the mode shown in the following table:
17641 @center @code{Open} and @code{Create} Call Modes
17643 @b{OPEN } @b{CREATE}
17644 Append_File "r+" "w+"
17646 Out_File (Direct_IO) "r+" "w"
17647 Out_File (all other cases) "w" "w"
17648 Inout_File "r+" "w+"
17652 If text file translation is required, then either @samp{b} or @samp{t}
17653 is added to the mode, depending on the setting of Text. Text file
17654 translation refers to the mapping of CR/LF sequences in an external file
17655 to LF characters internally. This mapping only occurs in DOS and
17656 DOS-like systems, and is not relevant to other systems.
17658 A special case occurs with Stream_IO@. As shown in the above table, the
17659 file is initially opened in @samp{r} or @samp{w} mode for the
17660 @code{In_File} and @code{Out_File} cases. If a @code{Set_Mode} operation
17661 subsequently requires switching from reading to writing or vice-versa,
17662 then the file is reopened in @samp{r+} mode to permit the required operation.
17664 @node Operations on C Streams
17665 @section Operations on C Streams
17666 The package @code{Interfaces.C_Streams} provides an Ada program with direct
17667 access to the C library functions for operations on C streams:
17669 @smallexample @c adanocomment
17670 package Interfaces.C_Streams is
17671 -- Note: the reason we do not use the types that are in
17672 -- Interfaces.C is that we want to avoid dragging in the
17673 -- code in this unit if possible.
17674 subtype chars is System.Address;
17675 -- Pointer to null-terminated array of characters
17676 subtype FILEs is System.Address;
17677 -- Corresponds to the C type FILE*
17678 subtype voids is System.Address;
17679 -- Corresponds to the C type void*
17680 subtype int is Integer;
17681 subtype long is Long_Integer;
17682 -- Note: the above types are subtypes deliberately, and it
17683 -- is part of this spec that the above correspondences are
17684 -- guaranteed. This means that it is legitimate to, for
17685 -- example, use Integer instead of int. We provide these
17686 -- synonyms for clarity, but in some cases it may be
17687 -- convenient to use the underlying types (for example to
17688 -- avoid an unnecessary dependency of a spec on the spec
17690 type size_t is mod 2 ** Standard'Address_Size;
17691 NULL_Stream : constant FILEs;
17692 -- Value returned (NULL in C) to indicate an
17693 -- fdopen/fopen/tmpfile error
17694 ----------------------------------
17695 -- Constants Defined in stdio.h --
17696 ----------------------------------
17697 EOF : constant int;
17698 -- Used by a number of routines to indicate error or
17700 IOFBF : constant int;
17701 IOLBF : constant int;
17702 IONBF : constant int;
17703 -- Used to indicate buffering mode for setvbuf call
17704 SEEK_CUR : constant int;
17705 SEEK_END : constant int;
17706 SEEK_SET : constant int;
17707 -- Used to indicate origin for fseek call
17708 function stdin return FILEs;
17709 function stdout return FILEs;
17710 function stderr return FILEs;
17711 -- Streams associated with standard files
17712 --------------------------
17713 -- Standard C functions --
17714 --------------------------
17715 -- The functions selected below are ones that are
17716 -- available in UNIX (but not necessarily in ANSI C).
17717 -- These are very thin interfaces
17718 -- which copy exactly the C headers. For more
17719 -- documentation on these functions, see the Microsoft C
17720 -- "Run-Time Library Reference" (Microsoft Press, 1990,
17721 -- ISBN 1-55615-225-6), which includes useful information
17722 -- on system compatibility.
17723 procedure clearerr (stream : FILEs);
17724 function fclose (stream : FILEs) return int;
17725 function fdopen (handle : int; mode : chars) return FILEs;
17726 function feof (stream : FILEs) return int;
17727 function ferror (stream : FILEs) return int;
17728 function fflush (stream : FILEs) return int;
17729 function fgetc (stream : FILEs) return int;
17730 function fgets (strng : chars; n : int; stream : FILEs)
17732 function fileno (stream : FILEs) return int;
17733 function fopen (filename : chars; Mode : chars)
17735 -- Note: to maintain target independence, use
17736 -- text_translation_required, a boolean variable defined in
17737 -- a-sysdep.c to deal with the target dependent text
17738 -- translation requirement. If this variable is set,
17739 -- then b/t should be appended to the standard mode
17740 -- argument to set the text translation mode off or on
17742 function fputc (C : int; stream : FILEs) return int;
17743 function fputs (Strng : chars; Stream : FILEs) return int;
17760 function ftell (stream : FILEs) return long;
17767 function isatty (handle : int) return int;
17768 procedure mktemp (template : chars);
17769 -- The return value (which is just a pointer to template)
17771 procedure rewind (stream : FILEs);
17772 function rmtmp return int;
17780 function tmpfile return FILEs;
17781 function ungetc (c : int; stream : FILEs) return int;
17782 function unlink (filename : chars) return int;
17783 ---------------------
17784 -- Extra functions --
17785 ---------------------
17786 -- These functions supply slightly thicker bindings than
17787 -- those above. They are derived from functions in the
17788 -- C Run-Time Library, but may do a bit more work than
17789 -- just directly calling one of the Library functions.
17790 function is_regular_file (handle : int) return int;
17791 -- Tests if given handle is for a regular file (result 1)
17792 -- or for a non-regular file (pipe or device, result 0).
17793 ---------------------------------
17794 -- Control of Text/Binary Mode --
17795 ---------------------------------
17796 -- If text_translation_required is true, then the following
17797 -- functions may be used to dynamically switch a file from
17798 -- binary to text mode or vice versa. These functions have
17799 -- no effect if text_translation_required is false (i.e.@: in
17800 -- normal UNIX mode). Use fileno to get a stream handle.
17801 procedure set_binary_mode (handle : int);
17802 procedure set_text_mode (handle : int);
17803 ----------------------------
17804 -- Full Path Name support --
17805 ----------------------------
17806 procedure full_name (nam : chars; buffer : chars);
17807 -- Given a NUL terminated string representing a file
17808 -- name, returns in buffer a NUL terminated string
17809 -- representing the full path name for the file name.
17810 -- On systems where it is relevant the drive is also
17811 -- part of the full path name. It is the responsibility
17812 -- of the caller to pass an actual parameter for buffer
17813 -- that is big enough for any full path name. Use
17814 -- max_path_len given below as the size of buffer.
17815 max_path_len : integer;
17816 -- Maximum length of an allowable full path name on the
17817 -- system, including a terminating NUL character.
17818 end Interfaces.C_Streams;
17821 @node Interfacing to C Streams
17822 @section Interfacing to C Streams
17825 The packages in this section permit interfacing Ada files to C Stream
17828 @smallexample @c ada
17829 with Interfaces.C_Streams;
17830 package Ada.Sequential_IO.C_Streams is
17831 function C_Stream (F : File_Type)
17832 return Interfaces.C_Streams.FILEs;
17834 (File : in out File_Type;
17835 Mode : in File_Mode;
17836 C_Stream : in Interfaces.C_Streams.FILEs;
17837 Form : in String := "");
17838 end Ada.Sequential_IO.C_Streams;
17840 with Interfaces.C_Streams;
17841 package Ada.Direct_IO.C_Streams is
17842 function C_Stream (F : File_Type)
17843 return Interfaces.C_Streams.FILEs;
17845 (File : in out File_Type;
17846 Mode : in File_Mode;
17847 C_Stream : in Interfaces.C_Streams.FILEs;
17848 Form : in String := "");
17849 end Ada.Direct_IO.C_Streams;
17851 with Interfaces.C_Streams;
17852 package Ada.Text_IO.C_Streams is
17853 function C_Stream (F : File_Type)
17854 return Interfaces.C_Streams.FILEs;
17856 (File : in out File_Type;
17857 Mode : in File_Mode;
17858 C_Stream : in Interfaces.C_Streams.FILEs;
17859 Form : in String := "");
17860 end Ada.Text_IO.C_Streams;
17862 with Interfaces.C_Streams;
17863 package Ada.Wide_Text_IO.C_Streams is
17864 function C_Stream (F : File_Type)
17865 return Interfaces.C_Streams.FILEs;
17867 (File : in out File_Type;
17868 Mode : in File_Mode;
17869 C_Stream : in Interfaces.C_Streams.FILEs;
17870 Form : in String := "");
17871 end Ada.Wide_Text_IO.C_Streams;
17873 with Interfaces.C_Streams;
17874 package Ada.Wide_Wide_Text_IO.C_Streams is
17875 function C_Stream (F : File_Type)
17876 return Interfaces.C_Streams.FILEs;
17878 (File : in out File_Type;
17879 Mode : in File_Mode;
17880 C_Stream : in Interfaces.C_Streams.FILEs;
17881 Form : in String := "");
17882 end Ada.Wide_Wide_Text_IO.C_Streams;
17884 with Interfaces.C_Streams;
17885 package Ada.Stream_IO.C_Streams is
17886 function C_Stream (F : File_Type)
17887 return Interfaces.C_Streams.FILEs;
17889 (File : in out File_Type;
17890 Mode : in File_Mode;
17891 C_Stream : in Interfaces.C_Streams.FILEs;
17892 Form : in String := "");
17893 end Ada.Stream_IO.C_Streams;
17897 In each of these six packages, the @code{C_Stream} function obtains the
17898 @code{FILE} pointer from a currently opened Ada file. It is then
17899 possible to use the @code{Interfaces.C_Streams} package to operate on
17900 this stream, or the stream can be passed to a C program which can
17901 operate on it directly. Of course the program is responsible for
17902 ensuring that only appropriate sequences of operations are executed.
17904 One particular use of relevance to an Ada program is that the
17905 @code{setvbuf} function can be used to control the buffering of the
17906 stream used by an Ada file. In the absence of such a call the standard
17907 default buffering is used.
17909 The @code{Open} procedures in these packages open a file giving an
17910 existing C Stream instead of a file name. Typically this stream is
17911 imported from a C program, allowing an Ada file to operate on an
17914 @node The GNAT Library
17915 @chapter The GNAT Library
17918 The GNAT library contains a number of general and special purpose packages.
17919 It represents functionality that the GNAT developers have found useful, and
17920 which is made available to GNAT users. The packages described here are fully
17921 supported, and upwards compatibility will be maintained in future releases,
17922 so you can use these facilities with the confidence that the same functionality
17923 will be available in future releases.
17925 The chapter here simply gives a brief summary of the facilities available.
17926 The full documentation is found in the spec file for the package. The full
17927 sources of these library packages, including both spec and body, are provided
17928 with all GNAT releases. For example, to find out the full specifications of
17929 the SPITBOL pattern matching capability, including a full tutorial and
17930 extensive examples, look in the @file{g-spipat.ads} file in the library.
17932 For each entry here, the package name (as it would appear in a @code{with}
17933 clause) is given, followed by the name of the corresponding spec file in
17934 parentheses. The packages are children in four hierarchies, @code{Ada},
17935 @code{Interfaces}, @code{System}, and @code{GNAT}, the latter being a
17936 GNAT-specific hierarchy.
17938 Note that an application program should only use packages in one of these
17939 four hierarchies if the package is defined in the Ada Reference Manual,
17940 or is listed in this section of the GNAT Programmers Reference Manual.
17941 All other units should be considered internal implementation units and
17942 should not be directly @code{with}'ed by application code. The use of
17943 a @code{with} statement that references one of these internal implementation
17944 units makes an application potentially dependent on changes in versions
17945 of GNAT, and will generate a warning message.
17948 * Ada.Characters.Latin_9 (a-chlat9.ads)::
17949 * Ada.Characters.Wide_Latin_1 (a-cwila1.ads)::
17950 * Ada.Characters.Wide_Latin_9 (a-cwila9.ads)::
17951 * Ada.Characters.Wide_Wide_Latin_1 (a-chzla1.ads)::
17952 * Ada.Characters.Wide_Wide_Latin_9 (a-chzla9.ads)::
17953 * Ada.Containers.Formal_Doubly_Linked_Lists (a-cfdlli.ads)::
17954 * Ada.Containers.Formal_Hashed_Maps (a-cfhama.ads)::
17955 * Ada.Containers.Formal_Hashed_Sets (a-cfhase.ads)::
17956 * Ada.Containers.Formal_Ordered_Maps (a-cforma.ads)::
17957 * Ada.Containers.Formal_Ordered_Sets (a-cforse.ads)::
17958 * Ada.Containers.Formal_Vectors (a-cofove.ads)::
17959 * Ada.Command_Line.Environment (a-colien.ads)::
17960 * Ada.Command_Line.Remove (a-colire.ads)::
17961 * Ada.Command_Line.Response_File (a-clrefi.ads)::
17962 * Ada.Direct_IO.C_Streams (a-diocst.ads)::
17963 * Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)::
17964 * Ada.Exceptions.Last_Chance_Handler (a-elchha.ads)::
17965 * Ada.Exceptions.Traceback (a-exctra.ads)::
17966 * Ada.Sequential_IO.C_Streams (a-siocst.ads)::
17967 * Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)::
17968 * Ada.Strings.Unbounded.Text_IO (a-suteio.ads)::
17969 * Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)::
17970 * Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO (a-szuzti.ads)::
17971 * Ada.Text_IO.C_Streams (a-tiocst.ads)::
17972 * Ada.Text_IO.Reset_Standard_Files (a-tirsfi.ads)::
17973 * Ada.Wide_Characters.Unicode (a-wichun.ads)::
17974 * Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)::
17975 * Ada.Wide_Text_IO.Reset_Standard_Files (a-wrstfi.ads)::
17976 * Ada.Wide_Wide_Characters.Unicode (a-zchuni.ads)::
17977 * Ada.Wide_Wide_Text_IO.C_Streams (a-ztcstr.ads)::
17978 * Ada.Wide_Wide_Text_IO.Reset_Standard_Files (a-zrstfi.ads)::
17979 * GNAT.Altivec (g-altive.ads)::
17980 * GNAT.Altivec.Conversions (g-altcon.ads)::
17981 * GNAT.Altivec.Vector_Operations (g-alveop.ads)::
17982 * GNAT.Altivec.Vector_Types (g-alvety.ads)::
17983 * GNAT.Altivec.Vector_Views (g-alvevi.ads)::
17984 * GNAT.Array_Split (g-arrspl.ads)::
17985 * GNAT.AWK (g-awk.ads)::
17986 * GNAT.Bounded_Buffers (g-boubuf.ads)::
17987 * GNAT.Bounded_Mailboxes (g-boumai.ads)::
17988 * GNAT.Bubble_Sort (g-bubsor.ads)::
17989 * GNAT.Bubble_Sort_A (g-busora.ads)::
17990 * GNAT.Bubble_Sort_G (g-busorg.ads)::
17991 * GNAT.Byte_Order_Mark (g-byorma.ads)::
17992 * GNAT.Byte_Swapping (g-bytswa.ads)::
17993 * GNAT.Calendar (g-calend.ads)::
17994 * GNAT.Calendar.Time_IO (g-catiio.ads)::
17995 * GNAT.Case_Util (g-casuti.ads)::
17996 * GNAT.CGI (g-cgi.ads)::
17997 * GNAT.CGI.Cookie (g-cgicoo.ads)::
17998 * GNAT.CGI.Debug (g-cgideb.ads)::
17999 * GNAT.Command_Line (g-comlin.ads)::
18000 * GNAT.Compiler_Version (g-comver.ads)::
18001 * GNAT.Ctrl_C (g-ctrl_c.ads)::
18002 * GNAT.CRC32 (g-crc32.ads)::
18003 * GNAT.Current_Exception (g-curexc.ads)::
18004 * GNAT.Debug_Pools (g-debpoo.ads)::
18005 * GNAT.Debug_Utilities (g-debuti.ads)::
18006 * GNAT.Decode_String (g-decstr.ads)::
18007 * GNAT.Decode_UTF8_String (g-deutst.ads)::
18008 * GNAT.Directory_Operations (g-dirope.ads)::
18009 * GNAT.Directory_Operations.Iteration (g-diopit.ads)::
18010 * GNAT.Dynamic_HTables (g-dynhta.ads)::
18011 * GNAT.Dynamic_Tables (g-dyntab.ads)::
18012 * GNAT.Encode_String (g-encstr.ads)::
18013 * GNAT.Encode_UTF8_String (g-enutst.ads)::
18014 * GNAT.Exception_Actions (g-excact.ads)::
18015 * GNAT.Exception_Traces (g-exctra.ads)::
18016 * GNAT.Exceptions (g-except.ads)::
18017 * GNAT.Expect (g-expect.ads)::
18018 * GNAT.Expect.TTY (g-exptty.ads)::
18019 * GNAT.Float_Control (g-flocon.ads)::
18020 * GNAT.Heap_Sort (g-heasor.ads)::
18021 * GNAT.Heap_Sort_A (g-hesora.ads)::
18022 * GNAT.Heap_Sort_G (g-hesorg.ads)::
18023 * GNAT.HTable (g-htable.ads)::
18024 * GNAT.IO (g-io.ads)::
18025 * GNAT.IO_Aux (g-io_aux.ads)::
18026 * GNAT.Lock_Files (g-locfil.ads)::
18027 * GNAT.MBBS_Discrete_Random (g-mbdira.ads)::
18028 * GNAT.MBBS_Float_Random (g-mbflra.ads)::
18029 * GNAT.MD5 (g-md5.ads)::
18030 * GNAT.Memory_Dump (g-memdum.ads)::
18031 * GNAT.Most_Recent_Exception (g-moreex.ads)::
18032 * GNAT.OS_Lib (g-os_lib.ads)::
18033 * GNAT.Perfect_Hash_Generators (g-pehage.ads)::
18034 * GNAT.Random_Numbers (g-rannum.ads)::
18035 * GNAT.Regexp (g-regexp.ads)::
18036 * GNAT.Registry (g-regist.ads)::
18037 * GNAT.Regpat (g-regpat.ads)::
18038 * GNAT.Secondary_Stack_Info (g-sestin.ads)::
18039 * GNAT.Semaphores (g-semaph.ads)::
18040 * GNAT.Serial_Communications (g-sercom.ads)::
18041 * GNAT.SHA1 (g-sha1.ads)::
18042 * GNAT.SHA224 (g-sha224.ads)::
18043 * GNAT.SHA256 (g-sha256.ads)::
18044 * GNAT.SHA384 (g-sha384.ads)::
18045 * GNAT.SHA512 (g-sha512.ads)::
18046 * GNAT.Signals (g-signal.ads)::
18047 * GNAT.Sockets (g-socket.ads)::
18048 * GNAT.Source_Info (g-souinf.ads)::
18049 * GNAT.Spelling_Checker (g-speche.ads)::
18050 * GNAT.Spelling_Checker_Generic (g-spchge.ads)::
18051 * GNAT.Spitbol.Patterns (g-spipat.ads)::
18052 * GNAT.Spitbol (g-spitbo.ads)::
18053 * GNAT.Spitbol.Table_Boolean (g-sptabo.ads)::
18054 * GNAT.Spitbol.Table_Integer (g-sptain.ads)::
18055 * GNAT.Spitbol.Table_VString (g-sptavs.ads)::
18056 * GNAT.SSE (g-sse.ads)::
18057 * GNAT.SSE.Vector_Types (g-ssvety.ads)::
18058 * GNAT.Strings (g-string.ads)::
18059 * GNAT.String_Split (g-strspl.ads)::
18060 * GNAT.Table (g-table.ads)::
18061 * GNAT.Task_Lock (g-tasloc.ads)::
18062 * GNAT.Threads (g-thread.ads)::
18063 * GNAT.Time_Stamp (g-timsta.ads)::
18064 * GNAT.Traceback (g-traceb.ads)::
18065 * GNAT.Traceback.Symbolic (g-trasym.ads)::
18066 * GNAT.UTF_32 (g-utf_32.ads)::
18067 * GNAT.UTF_32_Spelling_Checker (g-u3spch.ads)::
18068 * GNAT.Wide_Spelling_Checker (g-wispch.ads)::
18069 * GNAT.Wide_String_Split (g-wistsp.ads)::
18070 * GNAT.Wide_Wide_Spelling_Checker (g-zspche.ads)::
18071 * GNAT.Wide_Wide_String_Split (g-zistsp.ads)::
18072 * Interfaces.C.Extensions (i-cexten.ads)::
18073 * Interfaces.C.Streams (i-cstrea.ads)::
18074 * Interfaces.CPP (i-cpp.ads)::
18075 * Interfaces.Packed_Decimal (i-pacdec.ads)::
18076 * Interfaces.VxWorks (i-vxwork.ads)::
18077 * Interfaces.VxWorks.IO (i-vxwoio.ads)::
18078 * System.Address_Image (s-addima.ads)::
18079 * System.Assertions (s-assert.ads)::
18080 * System.Memory (s-memory.ads)::
18081 * System.Multiprocessors (s-multip.ads)::
18082 * System.Multiprocessors.Dispatching_Domains (s-mudido.ads)::
18083 * System.Partition_Interface (s-parint.ads)::
18084 * System.Pool_Global (s-pooglo.ads)::
18085 * System.Pool_Local (s-pooloc.ads)::
18086 * System.Restrictions (s-restri.ads)::
18087 * System.Rident (s-rident.ads)::
18088 * System.Strings.Stream_Ops (s-ststop.ads)::
18089 * System.Task_Info (s-tasinf.ads)::
18090 * System.Wch_Cnv (s-wchcnv.ads)::
18091 * System.Wch_Con (s-wchcon.ads)::
18094 @node Ada.Characters.Latin_9 (a-chlat9.ads)
18095 @section @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
18096 @cindex @code{Ada.Characters.Latin_9} (@file{a-chlat9.ads})
18097 @cindex Latin_9 constants for Character
18100 This child of @code{Ada.Characters}
18101 provides a set of definitions corresponding to those in the
18102 RM-defined package @code{Ada.Characters.Latin_1} but with the
18103 few modifications required for @code{Latin-9}
18104 The provision of such a package
18105 is specifically authorized by the Ada Reference Manual
18108 @node Ada.Characters.Wide_Latin_1 (a-cwila1.ads)
18109 @section @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
18110 @cindex @code{Ada.Characters.Wide_Latin_1} (@file{a-cwila1.ads})
18111 @cindex Latin_1 constants for Wide_Character
18114 This child of @code{Ada.Characters}
18115 provides a set of definitions corresponding to those in the
18116 RM-defined package @code{Ada.Characters.Latin_1} but with the
18117 types of the constants being @code{Wide_Character}
18118 instead of @code{Character}. The provision of such a package
18119 is specifically authorized by the Ada Reference Manual
18122 @node Ada.Characters.Wide_Latin_9 (a-cwila9.ads)
18123 @section @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
18124 @cindex @code{Ada.Characters.Wide_Latin_9} (@file{a-cwila1.ads})
18125 @cindex Latin_9 constants for Wide_Character
18128 This child of @code{Ada.Characters}
18129 provides a set of definitions corresponding to those in the
18130 GNAT defined package @code{Ada.Characters.Latin_9} but with the
18131 types of the constants being @code{Wide_Character}
18132 instead of @code{Character}. The provision of such a package
18133 is specifically authorized by the Ada Reference Manual
18136 @node Ada.Characters.Wide_Wide_Latin_1 (a-chzla1.ads)
18137 @section @code{Ada.Characters.Wide_Wide_Latin_1} (@file{a-chzla1.ads})
18138 @cindex @code{Ada.Characters.Wide_Wide_Latin_1} (@file{a-chzla1.ads})
18139 @cindex Latin_1 constants for Wide_Wide_Character
18142 This child of @code{Ada.Characters}
18143 provides a set of definitions corresponding to those in the
18144 RM-defined package @code{Ada.Characters.Latin_1} but with the
18145 types of the constants being @code{Wide_Wide_Character}
18146 instead of @code{Character}. The provision of such a package
18147 is specifically authorized by the Ada Reference Manual
18150 @node Ada.Characters.Wide_Wide_Latin_9 (a-chzla9.ads)
18151 @section @code{Ada.Characters.Wide_Wide_Latin_9} (@file{a-chzla9.ads})
18152 @cindex @code{Ada.Characters.Wide_Wide_Latin_9} (@file{a-chzla9.ads})
18153 @cindex Latin_9 constants for Wide_Wide_Character
18156 This child of @code{Ada.Characters}
18157 provides a set of definitions corresponding to those in the
18158 GNAT defined package @code{Ada.Characters.Latin_9} but with the
18159 types of the constants being @code{Wide_Wide_Character}
18160 instead of @code{Character}. The provision of such a package
18161 is specifically authorized by the Ada Reference Manual
18164 @node Ada.Containers.Formal_Doubly_Linked_Lists (a-cfdlli.ads)
18165 @section @code{Ada.Containers.Formal_Doubly_Linked_Lists} (@file{a-cfdlli.ads})
18166 @cindex @code{Ada.Containers.Formal_Doubly_Linked_Lists} (@file{a-cfdlli.ads})
18167 @cindex Formal container for doubly linked lists
18170 This child of @code{Ada.Containers} defines a modified version of the
18171 Ada 2005 container for doubly linked lists, meant to facilitate formal
18172 verification of code using such containers. The specification of this
18173 unit is compatible with SPARK 2014.
18175 Note that although this container was designed with formal verification
18176 in mind, it may well be generally useful in that it is a simplified more
18177 efficient version than the one defined in the standard. In particular it
18178 does not have the complex overhead required to detect cursor tampering.
18180 @node Ada.Containers.Formal_Hashed_Maps (a-cfhama.ads)
18181 @section @code{Ada.Containers.Formal_Hashed_Maps} (@file{a-cfhama.ads})
18182 @cindex @code{Ada.Containers.Formal_Hashed_Maps} (@file{a-cfhama.ads})
18183 @cindex Formal container for hashed maps
18186 This child of @code{Ada.Containers} defines a modified version of the
18187 Ada 2005 container for hashed maps, meant to facilitate formal
18188 verification of code using such containers. The specification of this
18189 unit is compatible with SPARK 2014.
18191 Note that although this container was designed with formal verification
18192 in mind, it may well be generally useful in that it is a simplified more
18193 efficient version than the one defined in the standard. In particular it
18194 does not have the complex overhead required to detect cursor tampering.
18196 @node Ada.Containers.Formal_Hashed_Sets (a-cfhase.ads)
18197 @section @code{Ada.Containers.Formal_Hashed_Sets} (@file{a-cfhase.ads})
18198 @cindex @code{Ada.Containers.Formal_Hashed_Sets} (@file{a-cfhase.ads})
18199 @cindex Formal container for hashed sets
18202 This child of @code{Ada.Containers} defines a modified version of the
18203 Ada 2005 container for hashed sets, meant to facilitate formal
18204 verification of code using such containers. The specification of this
18205 unit is compatible with SPARK 2014.
18207 Note that although this container was designed with formal verification
18208 in mind, it may well be generally useful in that it is a simplified more
18209 efficient version than the one defined in the standard. In particular it
18210 does not have the complex overhead required to detect cursor tampering.
18212 @node Ada.Containers.Formal_Ordered_Maps (a-cforma.ads)
18213 @section @code{Ada.Containers.Formal_Ordered_Maps} (@file{a-cforma.ads})
18214 @cindex @code{Ada.Containers.Formal_Ordered_Maps} (@file{a-cforma.ads})
18215 @cindex Formal container for ordered maps
18218 This child of @code{Ada.Containers} defines a modified version of the
18219 Ada 2005 container for ordered maps, meant to facilitate formal
18220 verification of code using such containers. The specification of this
18221 unit is compatible with SPARK 2014.
18223 Note that although this container was designed with formal verification
18224 in mind, it may well be generally useful in that it is a simplified more
18225 efficient version than the one defined in the standard. In particular it
18226 does not have the complex overhead required to detect cursor tampering.
18228 @node Ada.Containers.Formal_Ordered_Sets (a-cforse.ads)
18229 @section @code{Ada.Containers.Formal_Ordered_Sets} (@file{a-cforse.ads})
18230 @cindex @code{Ada.Containers.Formal_Ordered_Sets} (@file{a-cforse.ads})
18231 @cindex Formal container for ordered sets
18234 This child of @code{Ada.Containers} defines a modified version of the
18235 Ada 2005 container for ordered sets, meant to facilitate formal
18236 verification of code using such containers. The specification of this
18237 unit is compatible with SPARK 2014.
18239 Note that although this container was designed with formal verification
18240 in mind, it may well be generally useful in that it is a simplified more
18241 efficient version than the one defined in the standard. In particular it
18242 does not have the complex overhead required to detect cursor tampering.
18244 @node Ada.Containers.Formal_Vectors (a-cofove.ads)
18245 @section @code{Ada.Containers.Formal_Vectors} (@file{a-cofove.ads})
18246 @cindex @code{Ada.Containers.Formal_Vectors} (@file{a-cofove.ads})
18247 @cindex Formal container for vectors
18250 This child of @code{Ada.Containers} defines a modified version of the
18251 Ada 2005 container for vectors, meant to facilitate formal
18252 verification of code using such containers. The specification of this
18253 unit is compatible with SPARK 2014.
18255 Note that although this container was designed with formal verification
18256 in mind, it may well be generally useful in that it is a simplified more
18257 efficient version than the one defined in the standard. In particular it
18258 does not have the complex overhead required to detect cursor tampering.
18260 @node Ada.Command_Line.Environment (a-colien.ads)
18261 @section @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
18262 @cindex @code{Ada.Command_Line.Environment} (@file{a-colien.ads})
18263 @cindex Environment entries
18266 This child of @code{Ada.Command_Line}
18267 provides a mechanism for obtaining environment values on systems
18268 where this concept makes sense.
18270 @node Ada.Command_Line.Remove (a-colire.ads)
18271 @section @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
18272 @cindex @code{Ada.Command_Line.Remove} (@file{a-colire.ads})
18273 @cindex Removing command line arguments
18274 @cindex Command line, argument removal
18277 This child of @code{Ada.Command_Line}
18278 provides a mechanism for logically removing
18279 arguments from the argument list. Once removed, an argument is not visible
18280 to further calls on the subprograms in @code{Ada.Command_Line} will not
18281 see the removed argument.
18283 @node Ada.Command_Line.Response_File (a-clrefi.ads)
18284 @section @code{Ada.Command_Line.Response_File} (@file{a-clrefi.ads})
18285 @cindex @code{Ada.Command_Line.Response_File} (@file{a-clrefi.ads})
18286 @cindex Response file for command line
18287 @cindex Command line, response file
18288 @cindex Command line, handling long command lines
18291 This child of @code{Ada.Command_Line} provides a mechanism facilities for
18292 getting command line arguments from a text file, called a "response file".
18293 Using a response file allow passing a set of arguments to an executable longer
18294 than the maximum allowed by the system on the command line.
18296 @node Ada.Direct_IO.C_Streams (a-diocst.ads)
18297 @section @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
18298 @cindex @code{Ada.Direct_IO.C_Streams} (@file{a-diocst.ads})
18299 @cindex C Streams, Interfacing with Direct_IO
18302 This package provides subprograms that allow interfacing between
18303 C streams and @code{Direct_IO}. The stream identifier can be
18304 extracted from a file opened on the Ada side, and an Ada file
18305 can be constructed from a stream opened on the C side.
18307 @node Ada.Exceptions.Is_Null_Occurrence (a-einuoc.ads)
18308 @section @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
18309 @cindex @code{Ada.Exceptions.Is_Null_Occurrence} (@file{a-einuoc.ads})
18310 @cindex Null_Occurrence, testing for
18313 This child subprogram provides a way of testing for the null
18314 exception occurrence (@code{Null_Occurrence}) without raising
18317 @node Ada.Exceptions.Last_Chance_Handler (a-elchha.ads)
18318 @section @code{Ada.Exceptions.Last_Chance_Handler} (@file{a-elchha.ads})
18319 @cindex @code{Ada.Exceptions.Last_Chance_Handler} (@file{a-elchha.ads})
18320 @cindex Null_Occurrence, testing for
18323 This child subprogram is used for handling otherwise unhandled
18324 exceptions (hence the name last chance), and perform clean ups before
18325 terminating the program. Note that this subprogram never returns.
18327 @node Ada.Exceptions.Traceback (a-exctra.ads)
18328 @section @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
18329 @cindex @code{Ada.Exceptions.Traceback} (@file{a-exctra.ads})
18330 @cindex Traceback for Exception Occurrence
18333 This child package provides the subprogram (@code{Tracebacks}) to
18334 give a traceback array of addresses based on an exception
18337 @node Ada.Sequential_IO.C_Streams (a-siocst.ads)
18338 @section @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
18339 @cindex @code{Ada.Sequential_IO.C_Streams} (@file{a-siocst.ads})
18340 @cindex C Streams, Interfacing with Sequential_IO
18343 This package provides subprograms that allow interfacing between
18344 C streams and @code{Sequential_IO}. The stream identifier can be
18345 extracted from a file opened on the Ada side, and an Ada file
18346 can be constructed from a stream opened on the C side.
18348 @node Ada.Streams.Stream_IO.C_Streams (a-ssicst.ads)
18349 @section @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
18350 @cindex @code{Ada.Streams.Stream_IO.C_Streams} (@file{a-ssicst.ads})
18351 @cindex C Streams, Interfacing with Stream_IO
18354 This package provides subprograms that allow interfacing between
18355 C streams and @code{Stream_IO}. The stream identifier can be
18356 extracted from a file opened on the Ada side, and an Ada file
18357 can be constructed from a stream opened on the C side.
18359 @node Ada.Strings.Unbounded.Text_IO (a-suteio.ads)
18360 @section @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
18361 @cindex @code{Ada.Strings.Unbounded.Text_IO} (@file{a-suteio.ads})
18362 @cindex @code{Unbounded_String}, IO support
18363 @cindex @code{Text_IO}, extensions for unbounded strings
18366 This package provides subprograms for Text_IO for unbounded
18367 strings, avoiding the necessity for an intermediate operation
18368 with ordinary strings.
18370 @node Ada.Strings.Wide_Unbounded.Wide_Text_IO (a-swuwti.ads)
18371 @section @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
18372 @cindex @code{Ada.Strings.Wide_Unbounded.Wide_Text_IO} (@file{a-swuwti.ads})
18373 @cindex @code{Unbounded_Wide_String}, IO support
18374 @cindex @code{Text_IO}, extensions for unbounded wide strings
18377 This package provides subprograms for Text_IO for unbounded
18378 wide strings, avoiding the necessity for an intermediate operation
18379 with ordinary wide strings.
18381 @node Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO (a-szuzti.ads)
18382 @section @code{Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO} (@file{a-szuzti.ads})
18383 @cindex @code{Ada.Strings.Wide_Wide_Unbounded.Wide_Wide_Text_IO} (@file{a-szuzti.ads})
18384 @cindex @code{Unbounded_Wide_Wide_String}, IO support
18385 @cindex @code{Text_IO}, extensions for unbounded wide wide strings
18388 This package provides subprograms for Text_IO for unbounded
18389 wide wide strings, avoiding the necessity for an intermediate operation
18390 with ordinary wide wide strings.
18392 @node Ada.Text_IO.C_Streams (a-tiocst.ads)
18393 @section @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
18394 @cindex @code{Ada.Text_IO.C_Streams} (@file{a-tiocst.ads})
18395 @cindex C Streams, Interfacing with @code{Text_IO}
18398 This package provides subprograms that allow interfacing between
18399 C streams and @code{Text_IO}. The stream identifier can be
18400 extracted from a file opened on the Ada side, and an Ada file
18401 can be constructed from a stream opened on the C side.
18403 @node Ada.Text_IO.Reset_Standard_Files (a-tirsfi.ads)
18404 @section @code{Ada.Text_IO.Reset_Standard_Files} (@file{a-tirsfi.ads})
18405 @cindex @code{Ada.Text_IO.Reset_Standard_Files} (@file{a-tirsfi.ads})
18406 @cindex @code{Text_IO} resetting standard files
18409 This procedure is used to reset the status of the standard files used
18410 by Ada.Text_IO. This is useful in a situation (such as a restart in an
18411 embedded application) where the status of the files may change during
18412 execution (for example a standard input file may be redefined to be
18415 @node Ada.Wide_Characters.Unicode (a-wichun.ads)
18416 @section @code{Ada.Wide_Characters.Unicode} (@file{a-wichun.ads})
18417 @cindex @code{Ada.Wide_Characters.Unicode} (@file{a-wichun.ads})
18418 @cindex Unicode categorization, Wide_Character
18421 This package provides subprograms that allow categorization of
18422 Wide_Character values according to Unicode categories.
18424 @node Ada.Wide_Text_IO.C_Streams (a-wtcstr.ads)
18425 @section @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
18426 @cindex @code{Ada.Wide_Text_IO.C_Streams} (@file{a-wtcstr.ads})
18427 @cindex C Streams, Interfacing with @code{Wide_Text_IO}
18430 This package provides subprograms that allow interfacing between
18431 C streams and @code{Wide_Text_IO}. The stream identifier can be
18432 extracted from a file opened on the Ada side, and an Ada file
18433 can be constructed from a stream opened on the C side.
18435 @node Ada.Wide_Text_IO.Reset_Standard_Files (a-wrstfi.ads)
18436 @section @code{Ada.Wide_Text_IO.Reset_Standard_Files} (@file{a-wrstfi.ads})
18437 @cindex @code{Ada.Wide_Text_IO.Reset_Standard_Files} (@file{a-wrstfi.ads})
18438 @cindex @code{Wide_Text_IO} resetting standard files
18441 This procedure is used to reset the status of the standard files used
18442 by Ada.Wide_Text_IO. This is useful in a situation (such as a restart in an
18443 embedded application) where the status of the files may change during
18444 execution (for example a standard input file may be redefined to be
18447 @node Ada.Wide_Wide_Characters.Unicode (a-zchuni.ads)
18448 @section @code{Ada.Wide_Wide_Characters.Unicode} (@file{a-zchuni.ads})
18449 @cindex @code{Ada.Wide_Wide_Characters.Unicode} (@file{a-zchuni.ads})
18450 @cindex Unicode categorization, Wide_Wide_Character
18453 This package provides subprograms that allow categorization of
18454 Wide_Wide_Character values according to Unicode categories.
18456 @node Ada.Wide_Wide_Text_IO.C_Streams (a-ztcstr.ads)
18457 @section @code{Ada.Wide_Wide_Text_IO.C_Streams} (@file{a-ztcstr.ads})
18458 @cindex @code{Ada.Wide_Wide_Text_IO.C_Streams} (@file{a-ztcstr.ads})
18459 @cindex C Streams, Interfacing with @code{Wide_Wide_Text_IO}
18462 This package provides subprograms that allow interfacing between
18463 C streams and @code{Wide_Wide_Text_IO}. The stream identifier can be
18464 extracted from a file opened on the Ada side, and an Ada file
18465 can be constructed from a stream opened on the C side.
18467 @node Ada.Wide_Wide_Text_IO.Reset_Standard_Files (a-zrstfi.ads)
18468 @section @code{Ada.Wide_Wide_Text_IO.Reset_Standard_Files} (@file{a-zrstfi.ads})
18469 @cindex @code{Ada.Wide_Wide_Text_IO.Reset_Standard_Files} (@file{a-zrstfi.ads})
18470 @cindex @code{Wide_Wide_Text_IO} resetting standard files
18473 This procedure is used to reset the status of the standard files used
18474 by Ada.Wide_Wide_Text_IO. This is useful in a situation (such as a
18475 restart in an embedded application) where the status of the files may
18476 change during execution (for example a standard input file may be
18477 redefined to be interactive).
18479 @node GNAT.Altivec (g-altive.ads)
18480 @section @code{GNAT.Altivec} (@file{g-altive.ads})
18481 @cindex @code{GNAT.Altivec} (@file{g-altive.ads})
18485 This is the root package of the GNAT AltiVec binding. It provides
18486 definitions of constants and types common to all the versions of the
18489 @node GNAT.Altivec.Conversions (g-altcon.ads)
18490 @section @code{GNAT.Altivec.Conversions} (@file{g-altcon.ads})
18491 @cindex @code{GNAT.Altivec.Conversions} (@file{g-altcon.ads})
18495 This package provides the Vector/View conversion routines.
18497 @node GNAT.Altivec.Vector_Operations (g-alveop.ads)
18498 @section @code{GNAT.Altivec.Vector_Operations} (@file{g-alveop.ads})
18499 @cindex @code{GNAT.Altivec.Vector_Operations} (@file{g-alveop.ads})
18503 This package exposes the Ada interface to the AltiVec operations on
18504 vector objects. A soft emulation is included by default in the GNAT
18505 library. The hard binding is provided as a separate package. This unit
18506 is common to both bindings.
18508 @node GNAT.Altivec.Vector_Types (g-alvety.ads)
18509 @section @code{GNAT.Altivec.Vector_Types} (@file{g-alvety.ads})
18510 @cindex @code{GNAT.Altivec.Vector_Types} (@file{g-alvety.ads})
18514 This package exposes the various vector types part of the Ada binding
18515 to AltiVec facilities.
18517 @node GNAT.Altivec.Vector_Views (g-alvevi.ads)
18518 @section @code{GNAT.Altivec.Vector_Views} (@file{g-alvevi.ads})
18519 @cindex @code{GNAT.Altivec.Vector_Views} (@file{g-alvevi.ads})
18523 This package provides public 'View' data types from/to which private
18524 vector representations can be converted via
18525 GNAT.Altivec.Conversions. This allows convenient access to individual
18526 vector elements and provides a simple way to initialize vector
18529 @node GNAT.Array_Split (g-arrspl.ads)
18530 @section @code{GNAT.Array_Split} (@file{g-arrspl.ads})
18531 @cindex @code{GNAT.Array_Split} (@file{g-arrspl.ads})
18532 @cindex Array splitter
18535 Useful array-manipulation routines: given a set of separators, split
18536 an array wherever the separators appear, and provide direct access
18537 to the resulting slices.
18539 @node GNAT.AWK (g-awk.ads)
18540 @section @code{GNAT.AWK} (@file{g-awk.ads})
18541 @cindex @code{GNAT.AWK} (@file{g-awk.ads})
18546 Provides AWK-like parsing functions, with an easy interface for parsing one
18547 or more files containing formatted data. The file is viewed as a database
18548 where each record is a line and a field is a data element in this line.
18550 @node GNAT.Bounded_Buffers (g-boubuf.ads)
18551 @section @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
18552 @cindex @code{GNAT.Bounded_Buffers} (@file{g-boubuf.ads})
18554 @cindex Bounded Buffers
18557 Provides a concurrent generic bounded buffer abstraction. Instances are
18558 useful directly or as parts of the implementations of other abstractions,
18561 @node GNAT.Bounded_Mailboxes (g-boumai.ads)
18562 @section @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
18563 @cindex @code{GNAT.Bounded_Mailboxes} (@file{g-boumai.ads})
18568 Provides a thread-safe asynchronous intertask mailbox communication facility.
18570 @node GNAT.Bubble_Sort (g-bubsor.ads)
18571 @section @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
18572 @cindex @code{GNAT.Bubble_Sort} (@file{g-bubsor.ads})
18574 @cindex Bubble sort
18577 Provides a general implementation of bubble sort usable for sorting arbitrary
18578 data items. Exchange and comparison procedures are provided by passing
18579 access-to-procedure values.
18581 @node GNAT.Bubble_Sort_A (g-busora.ads)
18582 @section @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
18583 @cindex @code{GNAT.Bubble_Sort_A} (@file{g-busora.ads})
18585 @cindex Bubble sort
18588 Provides a general implementation of bubble sort usable for sorting arbitrary
18589 data items. Move and comparison procedures are provided by passing
18590 access-to-procedure values. This is an older version, retained for
18591 compatibility. Usually @code{GNAT.Bubble_Sort} will be preferable.
18593 @node GNAT.Bubble_Sort_G (g-busorg.ads)
18594 @section @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
18595 @cindex @code{GNAT.Bubble_Sort_G} (@file{g-busorg.ads})
18597 @cindex Bubble sort
18600 Similar to @code{Bubble_Sort_A} except that the move and sorting procedures
18601 are provided as generic parameters, this improves efficiency, especially
18602 if the procedures can be inlined, at the expense of duplicating code for
18603 multiple instantiations.
18605 @node GNAT.Byte_Order_Mark (g-byorma.ads)
18606 @section @code{GNAT.Byte_Order_Mark} (@file{g-byorma.ads})
18607 @cindex @code{GNAT.Byte_Order_Mark} (@file{g-byorma.ads})
18608 @cindex UTF-8 representation
18609 @cindex Wide characte representations
18612 Provides a routine which given a string, reads the start of the string to
18613 see whether it is one of the standard byte order marks (BOM's) which signal
18614 the encoding of the string. The routine includes detection of special XML
18615 sequences for various UCS input formats.
18617 @node GNAT.Byte_Swapping (g-bytswa.ads)
18618 @section @code{GNAT.Byte_Swapping} (@file{g-bytswa.ads})
18619 @cindex @code{GNAT.Byte_Swapping} (@file{g-bytswa.ads})
18620 @cindex Byte swapping
18624 General routines for swapping the bytes in 2-, 4-, and 8-byte quantities.
18625 Machine-specific implementations are available in some cases.
18627 @node GNAT.Calendar (g-calend.ads)
18628 @section @code{GNAT.Calendar} (@file{g-calend.ads})
18629 @cindex @code{GNAT.Calendar} (@file{g-calend.ads})
18630 @cindex @code{Calendar}
18633 Extends the facilities provided by @code{Ada.Calendar} to include handling
18634 of days of the week, an extended @code{Split} and @code{Time_Of} capability.
18635 Also provides conversion of @code{Ada.Calendar.Time} values to and from the
18636 C @code{timeval} format.
18638 @node GNAT.Calendar.Time_IO (g-catiio.ads)
18639 @section @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
18640 @cindex @code{Calendar}
18642 @cindex @code{GNAT.Calendar.Time_IO} (@file{g-catiio.ads})
18644 @node GNAT.CRC32 (g-crc32.ads)
18645 @section @code{GNAT.CRC32} (@file{g-crc32.ads})
18646 @cindex @code{GNAT.CRC32} (@file{g-crc32.ads})
18648 @cindex Cyclic Redundancy Check
18651 This package implements the CRC-32 algorithm. For a full description
18652 of this algorithm see
18653 ``Computation of Cyclic Redundancy Checks via Table Look-Up'',
18654 @cite{Communications of the ACM}, Vol.@: 31 No.@: 8, pp.@: 1008-1013,
18655 Aug.@: 1988. Sarwate, D.V@.
18657 @node GNAT.Case_Util (g-casuti.ads)
18658 @section @code{GNAT.Case_Util} (@file{g-casuti.ads})
18659 @cindex @code{GNAT.Case_Util} (@file{g-casuti.ads})
18660 @cindex Casing utilities
18661 @cindex Character handling (@code{GNAT.Case_Util})
18664 A set of simple routines for handling upper and lower casing of strings
18665 without the overhead of the full casing tables
18666 in @code{Ada.Characters.Handling}.
18668 @node GNAT.CGI (g-cgi.ads)
18669 @section @code{GNAT.CGI} (@file{g-cgi.ads})
18670 @cindex @code{GNAT.CGI} (@file{g-cgi.ads})
18671 @cindex CGI (Common Gateway Interface)
18674 This is a package for interfacing a GNAT program with a Web server via the
18675 Common Gateway Interface (CGI)@. Basically this package parses the CGI
18676 parameters, which are a set of key/value pairs sent by the Web server. It
18677 builds a table whose index is the key and provides some services to deal
18680 @node GNAT.CGI.Cookie (g-cgicoo.ads)
18681 @section @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
18682 @cindex @code{GNAT.CGI.Cookie} (@file{g-cgicoo.ads})
18683 @cindex CGI (Common Gateway Interface) cookie support
18684 @cindex Cookie support in CGI
18687 This is a package to interface a GNAT program with a Web server via the
18688 Common Gateway Interface (CGI). It exports services to deal with Web
18689 cookies (piece of information kept in the Web client software).
18691 @node GNAT.CGI.Debug (g-cgideb.ads)
18692 @section @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
18693 @cindex @code{GNAT.CGI.Debug} (@file{g-cgideb.ads})
18694 @cindex CGI (Common Gateway Interface) debugging
18697 This is a package to help debugging CGI (Common Gateway Interface)
18698 programs written in Ada.
18700 @node GNAT.Command_Line (g-comlin.ads)
18701 @section @code{GNAT.Command_Line} (@file{g-comlin.ads})
18702 @cindex @code{GNAT.Command_Line} (@file{g-comlin.ads})
18703 @cindex Command line
18706 Provides a high level interface to @code{Ada.Command_Line} facilities,
18707 including the ability to scan for named switches with optional parameters
18708 and expand file names using wild card notations.
18710 @node GNAT.Compiler_Version (g-comver.ads)
18711 @section @code{GNAT.Compiler_Version} (@file{g-comver.ads})
18712 @cindex @code{GNAT.Compiler_Version} (@file{g-comver.ads})
18713 @cindex Compiler Version
18714 @cindex Version, of compiler
18717 Provides a routine for obtaining the version of the compiler used to
18718 compile the program. More accurately this is the version of the binder
18719 used to bind the program (this will normally be the same as the version
18720 of the compiler if a consistent tool set is used to compile all units
18723 @node GNAT.Ctrl_C (g-ctrl_c.ads)
18724 @section @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
18725 @cindex @code{GNAT.Ctrl_C} (@file{g-ctrl_c.ads})
18729 Provides a simple interface to handle Ctrl-C keyboard events.
18731 @node GNAT.Current_Exception (g-curexc.ads)
18732 @section @code{GNAT.Current_Exception} (@file{g-curexc.ads})
18733 @cindex @code{GNAT.Current_Exception} (@file{g-curexc.ads})
18734 @cindex Current exception
18735 @cindex Exception retrieval
18738 Provides access to information on the current exception that has been raised
18739 without the need for using the Ada 95 / Ada 2005 exception choice parameter
18740 specification syntax.
18741 This is particularly useful in simulating typical facilities for
18742 obtaining information about exceptions provided by Ada 83 compilers.
18744 @node GNAT.Debug_Pools (g-debpoo.ads)
18745 @section @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
18746 @cindex @code{GNAT.Debug_Pools} (@file{g-debpoo.ads})
18748 @cindex Debug pools
18749 @cindex Memory corruption debugging
18752 Provide a debugging storage pools that helps tracking memory corruption
18753 problems. @xref{The GNAT Debug Pool Facility,,, gnat_ugn,
18754 @value{EDITION} User's Guide}.
18756 @node GNAT.Debug_Utilities (g-debuti.ads)
18757 @section @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
18758 @cindex @code{GNAT.Debug_Utilities} (@file{g-debuti.ads})
18762 Provides a few useful utilities for debugging purposes, including conversion
18763 to and from string images of address values. Supports both C and Ada formats
18764 for hexadecimal literals.
18766 @node GNAT.Decode_String (g-decstr.ads)
18767 @section @code{GNAT.Decode_String} (@file{g-decstr.ads})
18768 @cindex @code{GNAT.Decode_String} (@file{g-decstr.ads})
18769 @cindex Decoding strings
18770 @cindex String decoding
18771 @cindex Wide character encoding
18776 A generic package providing routines for decoding wide character and wide wide
18777 character strings encoded as sequences of 8-bit characters using a specified
18778 encoding method. Includes validation routines, and also routines for stepping
18779 to next or previous encoded character in an encoded string.
18780 Useful in conjunction with Unicode character coding. Note there is a
18781 preinstantiation for UTF-8. See next entry.
18783 @node GNAT.Decode_UTF8_String (g-deutst.ads)
18784 @section @code{GNAT.Decode_UTF8_String} (@file{g-deutst.ads})
18785 @cindex @code{GNAT.Decode_UTF8_String} (@file{g-deutst.ads})
18786 @cindex Decoding strings
18787 @cindex Decoding UTF-8 strings
18788 @cindex UTF-8 string decoding
18789 @cindex Wide character decoding
18794 A preinstantiation of GNAT.Decode_Strings for UTF-8 encoding.
18796 @node GNAT.Directory_Operations (g-dirope.ads)
18797 @section @code{GNAT.Directory_Operations} (@file{g-dirope.ads})
18798 @cindex @code{GNAT.Directory_Operations} (@file{g-dirope.ads})
18799 @cindex Directory operations
18802 Provides a set of routines for manipulating directories, including changing
18803 the current directory, making new directories, and scanning the files in a
18806 @node GNAT.Directory_Operations.Iteration (g-diopit.ads)
18807 @section @code{GNAT.Directory_Operations.Iteration} (@file{g-diopit.ads})
18808 @cindex @code{GNAT.Directory_Operations.Iteration} (@file{g-diopit.ads})
18809 @cindex Directory operations iteration
18812 A child unit of GNAT.Directory_Operations providing additional operations
18813 for iterating through directories.
18815 @node GNAT.Dynamic_HTables (g-dynhta.ads)
18816 @section @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
18817 @cindex @code{GNAT.Dynamic_HTables} (@file{g-dynhta.ads})
18818 @cindex Hash tables
18821 A generic implementation of hash tables that can be used to hash arbitrary
18822 data. Provided in two forms, a simple form with built in hash functions,
18823 and a more complex form in which the hash function is supplied.
18826 This package provides a facility similar to that of @code{GNAT.HTable},
18827 except that this package declares a type that can be used to define
18828 dynamic instances of the hash table, while an instantiation of
18829 @code{GNAT.HTable} creates a single instance of the hash table.
18831 @node GNAT.Dynamic_Tables (g-dyntab.ads)
18832 @section @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
18833 @cindex @code{GNAT.Dynamic_Tables} (@file{g-dyntab.ads})
18834 @cindex Table implementation
18835 @cindex Arrays, extendable
18838 A generic package providing a single dimension array abstraction where the
18839 length of the array can be dynamically modified.
18842 This package provides a facility similar to that of @code{GNAT.Table},
18843 except that this package declares a type that can be used to define
18844 dynamic instances of the table, while an instantiation of
18845 @code{GNAT.Table} creates a single instance of the table type.
18847 @node GNAT.Encode_String (g-encstr.ads)
18848 @section @code{GNAT.Encode_String} (@file{g-encstr.ads})
18849 @cindex @code{GNAT.Encode_String} (@file{g-encstr.ads})
18850 @cindex Encoding strings
18851 @cindex String encoding
18852 @cindex Wide character encoding
18857 A generic package providing routines for encoding wide character and wide
18858 wide character strings as sequences of 8-bit characters using a specified
18859 encoding method. Useful in conjunction with Unicode character coding.
18860 Note there is a preinstantiation for UTF-8. See next entry.
18862 @node GNAT.Encode_UTF8_String (g-enutst.ads)
18863 @section @code{GNAT.Encode_UTF8_String} (@file{g-enutst.ads})
18864 @cindex @code{GNAT.Encode_UTF8_String} (@file{g-enutst.ads})
18865 @cindex Encoding strings
18866 @cindex Encoding UTF-8 strings
18867 @cindex UTF-8 string encoding
18868 @cindex Wide character encoding
18873 A preinstantiation of GNAT.Encode_Strings for UTF-8 encoding.
18875 @node GNAT.Exception_Actions (g-excact.ads)
18876 @section @code{GNAT.Exception_Actions} (@file{g-excact.ads})
18877 @cindex @code{GNAT.Exception_Actions} (@file{g-excact.ads})
18878 @cindex Exception actions
18881 Provides callbacks when an exception is raised. Callbacks can be registered
18882 for specific exceptions, or when any exception is raised. This
18883 can be used for instance to force a core dump to ease debugging.
18885 @node GNAT.Exception_Traces (g-exctra.ads)
18886 @section @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
18887 @cindex @code{GNAT.Exception_Traces} (@file{g-exctra.ads})
18888 @cindex Exception traces
18892 Provides an interface allowing to control automatic output upon exception
18895 @node GNAT.Exceptions (g-except.ads)
18896 @section @code{GNAT.Exceptions} (@file{g-expect.ads})
18897 @cindex @code{GNAT.Exceptions} (@file{g-expect.ads})
18898 @cindex Exceptions, Pure
18899 @cindex Pure packages, exceptions
18902 Normally it is not possible to raise an exception with
18903 a message from a subprogram in a pure package, since the
18904 necessary types and subprograms are in @code{Ada.Exceptions}
18905 which is not a pure unit. @code{GNAT.Exceptions} provides a
18906 facility for getting around this limitation for a few
18907 predefined exceptions, and for example allow raising
18908 @code{Constraint_Error} with a message from a pure subprogram.
18910 @node GNAT.Expect (g-expect.ads)
18911 @section @code{GNAT.Expect} (@file{g-expect.ads})
18912 @cindex @code{GNAT.Expect} (@file{g-expect.ads})
18915 Provides a set of subprograms similar to what is available
18916 with the standard Tcl Expect tool.
18917 It allows you to easily spawn and communicate with an external process.
18918 You can send commands or inputs to the process, and compare the output
18919 with some expected regular expression. Currently @code{GNAT.Expect}
18920 is implemented on all native GNAT ports except for OpenVMS@.
18921 It is not implemented for cross ports, and in particular is not
18922 implemented for VxWorks or LynxOS@.
18924 @node GNAT.Expect.TTY (g-exptty.ads)
18925 @section @code{GNAT.Expect.TTY} (@file{g-exptty.ads})
18926 @cindex @code{GNAT.Expect.TTY} (@file{g-exptty.ads})
18929 As GNAT.Expect but using pseudo-terminal.
18930 Currently @code{GNAT.Expect.TTY} is implemented on all native GNAT
18931 ports except for OpenVMS@. It is not implemented for cross ports, and
18932 in particular is not implemented for VxWorks or LynxOS@.
18934 @node GNAT.Float_Control (g-flocon.ads)
18935 @section @code{GNAT.Float_Control} (@file{g-flocon.ads})
18936 @cindex @code{GNAT.Float_Control} (@file{g-flocon.ads})
18937 @cindex Floating-Point Processor
18940 Provides an interface for resetting the floating-point processor into the
18941 mode required for correct semantic operation in Ada. Some third party
18942 library calls may cause this mode to be modified, and the Reset procedure
18943 in this package can be used to reestablish the required mode.
18945 @node GNAT.Heap_Sort (g-heasor.ads)
18946 @section @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
18947 @cindex @code{GNAT.Heap_Sort} (@file{g-heasor.ads})
18951 Provides a general implementation of heap sort usable for sorting arbitrary
18952 data items. Exchange and comparison procedures are provided by passing
18953 access-to-procedure values. The algorithm used is a modified heap sort
18954 that performs approximately N*log(N) comparisons in the worst case.
18956 @node GNAT.Heap_Sort_A (g-hesora.ads)
18957 @section @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
18958 @cindex @code{GNAT.Heap_Sort_A} (@file{g-hesora.ads})
18962 Provides a general implementation of heap sort usable for sorting arbitrary
18963 data items. Move and comparison procedures are provided by passing
18964 access-to-procedure values. The algorithm used is a modified heap sort
18965 that performs approximately N*log(N) comparisons in the worst case.
18966 This differs from @code{GNAT.Heap_Sort} in having a less convenient
18967 interface, but may be slightly more efficient.
18969 @node GNAT.Heap_Sort_G (g-hesorg.ads)
18970 @section @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
18971 @cindex @code{GNAT.Heap_Sort_G} (@file{g-hesorg.ads})
18975 Similar to @code{Heap_Sort_A} except that the move and sorting procedures
18976 are provided as generic parameters, this improves efficiency, especially
18977 if the procedures can be inlined, at the expense of duplicating code for
18978 multiple instantiations.
18980 @node GNAT.HTable (g-htable.ads)
18981 @section @code{GNAT.HTable} (@file{g-htable.ads})
18982 @cindex @code{GNAT.HTable} (@file{g-htable.ads})
18983 @cindex Hash tables
18986 A generic implementation of hash tables that can be used to hash arbitrary
18987 data. Provides two approaches, one a simple static approach, and the other
18988 allowing arbitrary dynamic hash tables.
18990 @node GNAT.IO (g-io.ads)
18991 @section @code{GNAT.IO} (@file{g-io.ads})
18992 @cindex @code{GNAT.IO} (@file{g-io.ads})
18994 @cindex Input/Output facilities
18997 A simple preelaborable input-output package that provides a subset of
18998 simple Text_IO functions for reading characters and strings from
18999 Standard_Input, and writing characters, strings and integers to either
19000 Standard_Output or Standard_Error.
19002 @node GNAT.IO_Aux (g-io_aux.ads)
19003 @section @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
19004 @cindex @code{GNAT.IO_Aux} (@file{g-io_aux.ads})
19006 @cindex Input/Output facilities
19008 Provides some auxiliary functions for use with Text_IO, including a test
19009 for whether a file exists, and functions for reading a line of text.
19011 @node GNAT.Lock_Files (g-locfil.ads)
19012 @section @code{GNAT.Lock_Files} (@file{g-locfil.ads})
19013 @cindex @code{GNAT.Lock_Files} (@file{g-locfil.ads})
19014 @cindex File locking
19015 @cindex Locking using files
19018 Provides a general interface for using files as locks. Can be used for
19019 providing program level synchronization.
19021 @node GNAT.MBBS_Discrete_Random (g-mbdira.ads)
19022 @section @code{GNAT.MBBS_Discrete_Random} (@file{g-mbdira.ads})
19023 @cindex @code{GNAT.MBBS_Discrete_Random} (@file{g-mbdira.ads})
19024 @cindex Random number generation
19027 The original implementation of @code{Ada.Numerics.Discrete_Random}. Uses
19028 a modified version of the Blum-Blum-Shub generator.
19030 @node GNAT.MBBS_Float_Random (g-mbflra.ads)
19031 @section @code{GNAT.MBBS_Float_Random} (@file{g-mbflra.ads})
19032 @cindex @code{GNAT.MBBS_Float_Random} (@file{g-mbflra.ads})
19033 @cindex Random number generation
19036 The original implementation of @code{Ada.Numerics.Float_Random}. Uses
19037 a modified version of the Blum-Blum-Shub generator.
19039 @node GNAT.MD5 (g-md5.ads)
19040 @section @code{GNAT.MD5} (@file{g-md5.ads})
19041 @cindex @code{GNAT.MD5} (@file{g-md5.ads})
19042 @cindex Message Digest MD5
19045 Implements the MD5 Message-Digest Algorithm as described in RFC 1321.
19047 @node GNAT.Memory_Dump (g-memdum.ads)
19048 @section @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
19049 @cindex @code{GNAT.Memory_Dump} (@file{g-memdum.ads})
19050 @cindex Dump Memory
19053 Provides a convenient routine for dumping raw memory to either the
19054 standard output or standard error files. Uses GNAT.IO for actual
19057 @node GNAT.Most_Recent_Exception (g-moreex.ads)
19058 @section @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
19059 @cindex @code{GNAT.Most_Recent_Exception} (@file{g-moreex.ads})
19060 @cindex Exception, obtaining most recent
19063 Provides access to the most recently raised exception. Can be used for
19064 various logging purposes, including duplicating functionality of some
19065 Ada 83 implementation dependent extensions.
19067 @node GNAT.OS_Lib (g-os_lib.ads)
19068 @section @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
19069 @cindex @code{GNAT.OS_Lib} (@file{g-os_lib.ads})
19070 @cindex Operating System interface
19071 @cindex Spawn capability
19074 Provides a range of target independent operating system interface functions,
19075 including time/date management, file operations, subprocess management,
19076 including a portable spawn procedure, and access to environment variables
19077 and error return codes.
19079 @node GNAT.Perfect_Hash_Generators (g-pehage.ads)
19080 @section @code{GNAT.Perfect_Hash_Generators} (@file{g-pehage.ads})
19081 @cindex @code{GNAT.Perfect_Hash_Generators} (@file{g-pehage.ads})
19082 @cindex Hash functions
19085 Provides a generator of static minimal perfect hash functions. No
19086 collisions occur and each item can be retrieved from the table in one
19087 probe (perfect property). The hash table size corresponds to the exact
19088 size of the key set and no larger (minimal property). The key set has to
19089 be know in advance (static property). The hash functions are also order
19090 preserving. If w2 is inserted after w1 in the generator, their
19091 hashcode are in the same order. These hashing functions are very
19092 convenient for use with realtime applications.
19094 @node GNAT.Random_Numbers (g-rannum.ads)
19095 @section @code{GNAT.Random_Numbers} (@file{g-rannum.ads})
19096 @cindex @code{GNAT.Random_Numbers} (@file{g-rannum.ads})
19097 @cindex Random number generation
19100 Provides random number capabilities which extend those available in the
19101 standard Ada library and are more convenient to use.
19103 @node GNAT.Regexp (g-regexp.ads)
19104 @section @code{GNAT.Regexp} (@file{g-regexp.ads})
19105 @cindex @code{GNAT.Regexp} (@file{g-regexp.ads})
19106 @cindex Regular expressions
19107 @cindex Pattern matching
19110 A simple implementation of regular expressions, using a subset of regular
19111 expression syntax copied from familiar Unix style utilities. This is the
19112 simples of the three pattern matching packages provided, and is particularly
19113 suitable for ``file globbing'' applications.
19115 @node GNAT.Registry (g-regist.ads)
19116 @section @code{GNAT.Registry} (@file{g-regist.ads})
19117 @cindex @code{GNAT.Registry} (@file{g-regist.ads})
19118 @cindex Windows Registry
19121 This is a high level binding to the Windows registry. It is possible to
19122 do simple things like reading a key value, creating a new key. For full
19123 registry API, but at a lower level of abstraction, refer to the Win32.Winreg
19124 package provided with the Win32Ada binding
19126 @node GNAT.Regpat (g-regpat.ads)
19127 @section @code{GNAT.Regpat} (@file{g-regpat.ads})
19128 @cindex @code{GNAT.Regpat} (@file{g-regpat.ads})
19129 @cindex Regular expressions
19130 @cindex Pattern matching
19133 A complete implementation of Unix-style regular expression matching, copied
19134 from the original V7 style regular expression library written in C by
19135 Henry Spencer (and binary compatible with this C library).
19137 @node GNAT.Secondary_Stack_Info (g-sestin.ads)
19138 @section @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
19139 @cindex @code{GNAT.Secondary_Stack_Info} (@file{g-sestin.ads})
19140 @cindex Secondary Stack Info
19143 Provide the capability to query the high water mark of the current task's
19146 @node GNAT.Semaphores (g-semaph.ads)
19147 @section @code{GNAT.Semaphores} (@file{g-semaph.ads})
19148 @cindex @code{GNAT.Semaphores} (@file{g-semaph.ads})
19152 Provides classic counting and binary semaphores using protected types.
19154 @node GNAT.Serial_Communications (g-sercom.ads)
19155 @section @code{GNAT.Serial_Communications} (@file{g-sercom.ads})
19156 @cindex @code{GNAT.Serial_Communications} (@file{g-sercom.ads})
19157 @cindex Serial_Communications
19160 Provides a simple interface to send and receive data over a serial
19161 port. This is only supported on GNU/Linux and Windows.
19163 @node GNAT.SHA1 (g-sha1.ads)
19164 @section @code{GNAT.SHA1} (@file{g-sha1.ads})
19165 @cindex @code{GNAT.SHA1} (@file{g-sha1.ads})
19166 @cindex Secure Hash Algorithm SHA-1
19169 Implements the SHA-1 Secure Hash Algorithm as described in FIPS PUB 180-3
19172 @node GNAT.SHA224 (g-sha224.ads)
19173 @section @code{GNAT.SHA224} (@file{g-sha224.ads})
19174 @cindex @code{GNAT.SHA224} (@file{g-sha224.ads})
19175 @cindex Secure Hash Algorithm SHA-224
19178 Implements the SHA-224 Secure Hash Algorithm as described in FIPS PUB 180-3.
19180 @node GNAT.SHA256 (g-sha256.ads)
19181 @section @code{GNAT.SHA256} (@file{g-sha256.ads})
19182 @cindex @code{GNAT.SHA256} (@file{g-sha256.ads})
19183 @cindex Secure Hash Algorithm SHA-256
19186 Implements the SHA-256 Secure Hash Algorithm as described in FIPS PUB 180-3.
19188 @node GNAT.SHA384 (g-sha384.ads)
19189 @section @code{GNAT.SHA384} (@file{g-sha384.ads})
19190 @cindex @code{GNAT.SHA384} (@file{g-sha384.ads})
19191 @cindex Secure Hash Algorithm SHA-384
19194 Implements the SHA-384 Secure Hash Algorithm as described in FIPS PUB 180-3.
19196 @node GNAT.SHA512 (g-sha512.ads)
19197 @section @code{GNAT.SHA512} (@file{g-sha512.ads})
19198 @cindex @code{GNAT.SHA512} (@file{g-sha512.ads})
19199 @cindex Secure Hash Algorithm SHA-512
19202 Implements the SHA-512 Secure Hash Algorithm as described in FIPS PUB 180-3.
19204 @node GNAT.Signals (g-signal.ads)
19205 @section @code{GNAT.Signals} (@file{g-signal.ads})
19206 @cindex @code{GNAT.Signals} (@file{g-signal.ads})
19210 Provides the ability to manipulate the blocked status of signals on supported
19213 @node GNAT.Sockets (g-socket.ads)
19214 @section @code{GNAT.Sockets} (@file{g-socket.ads})
19215 @cindex @code{GNAT.Sockets} (@file{g-socket.ads})
19219 A high level and portable interface to develop sockets based applications.
19220 This package is based on the sockets thin binding found in
19221 @code{GNAT.Sockets.Thin}. Currently @code{GNAT.Sockets} is implemented
19222 on all native GNAT ports except for OpenVMS@. It is not implemented
19223 for the LynxOS@ cross port.
19225 @node GNAT.Source_Info (g-souinf.ads)
19226 @section @code{GNAT.Source_Info} (@file{g-souinf.ads})
19227 @cindex @code{GNAT.Source_Info} (@file{g-souinf.ads})
19228 @cindex Source Information
19231 Provides subprograms that give access to source code information known at
19232 compile time, such as the current file name and line number.
19234 @node GNAT.Spelling_Checker (g-speche.ads)
19235 @section @code{GNAT.Spelling_Checker} (@file{g-speche.ads})
19236 @cindex @code{GNAT.Spelling_Checker} (@file{g-speche.ads})
19237 @cindex Spell checking
19240 Provides a function for determining whether one string is a plausible
19241 near misspelling of another string.
19243 @node GNAT.Spelling_Checker_Generic (g-spchge.ads)
19244 @section @code{GNAT.Spelling_Checker_Generic} (@file{g-spchge.ads})
19245 @cindex @code{GNAT.Spelling_Checker_Generic} (@file{g-spchge.ads})
19246 @cindex Spell checking
19249 Provides a generic function that can be instantiated with a string type for
19250 determining whether one string is a plausible near misspelling of another
19253 @node GNAT.Spitbol.Patterns (g-spipat.ads)
19254 @section @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
19255 @cindex @code{GNAT.Spitbol.Patterns} (@file{g-spipat.ads})
19256 @cindex SPITBOL pattern matching
19257 @cindex Pattern matching
19260 A complete implementation of SNOBOL4 style pattern matching. This is the
19261 most elaborate of the pattern matching packages provided. It fully duplicates
19262 the SNOBOL4 dynamic pattern construction and matching capabilities, using the
19263 efficient algorithm developed by Robert Dewar for the SPITBOL system.
19265 @node GNAT.Spitbol (g-spitbo.ads)
19266 @section @code{GNAT.Spitbol} (@file{g-spitbo.ads})
19267 @cindex @code{GNAT.Spitbol} (@file{g-spitbo.ads})
19268 @cindex SPITBOL interface
19271 The top level package of the collection of SPITBOL-style functionality, this
19272 package provides basic SNOBOL4 string manipulation functions, such as
19273 Pad, Reverse, Trim, Substr capability, as well as a generic table function
19274 useful for constructing arbitrary mappings from strings in the style of
19275 the SNOBOL4 TABLE function.
19277 @node GNAT.Spitbol.Table_Boolean (g-sptabo.ads)
19278 @section @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
19279 @cindex @code{GNAT.Spitbol.Table_Boolean} (@file{g-sptabo.ads})
19280 @cindex Sets of strings
19281 @cindex SPITBOL Tables
19284 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
19285 for type @code{Standard.Boolean}, giving an implementation of sets of
19288 @node GNAT.Spitbol.Table_Integer (g-sptain.ads)
19289 @section @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
19290 @cindex @code{GNAT.Spitbol.Table_Integer} (@file{g-sptain.ads})
19291 @cindex Integer maps
19293 @cindex SPITBOL Tables
19296 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table}
19297 for type @code{Standard.Integer}, giving an implementation of maps
19298 from string to integer values.
19300 @node GNAT.Spitbol.Table_VString (g-sptavs.ads)
19301 @section @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
19302 @cindex @code{GNAT.Spitbol.Table_VString} (@file{g-sptavs.ads})
19303 @cindex String maps
19305 @cindex SPITBOL Tables
19308 A library level of instantiation of @code{GNAT.Spitbol.Patterns.Table} for
19309 a variable length string type, giving an implementation of general
19310 maps from strings to strings.
19312 @node GNAT.SSE (g-sse.ads)
19313 @section @code{GNAT.SSE} (@file{g-sse.ads})
19314 @cindex @code{GNAT.SSE} (@file{g-sse.ads})
19317 Root of a set of units aimed at offering Ada bindings to a subset of
19318 the Intel(r) Streaming SIMD Extensions with GNAT on the x86 family of
19319 targets. It exposes vector component types together with a general
19320 introduction to the binding contents and use.
19322 @node GNAT.SSE.Vector_Types (g-ssvety.ads)
19323 @section @code{GNAT.SSE.Vector_Types} (@file{g-ssvety.ads})
19324 @cindex @code{GNAT.SSE.Vector_Types} (@file{g-ssvety.ads})
19327 SSE vector types for use with SSE related intrinsics.
19329 @node GNAT.Strings (g-string.ads)
19330 @section @code{GNAT.Strings} (@file{g-string.ads})
19331 @cindex @code{GNAT.Strings} (@file{g-string.ads})
19334 Common String access types and related subprograms. Basically it
19335 defines a string access and an array of string access types.
19337 @node GNAT.String_Split (g-strspl.ads)
19338 @section @code{GNAT.String_Split} (@file{g-strspl.ads})
19339 @cindex @code{GNAT.String_Split} (@file{g-strspl.ads})
19340 @cindex String splitter
19343 Useful string manipulation routines: given a set of separators, split
19344 a string wherever the separators appear, and provide direct access
19345 to the resulting slices. This package is instantiated from
19346 @code{GNAT.Array_Split}.
19348 @node GNAT.Table (g-table.ads)
19349 @section @code{GNAT.Table} (@file{g-table.ads})
19350 @cindex @code{GNAT.Table} (@file{g-table.ads})
19351 @cindex Table implementation
19352 @cindex Arrays, extendable
19355 A generic package providing a single dimension array abstraction where the
19356 length of the array can be dynamically modified.
19359 This package provides a facility similar to that of @code{GNAT.Dynamic_Tables},
19360 except that this package declares a single instance of the table type,
19361 while an instantiation of @code{GNAT.Dynamic_Tables} creates a type that can be
19362 used to define dynamic instances of the table.
19364 @node GNAT.Task_Lock (g-tasloc.ads)
19365 @section @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
19366 @cindex @code{GNAT.Task_Lock} (@file{g-tasloc.ads})
19367 @cindex Task synchronization
19368 @cindex Task locking
19372 A very simple facility for locking and unlocking sections of code using a
19373 single global task lock. Appropriate for use in situations where contention
19374 between tasks is very rarely expected.
19376 @node GNAT.Time_Stamp (g-timsta.ads)
19377 @section @code{GNAT.Time_Stamp} (@file{g-timsta.ads})
19378 @cindex @code{GNAT.Time_Stamp} (@file{g-timsta.ads})
19380 @cindex Current time
19383 Provides a simple function that returns a string YYYY-MM-DD HH:MM:SS.SS that
19384 represents the current date and time in ISO 8601 format. This is a very simple
19385 routine with minimal code and there are no dependencies on any other unit.
19387 @node GNAT.Threads (g-thread.ads)
19388 @section @code{GNAT.Threads} (@file{g-thread.ads})
19389 @cindex @code{GNAT.Threads} (@file{g-thread.ads})
19390 @cindex Foreign threads
19391 @cindex Threads, foreign
19394 Provides facilities for dealing with foreign threads which need to be known
19395 by the GNAT run-time system. Consult the documentation of this package for
19396 further details if your program has threads that are created by a non-Ada
19397 environment which then accesses Ada code.
19399 @node GNAT.Traceback (g-traceb.ads)
19400 @section @code{GNAT.Traceback} (@file{g-traceb.ads})
19401 @cindex @code{GNAT.Traceback} (@file{g-traceb.ads})
19402 @cindex Trace back facilities
19405 Provides a facility for obtaining non-symbolic traceback information, useful
19406 in various debugging situations.
19408 @node GNAT.Traceback.Symbolic (g-trasym.ads)
19409 @section @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
19410 @cindex @code{GNAT.Traceback.Symbolic} (@file{g-trasym.ads})
19411 @cindex Trace back facilities
19413 @node GNAT.UTF_32 (g-utf_32.ads)
19414 @section @code{GNAT.UTF_32} (@file{g-table.ads})
19415 @cindex @code{GNAT.UTF_32} (@file{g-table.ads})
19416 @cindex Wide character codes
19419 This is a package intended to be used in conjunction with the
19420 @code{Wide_Character} type in Ada 95 and the
19421 @code{Wide_Wide_Character} type in Ada 2005 (available
19422 in @code{GNAT} in Ada 2005 mode). This package contains
19423 Unicode categorization routines, as well as lexical
19424 categorization routines corresponding to the Ada 2005
19425 lexical rules for identifiers and strings, and also a
19426 lower case to upper case fold routine corresponding to
19427 the Ada 2005 rules for identifier equivalence.
19429 @node GNAT.UTF_32_Spelling_Checker (g-u3spch.ads)
19430 @section @code{GNAT.Wide_Spelling_Checker} (@file{g-u3spch.ads})
19431 @cindex @code{GNAT.Wide_Spelling_Checker} (@file{g-u3spch.ads})
19432 @cindex Spell checking
19435 Provides a function for determining whether one wide wide string is a plausible
19436 near misspelling of another wide wide string, where the strings are represented
19437 using the UTF_32_String type defined in System.Wch_Cnv.
19439 @node GNAT.Wide_Spelling_Checker (g-wispch.ads)
19440 @section @code{GNAT.Wide_Spelling_Checker} (@file{g-wispch.ads})
19441 @cindex @code{GNAT.Wide_Spelling_Checker} (@file{g-wispch.ads})
19442 @cindex Spell checking
19445 Provides a function for determining whether one wide string is a plausible
19446 near misspelling of another wide string.
19448 @node GNAT.Wide_String_Split (g-wistsp.ads)
19449 @section @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
19450 @cindex @code{GNAT.Wide_String_Split} (@file{g-wistsp.ads})
19451 @cindex Wide_String splitter
19454 Useful wide string manipulation routines: given a set of separators, split
19455 a wide string wherever the separators appear, and provide direct access
19456 to the resulting slices. This package is instantiated from
19457 @code{GNAT.Array_Split}.
19459 @node GNAT.Wide_Wide_Spelling_Checker (g-zspche.ads)
19460 @section @code{GNAT.Wide_Wide_Spelling_Checker} (@file{g-zspche.ads})
19461 @cindex @code{GNAT.Wide_Wide_Spelling_Checker} (@file{g-zspche.ads})
19462 @cindex Spell checking
19465 Provides a function for determining whether one wide wide string is a plausible
19466 near misspelling of another wide wide string.
19468 @node GNAT.Wide_Wide_String_Split (g-zistsp.ads)
19469 @section @code{GNAT.Wide_Wide_String_Split} (@file{g-zistsp.ads})
19470 @cindex @code{GNAT.Wide_Wide_String_Split} (@file{g-zistsp.ads})
19471 @cindex Wide_Wide_String splitter
19474 Useful wide wide string manipulation routines: given a set of separators, split
19475 a wide wide string wherever the separators appear, and provide direct access
19476 to the resulting slices. This package is instantiated from
19477 @code{GNAT.Array_Split}.
19479 @node Interfaces.C.Extensions (i-cexten.ads)
19480 @section @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
19481 @cindex @code{Interfaces.C.Extensions} (@file{i-cexten.ads})
19484 This package contains additional C-related definitions, intended
19485 for use with either manually or automatically generated bindings
19488 @node Interfaces.C.Streams (i-cstrea.ads)
19489 @section @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
19490 @cindex @code{Interfaces.C.Streams} (@file{i-cstrea.ads})
19491 @cindex C streams, interfacing
19494 This package is a binding for the most commonly used operations
19497 @node Interfaces.CPP (i-cpp.ads)
19498 @section @code{Interfaces.CPP} (@file{i-cpp.ads})
19499 @cindex @code{Interfaces.CPP} (@file{i-cpp.ads})
19500 @cindex C++ interfacing
19501 @cindex Interfacing, to C++
19504 This package provides facilities for use in interfacing to C++. It
19505 is primarily intended to be used in connection with automated tools
19506 for the generation of C++ interfaces.
19508 @node Interfaces.Packed_Decimal (i-pacdec.ads)
19509 @section @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
19510 @cindex @code{Interfaces.Packed_Decimal} (@file{i-pacdec.ads})
19511 @cindex IBM Packed Format
19512 @cindex Packed Decimal
19515 This package provides a set of routines for conversions to and
19516 from a packed decimal format compatible with that used on IBM
19519 @node Interfaces.VxWorks (i-vxwork.ads)
19520 @section @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
19521 @cindex @code{Interfaces.VxWorks} (@file{i-vxwork.ads})
19522 @cindex Interfacing to VxWorks
19523 @cindex VxWorks, interfacing
19526 This package provides a limited binding to the VxWorks API.
19527 In particular, it interfaces with the
19528 VxWorks hardware interrupt facilities.
19530 @node Interfaces.VxWorks.IO (i-vxwoio.ads)
19531 @section @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
19532 @cindex @code{Interfaces.VxWorks.IO} (@file{i-vxwoio.ads})
19533 @cindex Interfacing to VxWorks' I/O
19534 @cindex VxWorks, I/O interfacing
19535 @cindex VxWorks, Get_Immediate
19536 @cindex Get_Immediate, VxWorks
19539 This package provides a binding to the ioctl (IO/Control)
19540 function of VxWorks, defining a set of option values and
19541 function codes. A particular use of this package is
19542 to enable the use of Get_Immediate under VxWorks.
19544 @node System.Address_Image (s-addima.ads)
19545 @section @code{System.Address_Image} (@file{s-addima.ads})
19546 @cindex @code{System.Address_Image} (@file{s-addima.ads})
19547 @cindex Address image
19548 @cindex Image, of an address
19551 This function provides a useful debugging
19552 function that gives an (implementation dependent)
19553 string which identifies an address.
19555 @node System.Assertions (s-assert.ads)
19556 @section @code{System.Assertions} (@file{s-assert.ads})
19557 @cindex @code{System.Assertions} (@file{s-assert.ads})
19559 @cindex Assert_Failure, exception
19562 This package provides the declaration of the exception raised
19563 by an run-time assertion failure, as well as the routine that
19564 is used internally to raise this assertion.
19566 @node System.Memory (s-memory.ads)
19567 @section @code{System.Memory} (@file{s-memory.ads})
19568 @cindex @code{System.Memory} (@file{s-memory.ads})
19569 @cindex Memory allocation
19572 This package provides the interface to the low level routines used
19573 by the generated code for allocation and freeing storage for the
19574 default storage pool (analogous to the C routines malloc and free.
19575 It also provides a reallocation interface analogous to the C routine
19576 realloc. The body of this unit may be modified to provide alternative
19577 allocation mechanisms for the default pool, and in addition, direct
19578 calls to this unit may be made for low level allocation uses (for
19579 example see the body of @code{GNAT.Tables}).
19581 @node System.Multiprocessors (s-multip.ads)
19582 @section @code{System.Multiprocessors} (@file{s-multip.ads})
19583 @cindex @code{System.Multiprocessors} (@file{s-multip.ads})
19584 @cindex Multiprocessor interface
19585 This is an Ada 2012 unit defined in the Ada 2012 Reference Manual, but
19586 in GNAT we also make it available in Ada 95 and Ada 2005 (where it is
19587 technically an implementation-defined addition).
19589 @node System.Multiprocessors.Dispatching_Domains (s-mudido.ads)
19590 @section @code{System.Multiprocessors.Dispatching_Domains} (@file{s-mudido.ads})
19591 @cindex @code{System.Multiprocessors.Dispatching_Domains} (@file{s-mudido.ads})
19592 @cindex Multiprocessor interface
19593 This is an Ada 2012 unit defined in the Ada 2012 Reference Manual, but
19594 in GNAT we also make it available in Ada 95 and Ada 2005 (where it is
19595 technically an implementation-defined addition).
19597 @node System.Partition_Interface (s-parint.ads)
19598 @section @code{System.Partition_Interface} (@file{s-parint.ads})
19599 @cindex @code{System.Partition_Interface} (@file{s-parint.ads})
19600 @cindex Partition interfacing functions
19603 This package provides facilities for partition interfacing. It
19604 is used primarily in a distribution context when using Annex E
19607 @node System.Pool_Global (s-pooglo.ads)
19608 @section @code{System.Pool_Global} (@file{s-pooglo.ads})
19609 @cindex @code{System.Pool_Global} (@file{s-pooglo.ads})
19610 @cindex Storage pool, global
19611 @cindex Global storage pool
19614 This package provides a storage pool that is equivalent to the default
19615 storage pool used for access types for which no pool is specifically
19616 declared. It uses malloc/free to allocate/free and does not attempt to
19617 do any automatic reclamation.
19619 @node System.Pool_Local (s-pooloc.ads)
19620 @section @code{System.Pool_Local} (@file{s-pooloc.ads})
19621 @cindex @code{System.Pool_Local} (@file{s-pooloc.ads})
19622 @cindex Storage pool, local
19623 @cindex Local storage pool
19626 This package provides a storage pool that is intended for use with locally
19627 defined access types. It uses malloc/free for allocate/free, and maintains
19628 a list of allocated blocks, so that all storage allocated for the pool can
19629 be freed automatically when the pool is finalized.
19631 @node System.Restrictions (s-restri.ads)
19632 @section @code{System.Restrictions} (@file{s-restri.ads})
19633 @cindex @code{System.Restrictions} (@file{s-restri.ads})
19634 @cindex Run-time restrictions access
19637 This package provides facilities for accessing at run time
19638 the status of restrictions specified at compile time for
19639 the partition. Information is available both with regard
19640 to actual restrictions specified, and with regard to
19641 compiler determined information on which restrictions
19642 are violated by one or more packages in the partition.
19644 @node System.Rident (s-rident.ads)
19645 @section @code{System.Rident} (@file{s-rident.ads})
19646 @cindex @code{System.Rident} (@file{s-rident.ads})
19647 @cindex Restrictions definitions
19650 This package provides definitions of the restrictions
19651 identifiers supported by GNAT, and also the format of
19652 the restrictions provided in package System.Restrictions.
19653 It is not normally necessary to @code{with} this generic package
19654 since the necessary instantiation is included in
19655 package System.Restrictions.
19657 @node System.Strings.Stream_Ops (s-ststop.ads)
19658 @section @code{System.Strings.Stream_Ops} (@file{s-ststop.ads})
19659 @cindex @code{System.Strings.Stream_Ops} (@file{s-ststop.ads})
19660 @cindex Stream operations
19661 @cindex String stream operations
19664 This package provides a set of stream subprograms for standard string types.
19665 It is intended primarily to support implicit use of such subprograms when
19666 stream attributes are applied to string types, but the subprograms in this
19667 package can be used directly by application programs.
19669 @node System.Task_Info (s-tasinf.ads)
19670 @section @code{System.Task_Info} (@file{s-tasinf.ads})
19671 @cindex @code{System.Task_Info} (@file{s-tasinf.ads})
19672 @cindex Task_Info pragma
19675 This package provides target dependent functionality that is used
19676 to support the @code{Task_Info} pragma
19678 @node System.Wch_Cnv (s-wchcnv.ads)
19679 @section @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
19680 @cindex @code{System.Wch_Cnv} (@file{s-wchcnv.ads})
19681 @cindex Wide Character, Representation
19682 @cindex Wide String, Conversion
19683 @cindex Representation of wide characters
19686 This package provides routines for converting between
19687 wide and wide wide characters and a representation as a value of type
19688 @code{Standard.String}, using a specified wide character
19689 encoding method. It uses definitions in
19690 package @code{System.Wch_Con}.
19692 @node System.Wch_Con (s-wchcon.ads)
19693 @section @code{System.Wch_Con} (@file{s-wchcon.ads})
19694 @cindex @code{System.Wch_Con} (@file{s-wchcon.ads})
19697 This package provides definitions and descriptions of
19698 the various methods used for encoding wide characters
19699 in ordinary strings. These definitions are used by
19700 the package @code{System.Wch_Cnv}.
19702 @node Interfacing to Other Languages
19703 @chapter Interfacing to Other Languages
19705 The facilities in annex B of the Ada Reference Manual are fully
19706 implemented in GNAT, and in addition, a full interface to C++ is
19710 * Interfacing to C::
19711 * Interfacing to C++::
19712 * Interfacing to COBOL::
19713 * Interfacing to Fortran::
19714 * Interfacing to non-GNAT Ada code::
19717 @node Interfacing to C
19718 @section Interfacing to C
19721 Interfacing to C with GNAT can use one of two approaches:
19725 The types in the package @code{Interfaces.C} may be used.
19727 Standard Ada types may be used directly. This may be less portable to
19728 other compilers, but will work on all GNAT compilers, which guarantee
19729 correspondence between the C and Ada types.
19733 Pragma @code{Convention C} may be applied to Ada types, but mostly has no
19734 effect, since this is the default. The following table shows the
19735 correspondence between Ada scalar types and the corresponding C types.
19740 @item Short_Integer
19742 @item Short_Short_Integer
19746 @item Long_Long_Integer
19754 @item Long_Long_Float
19755 This is the longest floating-point type supported by the hardware.
19759 Additionally, there are the following general correspondences between Ada
19763 Ada enumeration types map to C enumeration types directly if pragma
19764 @code{Convention C} is specified, which causes them to have int
19765 length. Without pragma @code{Convention C}, Ada enumeration types map to
19766 8, 16, or 32 bits (i.e.@: C types @code{signed char}, @code{short},
19767 @code{int}, respectively) depending on the number of values passed.
19768 This is the only case in which pragma @code{Convention C} affects the
19769 representation of an Ada type.
19772 Ada access types map to C pointers, except for the case of pointers to
19773 unconstrained types in Ada, which have no direct C equivalent.
19776 Ada arrays map directly to C arrays.
19779 Ada records map directly to C structures.
19782 Packed Ada records map to C structures where all members are bit fields
19783 of the length corresponding to the @code{@var{type}'Size} value in Ada.
19786 @node Interfacing to C++
19787 @section Interfacing to C++
19790 The interface to C++ makes use of the following pragmas, which are
19791 primarily intended to be constructed automatically using a binding generator
19792 tool, although it is possible to construct them by hand.
19794 Using these pragmas it is possible to achieve complete
19795 inter-operability between Ada tagged types and C++ class definitions.
19796 See @ref{Implementation Defined Pragmas}, for more details.
19799 @item pragma CPP_Class ([Entity =>] @var{LOCAL_NAME})
19800 The argument denotes an entity in the current declarative region that is
19801 declared as a tagged or untagged record type. It indicates that the type
19802 corresponds to an externally declared C++ class type, and is to be laid
19803 out the same way that C++ would lay out the type.
19805 Note: Pragma @code{CPP_Class} is currently obsolete. It is supported
19806 for backward compatibility but its functionality is available
19807 using pragma @code{Import} with @code{Convention} = @code{CPP}.
19809 @item pragma CPP_Constructor ([Entity =>] @var{LOCAL_NAME})
19810 This pragma identifies an imported function (imported in the usual way
19811 with pragma @code{Import}) as corresponding to a C++ constructor.
19814 A few restrictions are placed on the use of the @code{Access} attribute
19815 in conjunction with subprograms subject to convention @code{CPP}: the
19816 attribute may be used neither on primitive operations of a tagged
19817 record type with convention @code{CPP}, imported or not, nor on
19818 subprograms imported with pragma @code{CPP_Constructor}.
19820 In addition, C++ exceptions are propagated and can be handled in an
19821 @code{others} choice of an exception handler. The corresponding Ada
19822 occurrence has no message, and the simple name of the exception identity
19823 contains @samp{Foreign_Exception}. Finalization and awaiting dependent
19824 tasks works properly when such foreign exceptions are propagated.
19826 It is also possible to import a C++ exception using the following syntax:
19828 @smallexample @c ada
19829 LOCAL_NAME : exception;
19830 pragma Import (Cpp,
19831 [Entity =>] LOCAL_NAME,
19832 [External_Name =>] static_string_EXPRESSION);
19836 The @code{External_Name} is the name of the C++ RTTI symbol. You can then
19837 cover a specific C++ exception in an exception handler.
19839 @node Interfacing to COBOL
19840 @section Interfacing to COBOL
19843 Interfacing to COBOL is achieved as described in section B.4 of
19844 the Ada Reference Manual.
19846 @node Interfacing to Fortran
19847 @section Interfacing to Fortran
19850 Interfacing to Fortran is achieved as described in section B.5 of the
19851 Ada Reference Manual. The pragma @code{Convention Fortran}, applied to a
19852 multi-dimensional array causes the array to be stored in column-major
19853 order as required for convenient interface to Fortran.
19855 @node Interfacing to non-GNAT Ada code
19856 @section Interfacing to non-GNAT Ada code
19858 It is possible to specify the convention @code{Ada} in a pragma
19859 @code{Import} or pragma @code{Export}. However this refers to
19860 the calling conventions used by GNAT, which may or may not be
19861 similar enough to those used by some other Ada 83 / Ada 95 / Ada 2005
19862 compiler to allow interoperation.
19864 If arguments types are kept simple, and if the foreign compiler generally
19865 follows system calling conventions, then it may be possible to integrate
19866 files compiled by other Ada compilers, provided that the elaboration
19867 issues are adequately addressed (for example by eliminating the
19868 need for any load time elaboration).
19870 In particular, GNAT running on VMS is designed to
19871 be highly compatible with the DEC Ada 83 compiler, so this is one
19872 case in which it is possible to import foreign units of this type,
19873 provided that the data items passed are restricted to simple scalar
19874 values or simple record types without variants, or simple array
19875 types with fixed bounds.
19877 @node Specialized Needs Annexes
19878 @chapter Specialized Needs Annexes
19881 Ada 95 and Ada 2005 define a number of Specialized Needs Annexes, which are not
19882 required in all implementations. However, as described in this chapter,
19883 GNAT implements all of these annexes:
19886 @item Systems Programming (Annex C)
19887 The Systems Programming Annex is fully implemented.
19889 @item Real-Time Systems (Annex D)
19890 The Real-Time Systems Annex is fully implemented.
19892 @item Distributed Systems (Annex E)
19893 Stub generation is fully implemented in the GNAT compiler. In addition,
19894 a complete compatible PCS is available as part of the GLADE system,
19895 a separate product. When the two
19896 products are used in conjunction, this annex is fully implemented.
19898 @item Information Systems (Annex F)
19899 The Information Systems annex is fully implemented.
19901 @item Numerics (Annex G)
19902 The Numerics Annex is fully implemented.
19904 @item Safety and Security / High-Integrity Systems (Annex H)
19905 The Safety and Security Annex (termed the High-Integrity Systems Annex
19906 in Ada 2005) is fully implemented.
19909 @node Implementation of Specific Ada Features
19910 @chapter Implementation of Specific Ada Features
19913 This chapter describes the GNAT implementation of several Ada language
19917 * Machine Code Insertions::
19918 * GNAT Implementation of Tasking::
19919 * GNAT Implementation of Shared Passive Packages::
19920 * Code Generation for Array Aggregates::
19921 * The Size of Discriminated Records with Default Discriminants::
19922 * Strict Conformance to the Ada Reference Manual::
19925 @node Machine Code Insertions
19926 @section Machine Code Insertions
19927 @cindex Machine Code insertions
19930 Package @code{Machine_Code} provides machine code support as described
19931 in the Ada Reference Manual in two separate forms:
19934 Machine code statements, consisting of qualified expressions that
19935 fit the requirements of RM section 13.8.
19937 An intrinsic callable procedure, providing an alternative mechanism of
19938 including machine instructions in a subprogram.
19942 The two features are similar, and both are closely related to the mechanism
19943 provided by the asm instruction in the GNU C compiler. Full understanding
19944 and use of the facilities in this package requires understanding the asm
19945 instruction, see @ref{Extended Asm,, Assembler Instructions with C Expression
19946 Operands, gcc, Using the GNU Compiler Collection (GCC)}.
19948 Calls to the function @code{Asm} and the procedure @code{Asm} have identical
19949 semantic restrictions and effects as described below. Both are provided so
19950 that the procedure call can be used as a statement, and the function call
19951 can be used to form a code_statement.
19953 The first example given in the GCC documentation is the C @code{asm}
19956 asm ("fsinx %1 %0" : "=f" (result) : "f" (angle));
19960 The equivalent can be written for GNAT as:
19962 @smallexample @c ada
19963 Asm ("fsinx %1 %0",
19964 My_Float'Asm_Output ("=f", result),
19965 My_Float'Asm_Input ("f", angle));
19969 The first argument to @code{Asm} is the assembler template, and is
19970 identical to what is used in GNU C@. This string must be a static
19971 expression. The second argument is the output operand list. It is
19972 either a single @code{Asm_Output} attribute reference, or a list of such
19973 references enclosed in parentheses (technically an array aggregate of
19976 The @code{Asm_Output} attribute denotes a function that takes two
19977 parameters. The first is a string, the second is the name of a variable
19978 of the type designated by the attribute prefix. The first (string)
19979 argument is required to be a static expression and designates the
19980 constraint for the parameter (e.g.@: what kind of register is
19981 required). The second argument is the variable to be updated with the
19982 result. The possible values for constraint are the same as those used in
19983 the RTL, and are dependent on the configuration file used to build the
19984 GCC back end. If there are no output operands, then this argument may
19985 either be omitted, or explicitly given as @code{No_Output_Operands}.
19987 The second argument of @code{@var{my_float}'Asm_Output} functions as
19988 though it were an @code{out} parameter, which is a little curious, but
19989 all names have the form of expressions, so there is no syntactic
19990 irregularity, even though normally functions would not be permitted
19991 @code{out} parameters. The third argument is the list of input
19992 operands. It is either a single @code{Asm_Input} attribute reference, or
19993 a list of such references enclosed in parentheses (technically an array
19994 aggregate of such references).
19996 The @code{Asm_Input} attribute denotes a function that takes two
19997 parameters. The first is a string, the second is an expression of the
19998 type designated by the prefix. The first (string) argument is required
19999 to be a static expression, and is the constraint for the parameter,
20000 (e.g.@: what kind of register is required). The second argument is the
20001 value to be used as the input argument. The possible values for the
20002 constant are the same as those used in the RTL, and are dependent on
20003 the configuration file used to built the GCC back end.
20005 If there are no input operands, this argument may either be omitted, or
20006 explicitly given as @code{No_Input_Operands}. The fourth argument, not
20007 present in the above example, is a list of register names, called the
20008 @dfn{clobber} argument. This argument, if given, must be a static string
20009 expression, and is a space or comma separated list of names of registers
20010 that must be considered destroyed as a result of the @code{Asm} call. If
20011 this argument is the null string (the default value), then the code
20012 generator assumes that no additional registers are destroyed.
20014 The fifth argument, not present in the above example, called the
20015 @dfn{volatile} argument, is by default @code{False}. It can be set to
20016 the literal value @code{True} to indicate to the code generator that all
20017 optimizations with respect to the instruction specified should be
20018 suppressed, and that in particular, for an instruction that has outputs,
20019 the instruction will still be generated, even if none of the outputs are
20020 used. @xref{Extended Asm,, Assembler Instructions with C Expression Operands,
20021 gcc, Using the GNU Compiler Collection (GCC)}, for the full description.
20022 Generally it is strongly advisable to use Volatile for any ASM statement
20023 that is missing either input or output operands, or when two or more ASM
20024 statements appear in sequence, to avoid unwanted optimizations. A warning
20025 is generated if this advice is not followed.
20027 The @code{Asm} subprograms may be used in two ways. First the procedure
20028 forms can be used anywhere a procedure call would be valid, and
20029 correspond to what the RM calls ``intrinsic'' routines. Such calls can
20030 be used to intersperse machine instructions with other Ada statements.
20031 Second, the function forms, which return a dummy value of the limited
20032 private type @code{Asm_Insn}, can be used in code statements, and indeed
20033 this is the only context where such calls are allowed. Code statements
20034 appear as aggregates of the form:
20036 @smallexample @c ada
20037 Asm_Insn'(Asm (@dots{}));
20038 Asm_Insn'(Asm_Volatile (@dots{}));
20042 In accordance with RM rules, such code statements are allowed only
20043 within subprograms whose entire body consists of such statements. It is
20044 not permissible to intermix such statements with other Ada statements.
20046 Typically the form using intrinsic procedure calls is more convenient
20047 and more flexible. The code statement form is provided to meet the RM
20048 suggestion that such a facility should be made available. The following
20049 is the exact syntax of the call to @code{Asm}. As usual, if named notation
20050 is used, the arguments may be given in arbitrary order, following the
20051 normal rules for use of positional and named arguments)
20055 [Template =>] static_string_EXPRESSION
20056 [,[Outputs =>] OUTPUT_OPERAND_LIST ]
20057 [,[Inputs =>] INPUT_OPERAND_LIST ]
20058 [,[Clobber =>] static_string_EXPRESSION ]
20059 [,[Volatile =>] static_boolean_EXPRESSION] )
20061 OUTPUT_OPERAND_LIST ::=
20062 [PREFIX.]No_Output_Operands
20063 | OUTPUT_OPERAND_ATTRIBUTE
20064 | (OUTPUT_OPERAND_ATTRIBUTE @{,OUTPUT_OPERAND_ATTRIBUTE@})
20066 OUTPUT_OPERAND_ATTRIBUTE ::=
20067 SUBTYPE_MARK'Asm_Output (static_string_EXPRESSION, NAME)
20069 INPUT_OPERAND_LIST ::=
20070 [PREFIX.]No_Input_Operands
20071 | INPUT_OPERAND_ATTRIBUTE
20072 | (INPUT_OPERAND_ATTRIBUTE @{,INPUT_OPERAND_ATTRIBUTE@})
20074 INPUT_OPERAND_ATTRIBUTE ::=
20075 SUBTYPE_MARK'Asm_Input (static_string_EXPRESSION, EXPRESSION)
20079 The identifiers @code{No_Input_Operands} and @code{No_Output_Operands}
20080 are declared in the package @code{Machine_Code} and must be referenced
20081 according to normal visibility rules. In particular if there is no
20082 @code{use} clause for this package, then appropriate package name
20083 qualification is required.
20085 @node GNAT Implementation of Tasking
20086 @section GNAT Implementation of Tasking
20089 This chapter outlines the basic GNAT approach to tasking (in particular,
20090 a multi-layered library for portability) and discusses issues related
20091 to compliance with the Real-Time Systems Annex.
20094 * Mapping Ada Tasks onto the Underlying Kernel Threads::
20095 * Ensuring Compliance with the Real-Time Annex::
20098 @node Mapping Ada Tasks onto the Underlying Kernel Threads
20099 @subsection Mapping Ada Tasks onto the Underlying Kernel Threads
20102 GNAT's run-time support comprises two layers:
20105 @item GNARL (GNAT Run-time Layer)
20106 @item GNULL (GNAT Low-level Library)
20110 In GNAT, Ada's tasking services rely on a platform and OS independent
20111 layer known as GNARL@. This code is responsible for implementing the
20112 correct semantics of Ada's task creation, rendezvous, protected
20115 GNARL decomposes Ada's tasking semantics into simpler lower level
20116 operations such as create a thread, set the priority of a thread,
20117 yield, create a lock, lock/unlock, etc. The spec for these low-level
20118 operations constitutes GNULLI, the GNULL Interface. This interface is
20119 directly inspired from the POSIX real-time API@.
20121 If the underlying executive or OS implements the POSIX standard
20122 faithfully, the GNULL Interface maps as is to the services offered by
20123 the underlying kernel. Otherwise, some target dependent glue code maps
20124 the services offered by the underlying kernel to the semantics expected
20127 Whatever the underlying OS (VxWorks, UNIX, Windows, etc.) the
20128 key point is that each Ada task is mapped on a thread in the underlying
20129 kernel. For example, in the case of VxWorks, one Ada task = one VxWorks task.
20131 In addition Ada task priorities map onto the underlying thread priorities.
20132 Mapping Ada tasks onto the underlying kernel threads has several advantages:
20136 The underlying scheduler is used to schedule the Ada tasks. This
20137 makes Ada tasks as efficient as kernel threads from a scheduling
20141 Interaction with code written in C containing threads is eased
20142 since at the lowest level Ada tasks and C threads map onto the same
20143 underlying kernel concept.
20146 When an Ada task is blocked during I/O the remaining Ada tasks are
20150 On multiprocessor systems Ada tasks can execute in parallel.
20154 Some threads libraries offer a mechanism to fork a new process, with the
20155 child process duplicating the threads from the parent.
20157 support this functionality when the parent contains more than one task.
20158 @cindex Forking a new process
20160 @node Ensuring Compliance with the Real-Time Annex
20161 @subsection Ensuring Compliance with the Real-Time Annex
20162 @cindex Real-Time Systems Annex compliance
20165 Although mapping Ada tasks onto
20166 the underlying threads has significant advantages, it does create some
20167 complications when it comes to respecting the scheduling semantics
20168 specified in the real-time annex (Annex D).
20170 For instance the Annex D requirement for the @code{FIFO_Within_Priorities}
20171 scheduling policy states:
20174 @emph{When the active priority of a ready task that is not running
20175 changes, or the setting of its base priority takes effect, the
20176 task is removed from the ready queue for its old active priority
20177 and is added at the tail of the ready queue for its new active
20178 priority, except in the case where the active priority is lowered
20179 due to the loss of inherited priority, in which case the task is
20180 added at the head of the ready queue for its new active priority.}
20184 While most kernels do put tasks at the end of the priority queue when
20185 a task changes its priority, (which respects the main
20186 FIFO_Within_Priorities requirement), almost none keep a thread at the
20187 beginning of its priority queue when its priority drops from the loss
20188 of inherited priority.
20190 As a result most vendors have provided incomplete Annex D implementations.
20192 The GNAT run-time, has a nice cooperative solution to this problem
20193 which ensures that accurate FIFO_Within_Priorities semantics are
20196 The principle is as follows. When an Ada task T is about to start
20197 running, it checks whether some other Ada task R with the same
20198 priority as T has been suspended due to the loss of priority
20199 inheritance. If this is the case, T yields and is placed at the end of
20200 its priority queue. When R arrives at the front of the queue it
20203 Note that this simple scheme preserves the relative order of the tasks
20204 that were ready to execute in the priority queue where R has been
20207 @node GNAT Implementation of Shared Passive Packages
20208 @section GNAT Implementation of Shared Passive Packages
20209 @cindex Shared passive packages
20212 GNAT fully implements the pragma @code{Shared_Passive} for
20213 @cindex pragma @code{Shared_Passive}
20214 the purpose of designating shared passive packages.
20215 This allows the use of passive partitions in the
20216 context described in the Ada Reference Manual; i.e., for communication
20217 between separate partitions of a distributed application using the
20218 features in Annex E.
20220 @cindex Distribution Systems Annex
20222 However, the implementation approach used by GNAT provides for more
20223 extensive usage as follows:
20226 @item Communication between separate programs
20228 This allows separate programs to access the data in passive
20229 partitions, using protected objects for synchronization where
20230 needed. The only requirement is that the two programs have a
20231 common shared file system. It is even possible for programs
20232 running on different machines with different architectures
20233 (e.g.@: different endianness) to communicate via the data in
20234 a passive partition.
20236 @item Persistence between program runs
20238 The data in a passive package can persist from one run of a
20239 program to another, so that a later program sees the final
20240 values stored by a previous run of the same program.
20245 The implementation approach used is to store the data in files. A
20246 separate stream file is created for each object in the package, and
20247 an access to an object causes the corresponding file to be read or
20250 The environment variable @code{SHARED_MEMORY_DIRECTORY} should be
20251 @cindex @code{SHARED_MEMORY_DIRECTORY} environment variable
20252 set to the directory to be used for these files.
20253 The files in this directory
20254 have names that correspond to their fully qualified names. For
20255 example, if we have the package
20257 @smallexample @c ada
20259 pragma Shared_Passive (X);
20266 and the environment variable is set to @code{/stemp/}, then the files created
20267 will have the names:
20275 These files are created when a value is initially written to the object, and
20276 the files are retained until manually deleted. This provides the persistence
20277 semantics. If no file exists, it means that no partition has assigned a value
20278 to the variable; in this case the initial value declared in the package
20279 will be used. This model ensures that there are no issues in synchronizing
20280 the elaboration process, since elaboration of passive packages elaborates the
20281 initial values, but does not create the files.
20283 The files are written using normal @code{Stream_IO} access.
20284 If you want to be able
20285 to communicate between programs or partitions running on different
20286 architectures, then you should use the XDR versions of the stream attribute
20287 routines, since these are architecture independent.
20289 If active synchronization is required for access to the variables in the
20290 shared passive package, then as described in the Ada Reference Manual, the
20291 package may contain protected objects used for this purpose. In this case
20292 a lock file (whose name is @file{___lock} (three underscores)
20293 is created in the shared memory directory.
20294 @cindex @file{___lock} file (for shared passive packages)
20295 This is used to provide the required locking
20296 semantics for proper protected object synchronization.
20298 As of January 2003, GNAT supports shared passive packages on all platforms
20299 except for OpenVMS.
20301 @node Code Generation for Array Aggregates
20302 @section Code Generation for Array Aggregates
20305 * Static constant aggregates with static bounds::
20306 * Constant aggregates with unconstrained nominal types::
20307 * Aggregates with static bounds::
20308 * Aggregates with non-static bounds::
20309 * Aggregates in assignment statements::
20313 Aggregates have a rich syntax and allow the user to specify the values of
20314 complex data structures by means of a single construct. As a result, the
20315 code generated for aggregates can be quite complex and involve loops, case
20316 statements and multiple assignments. In the simplest cases, however, the
20317 compiler will recognize aggregates whose components and constraints are
20318 fully static, and in those cases the compiler will generate little or no
20319 executable code. The following is an outline of the code that GNAT generates
20320 for various aggregate constructs. For further details, you will find it
20321 useful to examine the output produced by the -gnatG flag to see the expanded
20322 source that is input to the code generator. You may also want to examine
20323 the assembly code generated at various levels of optimization.
20325 The code generated for aggregates depends on the context, the component values,
20326 and the type. In the context of an object declaration the code generated is
20327 generally simpler than in the case of an assignment. As a general rule, static
20328 component values and static subtypes also lead to simpler code.
20330 @node Static constant aggregates with static bounds
20331 @subsection Static constant aggregates with static bounds
20334 For the declarations:
20335 @smallexample @c ada
20336 type One_Dim is array (1..10) of integer;
20337 ar0 : constant One_Dim := (1, 2, 3, 4, 5, 6, 7, 8, 9, 0);
20341 GNAT generates no executable code: the constant ar0 is placed in static memory.
20342 The same is true for constant aggregates with named associations:
20344 @smallexample @c ada
20345 Cr1 : constant One_Dim := (4 => 16, 2 => 4, 3 => 9, 1 => 1, 5 .. 10 => 0);
20346 Cr3 : constant One_Dim := (others => 7777);
20350 The same is true for multidimensional constant arrays such as:
20352 @smallexample @c ada
20353 type two_dim is array (1..3, 1..3) of integer;
20354 Unit : constant two_dim := ( (1,0,0), (0,1,0), (0,0,1));
20358 The same is true for arrays of one-dimensional arrays: the following are
20361 @smallexample @c ada
20362 type ar1b is array (1..3) of boolean;
20363 type ar_ar is array (1..3) of ar1b;
20364 None : constant ar1b := (others => false); -- fully static
20365 None2 : constant ar_ar := (1..3 => None); -- fully static
20369 However, for multidimensional aggregates with named associations, GNAT will
20370 generate assignments and loops, even if all associations are static. The
20371 following two declarations generate a loop for the first dimension, and
20372 individual component assignments for the second dimension:
20374 @smallexample @c ada
20375 Zero1: constant two_dim := (1..3 => (1..3 => 0));
20376 Zero2: constant two_dim := (others => (others => 0));
20379 @node Constant aggregates with unconstrained nominal types
20380 @subsection Constant aggregates with unconstrained nominal types
20383 In such cases the aggregate itself establishes the subtype, so that
20384 associations with @code{others} cannot be used. GNAT determines the
20385 bounds for the actual subtype of the aggregate, and allocates the
20386 aggregate statically as well. No code is generated for the following:
20388 @smallexample @c ada
20389 type One_Unc is array (natural range <>) of integer;
20390 Cr_Unc : constant One_Unc := (12,24,36);
20393 @node Aggregates with static bounds
20394 @subsection Aggregates with static bounds
20397 In all previous examples the aggregate was the initial (and immutable) value
20398 of a constant. If the aggregate initializes a variable, then code is generated
20399 for it as a combination of individual assignments and loops over the target
20400 object. The declarations
20402 @smallexample @c ada
20403 Cr_Var1 : One_Dim := (2, 5, 7, 11, 0, 0, 0, 0, 0, 0);
20404 Cr_Var2 : One_Dim := (others > -1);
20408 generate the equivalent of
20410 @smallexample @c ada
20416 for I in Cr_Var2'range loop
20421 @node Aggregates with non-static bounds
20422 @subsection Aggregates with non-static bounds
20425 If the bounds of the aggregate are not statically compatible with the bounds
20426 of the nominal subtype of the target, then constraint checks have to be
20427 generated on the bounds. For a multidimensional array, constraint checks may
20428 have to be applied to sub-arrays individually, if they do not have statically
20429 compatible subtypes.
20431 @node Aggregates in assignment statements
20432 @subsection Aggregates in assignment statements
20435 In general, aggregate assignment requires the construction of a temporary,
20436 and a copy from the temporary to the target of the assignment. This is because
20437 it is not always possible to convert the assignment into a series of individual
20438 component assignments. For example, consider the simple case:
20440 @smallexample @c ada
20445 This cannot be converted into:
20447 @smallexample @c ada
20453 So the aggregate has to be built first in a separate location, and then
20454 copied into the target. GNAT recognizes simple cases where this intermediate
20455 step is not required, and the assignments can be performed in place, directly
20456 into the target. The following sufficient criteria are applied:
20460 The bounds of the aggregate are static, and the associations are static.
20462 The components of the aggregate are static constants, names of
20463 simple variables that are not renamings, or expressions not involving
20464 indexed components whose operands obey these rules.
20468 If any of these conditions are violated, the aggregate will be built in
20469 a temporary (created either by the front-end or the code generator) and then
20470 that temporary will be copied onto the target.
20472 @node The Size of Discriminated Records with Default Discriminants
20473 @section The Size of Discriminated Records with Default Discriminants
20476 If a discriminated type @code{T} has discriminants with default values, it is
20477 possible to declare an object of this type without providing an explicit
20480 @smallexample @c ada
20482 type Size is range 1..100;
20484 type Rec (D : Size := 15) is record
20485 Name : String (1..D);
20493 Such an object is said to be @emph{unconstrained}.
20494 The discriminant of the object
20495 can be modified by a full assignment to the object, as long as it preserves the
20496 relation between the value of the discriminant, and the value of the components
20499 @smallexample @c ada
20501 Word := (3, "yes");
20503 Word := (5, "maybe");
20505 Word := (5, "no"); -- raises Constraint_Error
20510 In order to support this behavior efficiently, an unconstrained object is
20511 given the maximum size that any value of the type requires. In the case
20512 above, @code{Word} has storage for the discriminant and for
20513 a @code{String} of length 100.
20514 It is important to note that unconstrained objects do not require dynamic
20515 allocation. It would be an improper implementation to place on the heap those
20516 components whose size depends on discriminants. (This improper implementation
20517 was used by some Ada83 compilers, where the @code{Name} component above
20519 been stored as a pointer to a dynamic string). Following the principle that
20520 dynamic storage management should never be introduced implicitly,
20521 an Ada compiler should reserve the full size for an unconstrained declared
20522 object, and place it on the stack.
20524 This maximum size approach
20525 has been a source of surprise to some users, who expect the default
20526 values of the discriminants to determine the size reserved for an
20527 unconstrained object: ``If the default is 15, why should the object occupy
20529 The answer, of course, is that the discriminant may be later modified,
20530 and its full range of values must be taken into account. This is why the
20535 type Rec (D : Positive := 15) is record
20536 Name : String (1..D);
20544 is flagged by the compiler with a warning:
20545 an attempt to create @code{Too_Large} will raise @code{Storage_Error},
20546 because the required size includes @code{Positive'Last}
20547 bytes. As the first example indicates, the proper approach is to declare an
20548 index type of ``reasonable'' range so that unconstrained objects are not too
20551 One final wrinkle: if the object is declared to be @code{aliased}, or if it is
20552 created in the heap by means of an allocator, then it is @emph{not}
20554 it is constrained by the default values of the discriminants, and those values
20555 cannot be modified by full assignment. This is because in the presence of
20556 aliasing all views of the object (which may be manipulated by different tasks,
20557 say) must be consistent, so it is imperative that the object, once created,
20560 @node Strict Conformance to the Ada Reference Manual
20561 @section Strict Conformance to the Ada Reference Manual
20564 The dynamic semantics defined by the Ada Reference Manual impose a set of
20565 run-time checks to be generated. By default, the GNAT compiler will insert many
20566 run-time checks into the compiled code, including most of those required by the
20567 Ada Reference Manual. However, there are three checks that are not enabled
20568 in the default mode for efficiency reasons: arithmetic overflow checking for
20569 integer operations (including division by zero), checks for access before
20570 elaboration on subprogram calls, and stack overflow checking (most operating
20571 systems do not perform this check by default).
20573 Strict conformance to the Ada Reference Manual can be achieved by adding
20574 three compiler options for overflow checking for integer operations
20575 (@option{-gnato}), dynamic checks for access-before-elaboration on subprogram
20576 calls and generic instantiations (@option{-gnatE}), and stack overflow
20577 checking (@option{-fstack-check}).
20579 Note that the result of a floating point arithmetic operation in overflow and
20580 invalid situations, when the @code{Machine_Overflows} attribute of the result
20581 type is @code{False}, is to generate IEEE NaN and infinite values. This is the
20582 case for machines compliant with the IEEE floating-point standard, but on
20583 machines that are not fully compliant with this standard, such as Alpha, the
20584 @option{-mieee} compiler flag must be used for achieving IEEE confirming
20585 behavior (although at the cost of a significant performance penalty), so
20586 infinite and NaN values are properly generated.
20589 @node Implementation of Ada 2012 Features
20590 @chapter Implementation of Ada 2012 Features
20591 @cindex Ada 2012 implementation status
20593 This chapter contains a complete list of Ada 2012 features that have been
20594 implemented as of GNAT version 6.4. Generally, these features are only
20595 available if the @option{-gnat12} (Ada 2012 features enabled) flag is set
20596 @cindex @option{-gnat12} option
20597 or if the configuration pragma @code{Ada_2012} is used.
20598 @cindex pragma @code{Ada_2012}
20599 @cindex configuration pragma @code{Ada_2012}
20600 @cindex @code{Ada_2012} configuration pragma
20601 However, new pragmas, attributes, and restrictions are
20602 unconditionally available, since the Ada 95 standard allows the addition of
20603 new pragmas, attributes, and restrictions (there are exceptions, which are
20604 documented in the individual descriptions), and also certain packages
20605 were made available in earlier versions of Ada.
20607 An ISO date (YYYY-MM-DD) appears in parentheses on the description line.
20608 This date shows the implementation date of the feature. Any wavefront
20609 subsequent to this date will contain the indicated feature, as will any
20610 subsequent releases. A date of 0000-00-00 means that GNAT has always
20611 implemented the feature, or implemented it as soon as it appeared as a
20612 binding interpretation.
20614 Each feature corresponds to an Ada Issue (``AI'') approved by the Ada
20615 standardization group (ISO/IEC JTC1/SC22/WG9) for inclusion in Ada 2012.
20616 The features are ordered based on the relevant sections of the Ada
20617 Reference Manual (``RM''). When a given AI relates to multiple points
20618 in the RM, the earliest is used.
20620 A complete description of the AIs may be found in
20621 @url{www.ada-auth.org/ai05-summary.html}.
20626 @emph{AI-0176 Quantified expressions (2010-09-29)}
20627 @cindex AI-0176 (Ada 2012 feature)
20630 Both universally and existentially quantified expressions are implemented.
20631 They use the new syntax for iterators proposed in AI05-139-2, as well as
20632 the standard Ada loop syntax.
20635 RM References: 1.01.04 (12) 2.09 (2/2) 4.04 (7) 4.05.09 (0)
20638 @emph{AI-0079 Allow @i{other_format} characters in source (2010-07-10)}
20639 @cindex AI-0079 (Ada 2012 feature)
20642 Wide characters in the unicode category @i{other_format} are now allowed in
20643 source programs between tokens, but not within a token such as an identifier.
20646 RM References: 2.01 (4/2) 2.02 (7)
20649 @emph{AI-0091 Do not allow @i{other_format} in identifiers (0000-00-00)}
20650 @cindex AI-0091 (Ada 2012 feature)
20653 Wide characters in the unicode category @i{other_format} are not permitted
20654 within an identifier, since this can be a security problem. The error
20655 message for this case has been improved to be more specific, but GNAT has
20656 never allowed such characters to appear in identifiers.
20659 RM References: 2.03 (3.1/2) 2.03 (4/2) 2.03 (5/2) 2.03 (5.1/2) 2.03 (5.2/2) 2.03 (5.3/2) 2.09 (2/2)
20662 @emph{AI-0100 Placement of pragmas (2010-07-01)}
20663 @cindex AI-0100 (Ada 2012 feature)
20666 This AI is an earlier version of AI-163. It simplifies the rules
20667 for legal placement of pragmas. In the case of lists that allow pragmas, if
20668 the list may have no elements, then the list may consist solely of pragmas.
20671 RM References: 2.08 (7)
20674 @emph{AI-0163 Pragmas in place of null (2010-07-01)}
20675 @cindex AI-0163 (Ada 2012 feature)
20678 A statement sequence may be composed entirely of pragmas. It is no longer
20679 necessary to add a dummy @code{null} statement to make the sequence legal.
20682 RM References: 2.08 (7) 2.08 (16)
20686 @emph{AI-0080 ``View of'' not needed if clear from context (0000-00-00)}
20687 @cindex AI-0080 (Ada 2012 feature)
20690 This is an editorial change only, described as non-testable in the AI.
20693 RM References: 3.01 (7)
20697 @emph{AI-0183 Aspect specifications (2010-08-16)}
20698 @cindex AI-0183 (Ada 2012 feature)
20701 Aspect specifications have been fully implemented except for pre and post-
20702 conditions, and type invariants, which have their own separate AI's. All
20703 forms of declarations listed in the AI are supported. The following is a
20704 list of the aspects supported (with GNAT implementation aspects marked)
20706 @multitable {@code{Preelaborable_Initialization}} {--GNAT}
20707 @item @code{Ada_2005} @tab -- GNAT
20708 @item @code{Ada_2012} @tab -- GNAT
20709 @item @code{Address} @tab
20710 @item @code{Alignment} @tab
20711 @item @code{Atomic} @tab
20712 @item @code{Atomic_Components} @tab
20713 @item @code{Bit_Order} @tab
20714 @item @code{Component_Size} @tab
20715 @item @code{Contract_Cases} @tab -- GNAT
20716 @item @code{Discard_Names} @tab
20717 @item @code{External_Tag} @tab
20718 @item @code{Favor_Top_Level} @tab -- GNAT
20719 @item @code{Inline} @tab
20720 @item @code{Inline_Always} @tab -- GNAT
20721 @item @code{Invariant} @tab -- GNAT
20722 @item @code{Machine_Radix} @tab
20723 @item @code{No_Return} @tab
20724 @item @code{Object_Size} @tab -- GNAT
20725 @item @code{Pack} @tab
20726 @item @code{Persistent_BSS} @tab -- GNAT
20727 @item @code{Post} @tab
20728 @item @code{Pre} @tab
20729 @item @code{Predicate} @tab
20730 @item @code{Preelaborable_Initialization} @tab
20731 @item @code{Pure_Function} @tab -- GNAT
20732 @item @code{Remote_Access_Type} @tab -- GNAT
20733 @item @code{Shared} @tab -- GNAT
20734 @item @code{Size} @tab
20735 @item @code{Storage_Pool} @tab
20736 @item @code{Storage_Size} @tab
20737 @item @code{Stream_Size} @tab
20738 @item @code{Suppress} @tab
20739 @item @code{Suppress_Debug_Info} @tab -- GNAT
20740 @item @code{Test_Case} @tab -- GNAT
20741 @item @code{Type_Invariant} @tab
20742 @item @code{Unchecked_Union} @tab
20743 @item @code{Universal_Aliasing} @tab -- GNAT
20744 @item @code{Unmodified} @tab -- GNAT
20745 @item @code{Unreferenced} @tab -- GNAT
20746 @item @code{Unreferenced_Objects} @tab -- GNAT
20747 @item @code{Unsuppress} @tab
20748 @item @code{Value_Size} @tab -- GNAT
20749 @item @code{Volatile} @tab
20750 @item @code{Volatile_Components}
20751 @item @code{Warnings} @tab -- GNAT
20755 Note that for aspects with an expression, e.g. @code{Size}, the expression is
20756 treated like a default expression (visibility is analyzed at the point of
20757 occurrence of the aspect, but evaluation of the expression occurs at the
20758 freeze point of the entity involved).
20761 RM References: 3.02.01 (3) 3.02.02 (2) 3.03.01 (2/2) 3.08 (6)
20762 3.09.03 (1.1/2) 6.01 (2/2) 6.07 (2/2) 9.05.02 (2/2) 7.01 (3) 7.03
20763 (2) 7.03 (3) 9.01 (2/2) 9.01 (3/2) 9.04 (2/2) 9.04 (3/2)
20764 9.05.02 (2/2) 11.01 (2) 12.01 (3) 12.03 (2/2) 12.04 (2/2) 12.05 (2)
20765 12.06 (2.1/2) 12.06 (2.2/2) 12.07 (2) 13.01 (0.1/2) 13.03 (5/1)
20770 @emph{AI-0128 Inequality is a primitive operation (0000-00-00)}
20771 @cindex AI-0128 (Ada 2012 feature)
20774 If an equality operator ("=") is declared for a type, then the implicitly
20775 declared inequality operator ("/=") is a primitive operation of the type.
20776 This is the only reasonable interpretation, and is the one always implemented
20777 by GNAT, but the RM was not entirely clear in making this point.
20780 RM References: 3.02.03 (6) 6.06 (6)
20783 @emph{AI-0003 Qualified expressions as names (2010-07-11)}
20784 @cindex AI-0003 (Ada 2012 feature)
20787 In Ada 2012, a qualified expression is considered to be syntactically a name,
20788 meaning that constructs such as @code{A'(F(X)).B} are now legal. This is
20789 useful in disambiguating some cases of overloading.
20792 RM References: 3.03 (11) 3.03 (21) 4.01 (2) 4.04 (7) 4.07 (3)
20796 @emph{AI-0120 Constant instance of protected object (0000-00-00)}
20797 @cindex AI-0120 (Ada 2012 feature)
20800 This is an RM editorial change only. The section that lists objects that are
20801 constant failed to include the current instance of a protected object
20802 within a protected function. This has always been treated as a constant
20806 RM References: 3.03 (21)
20809 @emph{AI-0008 General access to constrained objects (0000-00-00)}
20810 @cindex AI-0008 (Ada 2012 feature)
20813 The wording in the RM implied that if you have a general access to a
20814 constrained object, it could be used to modify the discriminants. This was
20815 obviously not intended. @code{Constraint_Error} should be raised, and GNAT
20816 has always done so in this situation.
20819 RM References: 3.03 (23) 3.10.02 (26/2) 4.01 (9) 6.04.01 (17) 8.05.01 (5/2)
20823 @emph{AI-0093 Additional rules use immutably limited (0000-00-00)}
20824 @cindex AI-0093 (Ada 2012 feature)
20827 This is an editorial change only, to make more widespread use of the Ada 2012
20828 ``immutably limited''.
20831 RM References: 3.03 (23.4/3)
20836 @emph{AI-0096 Deriving from formal private types (2010-07-20)}
20837 @cindex AI-0096 (Ada 2012 feature)
20840 In general it is illegal for a type derived from a formal limited type to be
20841 nonlimited. This AI makes an exception to this rule: derivation is legal
20842 if it appears in the private part of the generic, and the formal type is not
20843 tagged. If the type is tagged, the legality check must be applied to the
20844 private part of the package.
20847 RM References: 3.04 (5.1/2) 6.02 (7)
20851 @emph{AI-0181 Soft hyphen is a non-graphic character (2010-07-23)}
20852 @cindex AI-0181 (Ada 2012 feature)
20855 From Ada 2005 on, soft hyphen is considered a non-graphic character, which
20856 means that it has a special name (@code{SOFT_HYPHEN}) in conjunction with the
20857 @code{Image} and @code{Value} attributes for the character types. Strictly
20858 speaking this is an inconsistency with Ada 95, but in practice the use of
20859 these attributes is so obscure that it will not cause problems.
20862 RM References: 3.05.02 (2/2) A.01 (35/2) A.03.03 (21)
20866 @emph{AI-0182 Additional forms for @code{Character'Value} (0000-00-00)}
20867 @cindex AI-0182 (Ada 2012 feature)
20870 This AI allows @code{Character'Value} to accept the string @code{'?'} where
20871 @code{?} is any character including non-graphic control characters. GNAT has
20872 always accepted such strings. It also allows strings such as
20873 @code{HEX_00000041} to be accepted, but GNAT does not take advantage of this
20874 permission and raises @code{Constraint_Error}, as is certainly still
20878 RM References: 3.05 (56/2)
20882 @emph{AI-0214 Defaulted discriminants for limited tagged (2010-10-01)}
20883 @cindex AI-0214 (Ada 2012 feature)
20886 Ada 2012 relaxes the restriction that forbids discriminants of tagged types
20887 to have default expressions by allowing them when the type is limited. It
20888 is often useful to define a default value for a discriminant even though
20889 it can't be changed by assignment.
20892 RM References: 3.07 (9.1/2) 3.07.02 (3)
20896 @emph{AI-0102 Some implicit conversions are illegal (0000-00-00)}
20897 @cindex AI-0102 (Ada 2012 feature)
20900 It is illegal to assign an anonymous access constant to an anonymous access
20901 variable. The RM did not have a clear rule to prevent this, but GNAT has
20902 always generated an error for this usage.
20905 RM References: 3.07 (16) 3.07.01 (9) 6.04.01 (6) 8.06 (27/2)
20909 @emph{AI-0158 Generalizing membership tests (2010-09-16)}
20910 @cindex AI-0158 (Ada 2012 feature)
20913 This AI extends the syntax of membership tests to simplify complex conditions
20914 that can be expressed as membership in a subset of values of any type. It
20915 introduces syntax for a list of expressions that may be used in loop contexts
20919 RM References: 3.08.01 (5) 4.04 (3) 4.05.02 (3) 4.05.02 (5) 4.05.02 (27)
20923 @emph{AI-0173 Testing if tags represent abstract types (2010-07-03)}
20924 @cindex AI-0173 (Ada 2012 feature)
20927 The function @code{Ada.Tags.Type_Is_Abstract} returns @code{True} if invoked
20928 with the tag of an abstract type, and @code{False} otherwise.
20931 RM References: 3.09 (7.4/2) 3.09 (12.4/2)
20936 @emph{AI-0076 function with controlling result (0000-00-00)}
20937 @cindex AI-0076 (Ada 2012 feature)
20940 This is an editorial change only. The RM defines calls with controlling
20941 results, but uses the term ``function with controlling result'' without an
20942 explicit definition.
20945 RM References: 3.09.02 (2/2)
20949 @emph{AI-0126 Dispatching with no declared operation (0000-00-00)}
20950 @cindex AI-0126 (Ada 2012 feature)
20953 This AI clarifies dispatching rules, and simply confirms that dispatching
20954 executes the operation of the parent type when there is no explicitly or
20955 implicitly declared operation for the descendant type. This has always been
20956 the case in all versions of GNAT.
20959 RM References: 3.09.02 (20/2) 3.09.02 (20.1/2) 3.09.02 (20.2/2)
20963 @emph{AI-0097 Treatment of abstract null extension (2010-07-19)}
20964 @cindex AI-0097 (Ada 2012 feature)
20967 The RM as written implied that in some cases it was possible to create an
20968 object of an abstract type, by having an abstract extension inherit a non-
20969 abstract constructor from its parent type. This mistake has been corrected
20970 in GNAT and in the RM, and this construct is now illegal.
20973 RM References: 3.09.03 (4/2)
20977 @emph{AI-0203 Extended return cannot be abstract (0000-00-00)}
20978 @cindex AI-0203 (Ada 2012 feature)
20981 A return_subtype_indication cannot denote an abstract subtype. GNAT has never
20982 permitted such usage.
20985 RM References: 3.09.03 (8/3)
20989 @emph{AI-0198 Inheriting abstract operators (0000-00-00)}
20990 @cindex AI-0198 (Ada 2012 feature)
20993 This AI resolves a conflict between two rules involving inherited abstract
20994 operations and predefined operators. If a derived numeric type inherits
20995 an abstract operator, it overrides the predefined one. This interpretation
20996 was always the one implemented in GNAT.
20999 RM References: 3.09.03 (4/3)
21002 @emph{AI-0073 Functions returning abstract types (2010-07-10)}
21003 @cindex AI-0073 (Ada 2012 feature)
21006 This AI covers a number of issues regarding returning abstract types. In
21007 particular generic functions cannot have abstract result types or access
21008 result types designated an abstract type. There are some other cases which
21009 are detailed in the AI. Note that this binding interpretation has not been
21010 retrofitted to operate before Ada 2012 mode, since it caused a significant
21011 number of regressions.
21014 RM References: 3.09.03 (8) 3.09.03 (10) 6.05 (8/2)
21018 @emph{AI-0070 Elaboration of interface types (0000-00-00)}
21019 @cindex AI-0070 (Ada 2012 feature)
21022 This is an editorial change only, there are no testable consequences short of
21023 checking for the absence of generated code for an interface declaration.
21026 RM References: 3.09.04 (18/2)
21030 @emph{AI-0208 Characteristics of incomplete views (0000-00-00)}
21031 @cindex AI-0208 (Ada 2012 feature)
21034 The wording in the Ada 2005 RM concerning characteristics of incomplete views
21035 was incorrect and implied that some programs intended to be legal were now
21036 illegal. GNAT had never considered such programs illegal, so it has always
21037 implemented the intent of this AI.
21040 RM References: 3.10.01 (2.4/2) 3.10.01 (2.6/2)
21044 @emph{AI-0162 Incomplete type completed by partial view (2010-09-15)}
21045 @cindex AI-0162 (Ada 2012 feature)
21048 Incomplete types are made more useful by allowing them to be completed by
21049 private types and private extensions.
21052 RM References: 3.10.01 (2.5/2) 3.10.01 (2.6/2) 3.10.01 (3) 3.10.01 (4/2)
21057 @emph{AI-0098 Anonymous subprogram access restrictions (0000-00-00)}
21058 @cindex AI-0098 (Ada 2012 feature)
21061 An unintentional omission in the RM implied some inconsistent restrictions on
21062 the use of anonymous access to subprogram values. These restrictions were not
21063 intentional, and have never been enforced by GNAT.
21066 RM References: 3.10.01 (6) 3.10.01 (9.2/2)
21070 @emph{AI-0199 Aggregate with anonymous access components (2010-07-14)}
21071 @cindex AI-0199 (Ada 2012 feature)
21074 A choice list in a record aggregate can include several components of
21075 (distinct) anonymous access types as long as they have matching designated
21079 RM References: 4.03.01 (16)
21083 @emph{AI-0220 Needed components for aggregates (0000-00-00)}
21084 @cindex AI-0220 (Ada 2012 feature)
21087 This AI addresses a wording problem in the RM that appears to permit some
21088 complex cases of aggregates with non-static discriminants. GNAT has always
21089 implemented the intended semantics.
21092 RM References: 4.03.01 (17)
21095 @emph{AI-0147 Conditional expressions (2009-03-29)}
21096 @cindex AI-0147 (Ada 2012 feature)
21099 Conditional expressions are permitted. The form of such an expression is:
21102 (@b{if} @i{expr} @b{then} @i{expr} @{@b{elsif} @i{expr} @b{then} @i{expr}@} [@b{else} @i{expr}])
21105 The parentheses can be omitted in contexts where parentheses are present
21106 anyway, such as subprogram arguments and pragma arguments. If the @b{else}
21107 clause is omitted, @b{else True} is assumed;
21108 thus @code{(@b{if} A @b{then} B)} is a way to conveniently represent
21109 @emph{(A implies B)} in standard logic.
21112 RM References: 4.03.03 (15) 4.04 (1) 4.04 (7) 4.05.07 (0) 4.07 (2)
21113 4.07 (3) 4.09 (12) 4.09 (33) 5.03 (3) 5.03 (4) 7.05 (2.1/2)
21117 @emph{AI-0037 Out-of-range box associations in aggregate (0000-00-00)}
21118 @cindex AI-0037 (Ada 2012 feature)
21121 This AI confirms that an association of the form @code{Indx => <>} in an
21122 array aggregate must raise @code{Constraint_Error} if @code{Indx}
21123 is out of range. The RM specified a range check on other associations, but
21124 not when the value of the association was defaulted. GNAT has always inserted
21125 a constraint check on the index value.
21128 RM References: 4.03.03 (29)
21132 @emph{AI-0123 Composability of equality (2010-04-13)}
21133 @cindex AI-0123 (Ada 2012 feature)
21136 Equality of untagged record composes, so that the predefined equality for a
21137 composite type that includes a component of some untagged record type
21138 @code{R} uses the equality operation of @code{R} (which may be user-defined
21139 or predefined). This makes the behavior of untagged records identical to that
21140 of tagged types in this respect.
21142 This change is an incompatibility with previous versions of Ada, but it
21143 corrects a non-uniformity that was often a source of confusion. Analysis of
21144 a large number of industrial programs indicates that in those rare cases
21145 where a composite type had an untagged record component with a user-defined
21146 equality, either there was no use of the composite equality, or else the code
21147 expected the same composability as for tagged types, and thus had a bug that
21148 would be fixed by this change.
21151 RM References: 4.05.02 (9.7/2) 4.05.02 (14) 4.05.02 (15) 4.05.02 (24)
21156 @emph{AI-0088 The value of exponentiation (0000-00-00)}
21157 @cindex AI-0088 (Ada 2012 feature)
21160 This AI clarifies the equivalence rule given for the dynamic semantics of
21161 exponentiation: the value of the operation can be obtained by repeated
21162 multiplication, but the operation can be implemented otherwise (for example
21163 using the familiar divide-by-two-and-square algorithm, even if this is less
21164 accurate), and does not imply repeated reads of a volatile base.
21167 RM References: 4.05.06 (11)
21170 @emph{AI-0188 Case expressions (2010-01-09)}
21171 @cindex AI-0188 (Ada 2012 feature)
21174 Case expressions are permitted. This allows use of constructs such as:
21176 X := (@b{case} Y @b{is when} 1 => 2, @b{when} 2 => 3, @b{when others} => 31)
21180 RM References: 4.05.07 (0) 4.05.08 (0) 4.09 (12) 4.09 (33)
21183 @emph{AI-0104 Null exclusion and uninitialized allocator (2010-07-15)}
21184 @cindex AI-0104 (Ada 2012 feature)
21187 The assignment @code{Ptr := @b{new not null} Some_Ptr;} will raise
21188 @code{Constraint_Error} because the default value of the allocated object is
21189 @b{null}. This useless construct is illegal in Ada 2012.
21192 RM References: 4.08 (2)
21195 @emph{AI-0157 Allocation/Deallocation from empty pool (2010-07-11)}
21196 @cindex AI-0157 (Ada 2012 feature)
21199 Allocation and Deallocation from an empty storage pool (i.e. allocation or
21200 deallocation of a pointer for which a static storage size clause of zero
21201 has been given) is now illegal and is detected as such. GNAT
21202 previously gave a warning but not an error.
21205 RM References: 4.08 (5.3/2) 13.11.02 (4) 13.11.02 (17)
21208 @emph{AI-0179 Statement not required after label (2010-04-10)}
21209 @cindex AI-0179 (Ada 2012 feature)
21212 It is not necessary to have a statement following a label, so a label
21213 can appear at the end of a statement sequence without the need for putting a
21214 null statement afterwards, but it is not allowable to have only labels and
21215 no real statements in a statement sequence.
21218 RM References: 5.01 (2)
21222 @emph{AI-139-2 Syntactic sugar for iterators (2010-09-29)}
21223 @cindex AI-139-2 (Ada 2012 feature)
21226 The new syntax for iterating over arrays and containers is now implemented.
21227 Iteration over containers is for now limited to read-only iterators. Only
21228 default iterators are supported, with the syntax: @code{@b{for} Elem @b{of} C}.
21231 RM References: 5.05
21234 @emph{AI-0134 Profiles must match for full conformance (0000-00-00)}
21235 @cindex AI-0134 (Ada 2012 feature)
21238 For full conformance, the profiles of anonymous-access-to-subprogram
21239 parameters must match. GNAT has always enforced this rule.
21242 RM References: 6.03.01 (18)
21245 @emph{AI-0207 Mode conformance and access constant (0000-00-00)}
21246 @cindex AI-0207 (Ada 2012 feature)
21249 This AI confirms that access_to_constant indication must match for mode
21250 conformance. This was implemented in GNAT when the qualifier was originally
21251 introduced in Ada 2005.
21254 RM References: 6.03.01 (16/2)
21258 @emph{AI-0046 Null exclusion match for full conformance (2010-07-17)}
21259 @cindex AI-0046 (Ada 2012 feature)
21262 For full conformance, in the case of access parameters, the null exclusion
21263 must match (either both or neither must have @code{@b{not null}}).
21266 RM References: 6.03.02 (18)
21270 @emph{AI-0118 The association of parameter associations (0000-00-00)}
21271 @cindex AI-0118 (Ada 2012 feature)
21274 This AI clarifies the rules for named associations in subprogram calls and
21275 generic instantiations. The rules have been in place since Ada 83.
21278 RM References: 6.04.01 (2) 12.03 (9)
21282 @emph{AI-0196 Null exclusion tests for out parameters (0000-00-00)}
21283 @cindex AI-0196 (Ada 2012 feature)
21286 Null exclusion checks are not made for @code{@b{out}} parameters when
21287 evaluating the actual parameters. GNAT has never generated these checks.
21290 RM References: 6.04.01 (13)
21293 @emph{AI-0015 Constant return objects (0000-00-00)}
21294 @cindex AI-0015 (Ada 2012 feature)
21297 The return object declared in an @i{extended_return_statement} may be
21298 declared constant. This was always intended, and GNAT has always allowed it.
21301 RM References: 6.05 (2.1/2) 3.03 (10/2) 3.03 (21) 6.05 (5/2)
21306 @emph{AI-0032 Extended return for class-wide functions (0000-00-00)}
21307 @cindex AI-0032 (Ada 2012 feature)
21310 If a function returns a class-wide type, the object of an extended return
21311 statement can be declared with a specific type that is covered by the class-
21312 wide type. This has been implemented in GNAT since the introduction of
21313 extended returns. Note AI-0103 complements this AI by imposing matching
21314 rules for constrained return types.
21317 RM References: 6.05 (5.2/2) 6.05 (5.3/2) 6.05 (5.6/2) 6.05 (5.8/2)
21321 @emph{AI-0103 Static matching for extended return (2010-07-23)}
21322 @cindex AI-0103 (Ada 2012 feature)
21325 If the return subtype of a function is an elementary type or a constrained
21326 type, the subtype indication in an extended return statement must match
21327 statically this return subtype.
21330 RM References: 6.05 (5.2/2)
21334 @emph{AI-0058 Abnormal completion of an extended return (0000-00-00)}
21335 @cindex AI-0058 (Ada 2012 feature)
21338 The RM had some incorrect wording implying wrong treatment of abnormal
21339 completion in an extended return. GNAT has always implemented the intended
21340 correct semantics as described by this AI.
21343 RM References: 6.05 (22/2)
21347 @emph{AI-0050 Raising Constraint_Error early for function call (0000-00-00)}
21348 @cindex AI-0050 (Ada 2012 feature)
21351 The implementation permissions for raising @code{Constraint_Error} early on a function call when it was clear an exception would be raised were over-permissive and allowed mishandling of discriminants in some cases. GNAT did
21352 not take advantage of these incorrect permissions in any case.
21355 RM References: 6.05 (24/2)
21359 @emph{AI-0125 Nonoverridable operations of an ancestor (2010-09-28)}
21360 @cindex AI-0125 (Ada 2012 feature)
21363 In Ada 2012, the declaration of a primitive operation of a type extension
21364 or private extension can also override an inherited primitive that is not
21365 visible at the point of this declaration.
21368 RM References: 7.03.01 (6) 8.03 (23) 8.03.01 (5/2) 8.03.01 (6/2)
21371 @emph{AI-0062 Null exclusions and deferred constants (0000-00-00)}
21372 @cindex AI-0062 (Ada 2012 feature)
21375 A full constant may have a null exclusion even if its associated deferred
21376 constant does not. GNAT has always allowed this.
21379 RM References: 7.04 (6/2) 7.04 (7.1/2)
21383 @emph{AI-0178 Incomplete views are limited (0000-00-00)}
21384 @cindex AI-0178 (Ada 2012 feature)
21387 This AI clarifies the role of incomplete views and plugs an omission in the
21388 RM. GNAT always correctly restricted the use of incomplete views and types.
21391 RM References: 7.05 (3/2) 7.05 (6/2)
21394 @emph{AI-0087 Actual for formal nonlimited derived type (2010-07-15)}
21395 @cindex AI-0087 (Ada 2012 feature)
21398 The actual for a formal nonlimited derived type cannot be limited. In
21399 particular, a formal derived type that extends a limited interface but which
21400 is not explicitly limited cannot be instantiated with a limited type.
21403 RM References: 7.05 (5/2) 12.05.01 (5.1/2)
21406 @emph{AI-0099 Tag determines whether finalization needed (0000-00-00)}
21407 @cindex AI-0099 (Ada 2012 feature)
21410 This AI clarifies that ``needs finalization'' is part of dynamic semantics,
21411 and therefore depends on the run-time characteristics of an object (i.e. its
21412 tag) and not on its nominal type. As the AI indicates: ``we do not expect
21413 this to affect any implementation''.
21416 RM References: 7.06.01 (6) 7.06.01 (7) 7.06.01 (8) 7.06.01 (9/2)
21421 @emph{AI-0064 Redundant finalization rule (0000-00-00)}
21422 @cindex AI-0064 (Ada 2012 feature)
21425 This is an editorial change only. The intended behavior is already checked
21426 by an existing ACATS test, which GNAT has always executed correctly.
21429 RM References: 7.06.01 (17.1/1)
21432 @emph{AI-0026 Missing rules for Unchecked_Union (2010-07-07)}
21433 @cindex AI-0026 (Ada 2012 feature)
21436 Record representation clauses concerning Unchecked_Union types cannot mention
21437 the discriminant of the type. The type of a component declared in the variant
21438 part of an Unchecked_Union cannot be controlled, have controlled components,
21439 nor have protected or task parts. If an Unchecked_Union type is declared
21440 within the body of a generic unit or its descendants, then the type of a
21441 component declared in the variant part cannot be a formal private type or a
21442 formal private extension declared within the same generic unit.
21445 RM References: 7.06 (9.4/2) B.03.03 (9/2) B.03.03 (10/2)
21449 @emph{AI-0205 Extended return declares visible name (0000-00-00)}
21450 @cindex AI-0205 (Ada 2012 feature)
21453 This AI corrects a simple omission in the RM. Return objects have always
21454 been visible within an extended return statement.
21457 RM References: 8.03 (17)
21461 @emph{AI-0042 Overriding versus implemented-by (0000-00-00)}
21462 @cindex AI-0042 (Ada 2012 feature)
21465 This AI fixes a wording gap in the RM. An operation of a synchronized
21466 interface can be implemented by a protected or task entry, but the abstract
21467 operation is not being overridden in the usual sense, and it must be stated
21468 separately that this implementation is legal. This has always been the case
21472 RM References: 9.01 (9.2/2) 9.04 (11.1/2)
21475 @emph{AI-0030 Requeue on synchronized interfaces (2010-07-19)}
21476 @cindex AI-0030 (Ada 2012 feature)
21479 Requeue is permitted to a protected, synchronized or task interface primitive
21480 providing it is known that the overriding operation is an entry. Otherwise
21481 the requeue statement has the same effect as a procedure call. Use of pragma
21482 @code{Implemented} provides a way to impose a static requirement on the
21483 overriding operation by adhering to one of the implementation kinds: entry,
21484 protected procedure or any of the above.
21487 RM References: 9.05 (9) 9.05.04 (2) 9.05.04 (3) 9.05.04 (5)
21488 9.05.04 (6) 9.05.04 (7) 9.05.04 (12)
21492 @emph{AI-0201 Independence of atomic object components (2010-07-22)}
21493 @cindex AI-0201 (Ada 2012 feature)
21496 If an Atomic object has a pragma @code{Pack} or a @code{Component_Size}
21497 attribute, then individual components may not be addressable by independent
21498 tasks. However, if the representation clause has no effect (is confirming),
21499 then independence is not compromised. Furthermore, in GNAT, specification of
21500 other appropriately addressable component sizes (e.g. 16 for 8-bit
21501 characters) also preserves independence. GNAT now gives very clear warnings
21502 both for the declaration of such a type, and for any assignment to its components.
21505 RM References: 9.10 (1/3) C.06 (22/2) C.06 (23/2)
21508 @emph{AI-0009 Pragma Independent[_Components] (2010-07-23)}
21509 @cindex AI-0009 (Ada 2012 feature)
21512 This AI introduces the new pragmas @code{Independent} and
21513 @code{Independent_Components},
21514 which control guaranteeing independence of access to objects and components.
21515 The AI also requires independence not unaffected by confirming rep clauses.
21518 RM References: 9.10 (1) 13.01 (15/1) 13.02 (9) 13.03 (13) C.06 (2)
21519 C.06 (4) C.06 (6) C.06 (9) C.06 (13) C.06 (14)
21523 @emph{AI-0072 Task signalling using 'Terminated (0000-00-00)}
21524 @cindex AI-0072 (Ada 2012 feature)
21527 This AI clarifies that task signalling for reading @code{'Terminated} only
21528 occurs if the result is True. GNAT semantics has always been consistent with
21529 this notion of task signalling.
21532 RM References: 9.10 (6.1/1)
21535 @emph{AI-0108 Limited incomplete view and discriminants (0000-00-00)}
21536 @cindex AI-0108 (Ada 2012 feature)
21539 This AI confirms that an incomplete type from a limited view does not have
21540 discriminants. This has always been the case in GNAT.
21543 RM References: 10.01.01 (12.3/2)
21546 @emph{AI-0129 Limited views and incomplete types (0000-00-00)}
21547 @cindex AI-0129 (Ada 2012 feature)
21550 This AI clarifies the description of limited views: a limited view of a
21551 package includes only one view of a type that has an incomplete declaration
21552 and a full declaration (there is no possible ambiguity in a client package).
21553 This AI also fixes an omission: a nested package in the private part has no
21554 limited view. GNAT always implemented this correctly.
21557 RM References: 10.01.01 (12.2/2) 10.01.01 (12.3/2)
21562 @emph{AI-0077 Limited withs and scope of declarations (0000-00-00)}
21563 @cindex AI-0077 (Ada 2012 feature)
21566 This AI clarifies that a declaration does not include a context clause,
21567 and confirms that it is illegal to have a context in which both a limited
21568 and a nonlimited view of a package are accessible. Such double visibility
21569 was always rejected by GNAT.
21572 RM References: 10.01.02 (12/2) 10.01.02 (21/2) 10.01.02 (22/2)
21575 @emph{AI-0122 Private with and children of generics (0000-00-00)}
21576 @cindex AI-0122 (Ada 2012 feature)
21579 This AI clarifies the visibility of private children of generic units within
21580 instantiations of a parent. GNAT has always handled this correctly.
21583 RM References: 10.01.02 (12/2)
21588 @emph{AI-0040 Limited with clauses on descendant (0000-00-00)}
21589 @cindex AI-0040 (Ada 2012 feature)
21592 This AI confirms that a limited with clause in a child unit cannot name
21593 an ancestor of the unit. This has always been checked in GNAT.
21596 RM References: 10.01.02 (20/2)
21599 @emph{AI-0132 Placement of library unit pragmas (0000-00-00)}
21600 @cindex AI-0132 (Ada 2012 feature)
21603 This AI fills a gap in the description of library unit pragmas. The pragma
21604 clearly must apply to a library unit, even if it does not carry the name
21605 of the enclosing unit. GNAT has always enforced the required check.
21608 RM References: 10.01.05 (7)
21612 @emph{AI-0034 Categorization of limited views (0000-00-00)}
21613 @cindex AI-0034 (Ada 2012 feature)
21616 The RM makes certain limited with clauses illegal because of categorization
21617 considerations, when the corresponding normal with would be legal. This is
21618 not intended, and GNAT has always implemented the recommended behavior.
21621 RM References: 10.02.01 (11/1) 10.02.01 (17/2)
21625 @emph{AI-0035 Inconsistencies with Pure units (0000-00-00)}
21626 @cindex AI-0035 (Ada 2012 feature)
21629 This AI remedies some inconsistencies in the legality rules for Pure units.
21630 Derived access types are legal in a pure unit (on the assumption that the
21631 rule for a zero storage pool size has been enforced on the ancestor type).
21632 The rules are enforced in generic instances and in subunits. GNAT has always
21633 implemented the recommended behavior.
21636 RM References: 10.02.01 (15.1/2) 10.02.01 (15.4/2) 10.02.01 (15.5/2) 10.02.01 (17/2)
21640 @emph{AI-0219 Pure permissions and limited parameters (2010-05-25)}
21641 @cindex AI-0219 (Ada 2012 feature)
21644 This AI refines the rules for the cases with limited parameters which do not
21645 allow the implementations to omit ``redundant''. GNAT now properly conforms
21646 to the requirements of this binding interpretation.
21649 RM References: 10.02.01 (18/2)
21652 @emph{AI-0043 Rules about raising exceptions (0000-00-00)}
21653 @cindex AI-0043 (Ada 2012 feature)
21656 This AI covers various omissions in the RM regarding the raising of
21657 exceptions. GNAT has always implemented the intended semantics.
21660 RM References: 11.04.01 (10.1/2) 11 (2)
21664 @emph{AI-0200 Mismatches in formal package declarations (0000-00-00)}
21665 @cindex AI-0200 (Ada 2012 feature)
21668 This AI plugs a gap in the RM which appeared to allow some obviously intended
21669 illegal instantiations. GNAT has never allowed these instantiations.
21672 RM References: 12.07 (16)
21676 @emph{AI-0112 Detection of duplicate pragmas (2010-07-24)}
21677 @cindex AI-0112 (Ada 2012 feature)
21680 This AI concerns giving names to various representation aspects, but the
21681 practical effect is simply to make the use of duplicate
21682 @code{Atomic}[@code{_Components}],
21683 @code{Volatile}[@code{_Components}] and
21684 @code{Independent}[@code{_Components}] pragmas illegal, and GNAT
21685 now performs this required check.
21688 RM References: 13.01 (8)
21691 @emph{AI-0106 No representation pragmas on generic formals (0000-00-00)}
21692 @cindex AI-0106 (Ada 2012 feature)
21695 The RM appeared to allow representation pragmas on generic formal parameters,
21696 but this was not intended, and GNAT has never permitted this usage.
21699 RM References: 13.01 (9.1/1)
21703 @emph{AI-0012 Pack/Component_Size for aliased/atomic (2010-07-15)}
21704 @cindex AI-0012 (Ada 2012 feature)
21707 It is now illegal to give an inappropriate component size or a pragma
21708 @code{Pack} that attempts to change the component size in the case of atomic
21709 or aliased components. Previously GNAT ignored such an attempt with a
21713 RM References: 13.02 (6.1/2) 13.02 (7) C.06 (10) C.06 (11) C.06 (21)
21717 @emph{AI-0039 Stream attributes cannot be dynamic (0000-00-00)}
21718 @cindex AI-0039 (Ada 2012 feature)
21721 The RM permitted the use of dynamic expressions (such as @code{ptr.@b{all})}
21722 for stream attributes, but these were never useful and are now illegal. GNAT
21723 has always regarded such expressions as illegal.
21726 RM References: 13.03 (4) 13.03 (6) 13.13.02 (38/2)
21730 @emph{AI-0095 Address of intrinsic subprograms (0000-00-00)}
21731 @cindex AI-0095 (Ada 2012 feature)
21734 The prefix of @code{'Address} cannot statically denote a subprogram with
21735 convention @code{Intrinsic}. The use of the @code{Address} attribute raises
21736 @code{Program_Error} if the prefix denotes a subprogram with convention
21740 RM References: 13.03 (11/1)
21744 @emph{AI-0116 Alignment of class-wide objects (0000-00-00)}
21745 @cindex AI-0116 (Ada 2012 feature)
21748 This AI requires that the alignment of a class-wide object be no greater
21749 than the alignment of any type in the class. GNAT has always followed this
21753 RM References: 13.03 (29) 13.11 (16)
21757 @emph{AI-0146 Type invariants (2009-09-21)}
21758 @cindex AI-0146 (Ada 2012 feature)
21761 Type invariants may be specified for private types using the aspect notation.
21762 Aspect @code{Type_Invariant} may be specified for any private type,
21763 @code{Type_Invariant'Class} can
21764 only be specified for tagged types, and is inherited by any descendent of the
21765 tagged types. The invariant is a boolean expression that is tested for being
21766 true in the following situations: conversions to the private type, object
21767 declarations for the private type that are default initialized, and
21769 parameters and returned result on return from any primitive operation for
21770 the type that is visible to a client.
21771 GNAT defines the synonyms @code{Invariant} for @code{Type_Invariant} and
21772 @code{Invariant'Class} for @code{Type_Invariant'Class}.
21775 RM References: 13.03.03 (00)
21778 @emph{AI-0078 Relax Unchecked_Conversion alignment rules (0000-00-00)}
21779 @cindex AI-0078 (Ada 2012 feature)
21782 In Ada 2012, compilers are required to support unchecked conversion where the
21783 target alignment is a multiple of the source alignment. GNAT always supported
21784 this case (and indeed all cases of differing alignments, doing copies where
21785 required if the alignment was reduced).
21788 RM References: 13.09 (7)
21792 @emph{AI-0195 Invalid value handling is implementation defined (2010-07-03)}
21793 @cindex AI-0195 (Ada 2012 feature)
21796 The handling of invalid values is now designated to be implementation
21797 defined. This is a documentation change only, requiring Annex M in the GNAT
21798 Reference Manual to document this handling.
21799 In GNAT, checks for invalid values are made
21800 only when necessary to avoid erroneous behavior. Operations like assignments
21801 which cannot cause erroneous behavior ignore the possibility of invalid
21802 values and do not do a check. The date given above applies only to the
21803 documentation change, this behavior has always been implemented by GNAT.
21806 RM References: 13.09.01 (10)
21809 @emph{AI-0193 Alignment of allocators (2010-09-16)}
21810 @cindex AI-0193 (Ada 2012 feature)
21813 This AI introduces a new attribute @code{Max_Alignment_For_Allocation},
21814 analogous to @code{Max_Size_In_Storage_Elements}, but for alignment instead
21818 RM References: 13.11 (16) 13.11 (21) 13.11.01 (0) 13.11.01 (1)
21819 13.11.01 (2) 13.11.01 (3)
21823 @emph{AI-0177 Parameterized expressions (2010-07-10)}
21824 @cindex AI-0177 (Ada 2012 feature)
21827 The new Ada 2012 notion of parameterized expressions is implemented. The form
21830 @i{function specification} @b{is} (@i{expression})
21834 This is exactly equivalent to the
21835 corresponding function body that returns the expression, but it can appear
21836 in a package spec. Note that the expression must be parenthesized.
21839 RM References: 13.11.01 (3/2)
21842 @emph{AI-0033 Attach/Interrupt_Handler in generic (2010-07-24)}
21843 @cindex AI-0033 (Ada 2012 feature)
21846 Neither of these two pragmas may appear within a generic template, because
21847 the generic might be instantiated at other than the library level.
21850 RM References: 13.11.02 (16) C.03.01 (7/2) C.03.01 (8/2)
21854 @emph{AI-0161 Restriction No_Default_Stream_Attributes (2010-09-11)}
21855 @cindex AI-0161 (Ada 2012 feature)
21858 A new restriction @code{No_Default_Stream_Attributes} prevents the use of any
21859 of the default stream attributes for elementary types. If this restriction is
21860 in force, then it is necessary to provide explicit subprograms for any
21861 stream attributes used.
21864 RM References: 13.12.01 (4/2) 13.13.02 (40/2) 13.13.02 (52/2)
21867 @emph{AI-0194 Value of Stream_Size attribute (0000-00-00)}
21868 @cindex AI-0194 (Ada 2012 feature)
21871 The @code{Stream_Size} attribute returns the default number of bits in the
21872 stream representation of the given type.
21873 This value is not affected by the presence
21874 of stream subprogram attributes for the type. GNAT has always implemented
21875 this interpretation.
21878 RM References: 13.13.02 (1.2/2)
21881 @emph{AI-0109 Redundant check in S'Class'Input (0000-00-00)}
21882 @cindex AI-0109 (Ada 2012 feature)
21885 This AI is an editorial change only. It removes the need for a tag check
21886 that can never fail.
21889 RM References: 13.13.02 (34/2)
21892 @emph{AI-0007 Stream read and private scalar types (0000-00-00)}
21893 @cindex AI-0007 (Ada 2012 feature)
21896 The RM as written appeared to limit the possibilities of declaring read
21897 attribute procedures for private scalar types. This limitation was not
21898 intended, and has never been enforced by GNAT.
21901 RM References: 13.13.02 (50/2) 13.13.02 (51/2)
21905 @emph{AI-0065 Remote access types and external streaming (0000-00-00)}
21906 @cindex AI-0065 (Ada 2012 feature)
21909 This AI clarifies the fact that all remote access types support external
21910 streaming. This fixes an obvious oversight in the definition of the
21911 language, and GNAT always implemented the intended correct rules.
21914 RM References: 13.13.02 (52/2)
21917 @emph{AI-0019 Freezing of primitives for tagged types (0000-00-00)}
21918 @cindex AI-0019 (Ada 2012 feature)
21921 The RM suggests that primitive subprograms of a specific tagged type are
21922 frozen when the tagged type is frozen. This would be an incompatible change
21923 and is not intended. GNAT has never attempted this kind of freezing and its
21924 behavior is consistent with the recommendation of this AI.
21927 RM References: 13.14 (2) 13.14 (3/1) 13.14 (8.1/1) 13.14 (10) 13.14 (14) 13.14 (15.1/2)
21930 @emph{AI-0017 Freezing and incomplete types (0000-00-00)}
21931 @cindex AI-0017 (Ada 2012 feature)
21934 So-called ``Taft-amendment types'' (i.e., types that are completed in package
21935 bodies) are not frozen by the occurrence of bodies in the
21936 enclosing declarative part. GNAT always implemented this properly.
21939 RM References: 13.14 (3/1)
21943 @emph{AI-0060 Extended definition of remote access types (0000-00-00)}
21944 @cindex AI-0060 (Ada 2012 feature)
21947 This AI extends the definition of remote access types to include access
21948 to limited, synchronized, protected or task class-wide interface types.
21949 GNAT already implemented this extension.
21952 RM References: A (4) E.02.02 (9/1) E.02.02 (9.2/1) E.02.02 (14/2) E.02.02 (18)
21955 @emph{AI-0114 Classification of letters (0000-00-00)}
21956 @cindex AI-0114 (Ada 2012 feature)
21959 The code points 170 (@code{FEMININE ORDINAL INDICATOR}),
21960 181 (@code{MICRO SIGN}), and
21961 186 (@code{MASCULINE ORDINAL INDICATOR}) are technically considered
21962 lower case letters by Unicode.
21963 However, they are not allowed in identifiers, and they
21964 return @code{False} to @code{Ada.Characters.Handling.Is_Letter/Is_Lower}.
21965 This behavior is consistent with that defined in Ada 95.
21968 RM References: A.03.02 (59) A.04.06 (7)
21972 @emph{AI-0185 Ada.Wide_[Wide_]Characters.Handling (2010-07-06)}
21973 @cindex AI-0185 (Ada 2012 feature)
21976 Two new packages @code{Ada.Wide_[Wide_]Characters.Handling} provide
21977 classification functions for @code{Wide_Character} and
21978 @code{Wide_Wide_Character}, as well as providing
21979 case folding routines for @code{Wide_[Wide_]Character} and
21980 @code{Wide_[Wide_]String}.
21983 RM References: A.03.05 (0) A.03.06 (0)
21987 @emph{AI-0031 Add From parameter to Find_Token (2010-07-25)}
21988 @cindex AI-0031 (Ada 2012 feature)
21991 A new version of @code{Find_Token} is added to all relevant string packages,
21992 with an extra parameter @code{From}. Instead of starting at the first
21993 character of the string, the search for a matching Token starts at the
21994 character indexed by the value of @code{From}.
21995 These procedures are available in all versions of Ada
21996 but if used in versions earlier than Ada 2012 they will generate a warning
21997 that an Ada 2012 subprogram is being used.
22000 RM References: A.04.03 (16) A.04.03 (67) A.04.03 (68/1) A.04.04 (51)
22005 @emph{AI-0056 Index on null string returns zero (0000-00-00)}
22006 @cindex AI-0056 (Ada 2012 feature)
22009 The wording in the Ada 2005 RM implied an incompatible handling of the
22010 @code{Index} functions, resulting in raising an exception instead of
22011 returning zero in some situations.
22012 This was not intended and has been corrected.
22013 GNAT always returned zero, and is thus consistent with this AI.
22016 RM References: A.04.03 (56.2/2) A.04.03 (58.5/2)
22020 @emph{AI-0137 String encoding package (2010-03-25)}
22021 @cindex AI-0137 (Ada 2012 feature)
22024 The packages @code{Ada.Strings.UTF_Encoding}, together with its child
22025 packages, @code{Conversions}, @code{Strings}, @code{Wide_Strings},
22026 and @code{Wide_Wide_Strings} have been
22027 implemented. These packages (whose documentation can be found in the spec
22028 files @file{a-stuten.ads}, @file{a-suenco.ads}, @file{a-suenst.ads},
22029 @file{a-suewst.ads}, @file{a-suezst.ads}) allow encoding and decoding of
22030 @code{String}, @code{Wide_String}, and @code{Wide_Wide_String}
22031 values using UTF coding schemes (including UTF-8, UTF-16LE, UTF-16BE, and
22032 UTF-16), as well as conversions between the different UTF encodings. With
22033 the exception of @code{Wide_Wide_Strings}, these packages are available in
22034 Ada 95 and Ada 2005 mode as well as Ada 2012 mode.
22035 The @code{Wide_Wide_Strings package}
22036 is available in Ada 2005 mode as well as Ada 2012 mode (but not in Ada 95
22037 mode since it uses @code{Wide_Wide_Character}).
22040 RM References: A.04.11
22043 @emph{AI-0038 Minor errors in Text_IO (0000-00-00)}
22044 @cindex AI-0038 (Ada 2012 feature)
22047 These are minor errors in the description on three points. The intent on
22048 all these points has always been clear, and GNAT has always implemented the
22049 correct intended semantics.
22052 RM References: A.10.05 (37) A.10.07 (8/1) A.10.07 (10) A.10.07 (12) A.10.08 (10) A.10.08 (24)
22055 @emph{AI-0044 Restrictions on container instantiations (0000-00-00)}
22056 @cindex AI-0044 (Ada 2012 feature)
22059 This AI places restrictions on allowed instantiations of generic containers.
22060 These restrictions are not checked by the compiler, so there is nothing to
22061 change in the implementation. This affects only the RM documentation.
22064 RM References: A.18 (4/2) A.18.02 (231/2) A.18.03 (145/2) A.18.06 (56/2) A.18.08 (66/2) A.18.09 (79/2) A.18.26 (5/2) A.18.26 (9/2)
22067 @emph{AI-0127 Adding Locale Capabilities (2010-09-29)}
22068 @cindex AI-0127 (Ada 2012 feature)
22071 This package provides an interface for identifying the current locale.
22074 RM References: A.19 A.19.01 A.19.02 A.19.03 A.19.05 A.19.06
22075 A.19.07 A.19.08 A.19.09 A.19.10 A.19.11 A.19.12 A.19.13
22080 @emph{AI-0002 Export C with unconstrained arrays (0000-00-00)}
22081 @cindex AI-0002 (Ada 2012 feature)
22084 The compiler is not required to support exporting an Ada subprogram with
22085 convention C if there are parameters or a return type of an unconstrained
22086 array type (such as @code{String}). GNAT allows such declarations but
22087 generates warnings. It is possible, but complicated, to write the
22088 corresponding C code and certainly such code would be specific to GNAT and
22092 RM References: B.01 (17) B.03 (62) B.03 (71.1/2)
22096 @emph{AI-0216 No_Task_Hierarchy forbids local tasks (0000-00-00)}
22097 @cindex AI05-0216 (Ada 2012 feature)
22100 It is clearly the intention that @code{No_Task_Hierarchy} is intended to
22101 forbid tasks declared locally within subprograms, or functions returning task
22102 objects, and that is the implementation that GNAT has always provided.
22103 However the language in the RM was not sufficiently clear on this point.
22104 Thus this is a documentation change in the RM only.
22107 RM References: D.07 (3/3)
22110 @emph{AI-0211 No_Relative_Delays forbids Set_Handler use (2010-07-09)}
22111 @cindex AI-0211 (Ada 2012 feature)
22114 The restriction @code{No_Relative_Delays} forbids any calls to the subprogram
22115 @code{Ada.Real_Time.Timing_Events.Set_Handler}.
22118 RM References: D.07 (5) D.07 (10/2) D.07 (10.4/2) D.07 (10.7/2)
22121 @emph{AI-0190 pragma Default_Storage_Pool (2010-09-15)}
22122 @cindex AI-0190 (Ada 2012 feature)
22125 This AI introduces a new pragma @code{Default_Storage_Pool}, which can be
22126 used to control storage pools globally.
22127 In particular, you can force every access
22128 type that is used for allocation (@b{new}) to have an explicit storage pool,
22129 or you can declare a pool globally to be used for all access types that lack
22133 RM References: D.07 (8)
22136 @emph{AI-0189 No_Allocators_After_Elaboration (2010-01-23)}
22137 @cindex AI-0189 (Ada 2012 feature)
22140 This AI introduces a new restriction @code{No_Allocators_After_Elaboration},
22141 which says that no dynamic allocation will occur once elaboration is
22143 In general this requires a run-time check, which is not required, and which
22144 GNAT does not attempt. But the static cases of allocators in a task body or
22145 in the body of the main program are detected and flagged at compile or bind
22149 RM References: D.07 (19.1/2) H.04 (23.3/2)
22152 @emph{AI-0171 Pragma CPU and Ravenscar Profile (2010-09-24)}
22153 @cindex AI-0171 (Ada 2012 feature)
22156 A new package @code{System.Multiprocessors} is added, together with the
22157 definition of pragma @code{CPU} for controlling task affinity. A new no
22158 dependence restriction, on @code{System.Multiprocessors.Dispatching_Domains},
22159 is added to the Ravenscar profile.
22162 RM References: D.13.01 (4/2) D.16
22166 @emph{AI-0210 Correct Timing_Events metric (0000-00-00)}
22167 @cindex AI-0210 (Ada 2012 feature)
22170 This is a documentation only issue regarding wording of metric requirements,
22171 that does not affect the implementation of the compiler.
22174 RM References: D.15 (24/2)
22178 @emph{AI-0206 Remote types packages and preelaborate (2010-07-24)}
22179 @cindex AI-0206 (Ada 2012 feature)
22182 Remote types packages are now allowed to depend on preelaborated packages.
22183 This was formerly considered illegal.
22186 RM References: E.02.02 (6)
22191 @emph{AI-0152 Restriction No_Anonymous_Allocators (2010-09-08)}
22192 @cindex AI-0152 (Ada 2012 feature)
22195 Restriction @code{No_Anonymous_Allocators} prevents the use of allocators
22196 where the type of the returned value is an anonymous access type.
22199 RM References: H.04 (8/1)
22203 @node Obsolescent Features
22204 @chapter Obsolescent Features
22207 This chapter describes features that are provided by GNAT, but are
22208 considered obsolescent since there are preferred ways of achieving
22209 the same effect. These features are provided solely for historical
22210 compatibility purposes.
22213 * pragma No_Run_Time::
22214 * pragma Ravenscar::
22215 * pragma Restricted_Run_Time::
22218 @node pragma No_Run_Time
22219 @section pragma No_Run_Time
22221 The pragma @code{No_Run_Time} is used to achieve an affect similar
22222 to the use of the "Zero Foot Print" configurable run time, but without
22223 requiring a specially configured run time. The result of using this
22224 pragma, which must be used for all units in a partition, is to restrict
22225 the use of any language features requiring run-time support code. The
22226 preferred usage is to use an appropriately configured run-time that
22227 includes just those features that are to be made accessible.
22229 @node pragma Ravenscar
22230 @section pragma Ravenscar
22232 The pragma @code{Ravenscar} has exactly the same effect as pragma
22233 @code{Profile (Ravenscar)}. The latter usage is preferred since it
22234 is part of the new Ada 2005 standard.
22236 @node pragma Restricted_Run_Time
22237 @section pragma Restricted_Run_Time
22239 The pragma @code{Restricted_Run_Time} has exactly the same effect as
22240 pragma @code{Profile (Restricted)}. The latter usage is
22241 preferred since the Ada 2005 pragma @code{Profile} is intended for
22242 this kind of implementation dependent addition.
22245 @c GNU Free Documentation License
22247 @node Index,,GNU Free Documentation License, Top
22255 tablishes the following set of restrictions: