xref: /netbsd-src/external/gpl3/gcc.old/dist/gcc/doc/generic.texi (revision b7b7574d3bf8eeb51a1fa3977b59142ec6434a55)
1@c Copyright (c) 2004, 2005, 2007, 2008, 2010 Free Software Foundation, Inc.
2@c Free Software Foundation, Inc.
3@c This is part of the GCC manual.
4@c For copying conditions, see the file gcc.texi.
5
6@c ---------------------------------------------------------------------
7@c GENERIC
8@c ---------------------------------------------------------------------
9
10@node GENERIC
11@chapter GENERIC
12@cindex GENERIC
13
14The purpose of GENERIC is simply to provide a
15language-independent way of representing an entire function in
16trees.  To this end, it was necessary to add a few new tree codes
17to the back end, but most everything was already there.  If you
18can express it with the codes in @code{gcc/tree.def}, it's
19GENERIC@.
20
21Early on, there was a great deal of debate about how to think
22about statements in a tree IL@.  In GENERIC, a statement is
23defined as any expression whose value, if any, is ignored.  A
24statement will always have @code{TREE_SIDE_EFFECTS} set (or it
25will be discarded), but a non-statement expression may also have
26side effects.  A @code{CALL_EXPR}, for instance.
27
28It would be possible for some local optimizations to work on the
29GENERIC form of a function; indeed, the adapted tree inliner
30works fine on GENERIC, but the current compiler performs inlining
31after lowering to GIMPLE (a restricted form described in the next
32section). Indeed, currently the frontends perform this lowering
33before handing off to @code{tree_rest_of_compilation}, but this
34seems inelegant.
35
36@menu
37* Deficiencies::                Topics net yet covered in this document.
38* Tree overview::               All about @code{tree}s.
39* Types::                       Fundamental and aggregate types.
40* Declarations::                Type declarations and variables.
41* Attributes::                  Declaration and type attributes.
42* Expressions: Expression trees.            Operating on data.
43* Statements::                  Control flow and related trees.
44* Functions::           	Function bodies, linkage, and other aspects.
45* Language-dependent trees::    Topics and trees specific to language front ends.
46* C and C++ Trees::     	Trees specific to C and C++.
47* Java Trees:: 	                Trees specific to Java.
48@end menu
49
50@c ---------------------------------------------------------------------
51@c Deficiencies
52@c ---------------------------------------------------------------------
53
54@node Deficiencies
55@section Deficiencies
56
57There are many places in which this document is incomplet and incorrekt.
58It is, as of yet, only @emph{preliminary} documentation.
59
60@c ---------------------------------------------------------------------
61@c Overview
62@c ---------------------------------------------------------------------
63
64@node Tree overview
65@section Overview
66@cindex tree
67@findex TREE_CODE
68
69The central data structure used by the internal representation is the
70@code{tree}.  These nodes, while all of the C type @code{tree}, are of
71many varieties.  A @code{tree} is a pointer type, but the object to
72which it points may be of a variety of types.  From this point forward,
73we will refer to trees in ordinary type, rather than in @code{this
74font}, except when talking about the actual C type @code{tree}.
75
76You can tell what kind of node a particular tree is by using the
77@code{TREE_CODE} macro.  Many, many macros take trees as input and
78return trees as output.  However, most macros require a certain kind of
79tree node as input.  In other words, there is a type-system for trees,
80but it is not reflected in the C type-system.
81
82For safety, it is useful to configure GCC with @option{--enable-checking}.
83Although this results in a significant performance penalty (since all
84tree types are checked at run-time), and is therefore inappropriate in a
85release version, it is extremely helpful during the development process.
86
87Many macros behave as predicates.  Many, although not all, of these
88predicates end in @samp{_P}.  Do not rely on the result type of these
89macros being of any particular type.  You may, however, rely on the fact
90that the type can be compared to @code{0}, so that statements like
91@smallexample
92if (TEST_P (t) && !TEST_P (y))
93  x = 1;
94@end smallexample
95@noindent
96and
97@smallexample
98int i = (TEST_P (t) != 0);
99@end smallexample
100@noindent
101are legal.  Macros that return @code{int} values now may be changed to
102return @code{tree} values, or other pointers in the future.  Even those
103that continue to return @code{int} may return multiple nonzero codes
104where previously they returned only zero and one.  Therefore, you should
105not write code like
106@smallexample
107if (TEST_P (t) == 1)
108@end smallexample
109@noindent
110as this code is not guaranteed to work correctly in the future.
111
112You should not take the address of values returned by the macros or
113functions described here.  In particular, no guarantee is given that the
114values are lvalues.
115
116In general, the names of macros are all in uppercase, while the names of
117functions are entirely in lowercase.  There are rare exceptions to this
118rule.  You should assume that any macro or function whose name is made
119up entirely of uppercase letters may evaluate its arguments more than
120once.  You may assume that a macro or function whose name is made up
121entirely of lowercase letters will evaluate its arguments only once.
122
123The @code{error_mark_node} is a special tree.  Its tree code is
124@code{ERROR_MARK}, but since there is only ever one node with that code,
125the usual practice is to compare the tree against
126@code{error_mark_node}.  (This test is just a test for pointer
127equality.)  If an error has occurred during front-end processing the
128flag @code{errorcount} will be set.  If the front end has encountered
129code it cannot handle, it will issue a message to the user and set
130@code{sorrycount}.  When these flags are set, any macro or function
131which normally returns a tree of a particular kind may instead return
132the @code{error_mark_node}.  Thus, if you intend to do any processing of
133erroneous code, you must be prepared to deal with the
134@code{error_mark_node}.
135
136Occasionally, a particular tree slot (like an operand to an expression,
137or a particular field in a declaration) will be referred to as
138``reserved for the back end''.  These slots are used to store RTL when
139the tree is converted to RTL for use by the GCC back end.  However, if
140that process is not taking place (e.g., if the front end is being hooked
141up to an intelligent editor), then those slots may be used by the
142back end presently in use.
143
144If you encounter situations that do not match this documentation, such
145as tree nodes of types not mentioned here, or macros documented to
146return entities of a particular kind that instead return entities of
147some different kind, you have found a bug, either in the front end or in
148the documentation.  Please report these bugs as you would any other
149bug.
150
151@menu
152* Macros and Functions::Macros and functions that can be used with all trees.
153* Identifiers::         The names of things.
154* Containers::          Lists and vectors.
155@end menu
156
157@c ---------------------------------------------------------------------
158@c Trees
159@c ---------------------------------------------------------------------
160
161@node Macros and Functions
162@subsection Trees
163@cindex tree
164@findex TREE_CHAIN
165@findex TREE_TYPE
166
167All GENERIC trees have two fields in common.  First, @code{TREE_CHAIN}
168is a pointer that can be used as a singly-linked list to other trees.
169The other is @code{TREE_TYPE}.  Many trees store the type of an
170expression or declaration in this field.
171
172These are some other functions for handling trees:
173
174@ftable @code
175
176@item tree_size
177Return the number of bytes a tree takes.
178
179@item build0
180@itemx build1
181@itemx build2
182@itemx build3
183@itemx build4
184@itemx build5
185@itemx build6
186
187These functions build a tree and supply values to put in each
188parameter.  The basic signature is @samp{@w{code, type, [operands]}}.
189@code{code} is the @code{TREE_CODE}, and @code{type} is a tree
190representing the @code{TREE_TYPE}.  These are followed by the
191operands, each of which is also a tree.
192
193@end ftable
194
195
196@c ---------------------------------------------------------------------
197@c Identifiers
198@c ---------------------------------------------------------------------
199
200@node Identifiers
201@subsection Identifiers
202@cindex identifier
203@cindex name
204@tindex IDENTIFIER_NODE
205
206An @code{IDENTIFIER_NODE} represents a slightly more general concept
207that the standard C or C++ concept of identifier.  In particular, an
208@code{IDENTIFIER_NODE} may contain a @samp{$}, or other extraordinary
209characters.
210
211There are never two distinct @code{IDENTIFIER_NODE}s representing the
212same identifier.  Therefore, you may use pointer equality to compare
213@code{IDENTIFIER_NODE}s, rather than using a routine like
214@code{strcmp}.  Use @code{get_identifier} to obtain the unique
215@code{IDENTIFIER_NODE} for a supplied string.
216
217You can use the following macros to access identifiers:
218@ftable @code
219@item IDENTIFIER_POINTER
220The string represented by the identifier, represented as a
221@code{char*}.  This string is always @code{NUL}-terminated, and contains
222no embedded @code{NUL} characters.
223
224@item IDENTIFIER_LENGTH
225The length of the string returned by @code{IDENTIFIER_POINTER}, not
226including the trailing @code{NUL}.  This value of
227@code{IDENTIFIER_LENGTH (x)} is always the same as @code{strlen
228(IDENTIFIER_POINTER (x))}.
229
230@item IDENTIFIER_OPNAME_P
231This predicate holds if the identifier represents the name of an
232overloaded operator.  In this case, you should not depend on the
233contents of either the @code{IDENTIFIER_POINTER} or the
234@code{IDENTIFIER_LENGTH}.
235
236@item IDENTIFIER_TYPENAME_P
237This predicate holds if the identifier represents the name of a
238user-defined conversion operator.  In this case, the @code{TREE_TYPE} of
239the @code{IDENTIFIER_NODE} holds the type to which the conversion
240operator converts.
241
242@end ftable
243
244@c ---------------------------------------------------------------------
245@c Containers
246@c ---------------------------------------------------------------------
247
248@node Containers
249@subsection Containers
250@cindex container
251@cindex list
252@cindex vector
253@tindex TREE_LIST
254@tindex TREE_VEC
255@findex TREE_PURPOSE
256@findex TREE_VALUE
257@findex TREE_VEC_LENGTH
258@findex TREE_VEC_ELT
259
260Two common container data structures can be represented directly with
261tree nodes.  A @code{TREE_LIST} is a singly linked list containing two
262trees per node.  These are the @code{TREE_PURPOSE} and @code{TREE_VALUE}
263of each node.  (Often, the @code{TREE_PURPOSE} contains some kind of
264tag, or additional information, while the @code{TREE_VALUE} contains the
265majority of the payload.  In other cases, the @code{TREE_PURPOSE} is
266simply @code{NULL_TREE}, while in still others both the
267@code{TREE_PURPOSE} and @code{TREE_VALUE} are of equal stature.)  Given
268one @code{TREE_LIST} node, the next node is found by following the
269@code{TREE_CHAIN}.  If the @code{TREE_CHAIN} is @code{NULL_TREE}, then
270you have reached the end of the list.
271
272A @code{TREE_VEC} is a simple vector.  The @code{TREE_VEC_LENGTH} is an
273integer (not a tree) giving the number of nodes in the vector.  The
274nodes themselves are accessed using the @code{TREE_VEC_ELT} macro, which
275takes two arguments.  The first is the @code{TREE_VEC} in question; the
276second is an integer indicating which element in the vector is desired.
277The elements are indexed from zero.
278
279@c ---------------------------------------------------------------------
280@c Types
281@c ---------------------------------------------------------------------
282
283@node Types
284@section Types
285@cindex type
286@cindex pointer
287@cindex reference
288@cindex fundamental type
289@cindex array
290@tindex VOID_TYPE
291@tindex INTEGER_TYPE
292@tindex TYPE_MIN_VALUE
293@tindex TYPE_MAX_VALUE
294@tindex REAL_TYPE
295@tindex FIXED_POINT_TYPE
296@tindex COMPLEX_TYPE
297@tindex ENUMERAL_TYPE
298@tindex BOOLEAN_TYPE
299@tindex POINTER_TYPE
300@tindex REFERENCE_TYPE
301@tindex FUNCTION_TYPE
302@tindex METHOD_TYPE
303@tindex ARRAY_TYPE
304@tindex RECORD_TYPE
305@tindex UNION_TYPE
306@tindex UNKNOWN_TYPE
307@tindex OFFSET_TYPE
308@findex TYPE_UNQUALIFIED
309@findex TYPE_QUAL_CONST
310@findex TYPE_QUAL_VOLATILE
311@findex TYPE_QUAL_RESTRICT
312@findex TYPE_MAIN_VARIANT
313@cindex qualified type
314@findex TYPE_SIZE
315@findex TYPE_ALIGN
316@findex TYPE_PRECISION
317@findex TYPE_ARG_TYPES
318@findex TYPE_METHOD_BASETYPE
319@findex TYPE_OFFSET_BASETYPE
320@findex TREE_TYPE
321@findex TYPE_CONTEXT
322@findex TYPE_NAME
323@findex TYPENAME_TYPE_FULLNAME
324@findex TYPE_FIELDS
325@findex TYPE_CANONICAL
326@findex TYPE_STRUCTURAL_EQUALITY_P
327@findex SET_TYPE_STRUCTURAL_EQUALITY
328
329All types have corresponding tree nodes.  However, you should not assume
330that there is exactly one tree node corresponding to each type.  There
331are often multiple nodes corresponding to the same type.
332
333For the most part, different kinds of types have different tree codes.
334(For example, pointer types use a @code{POINTER_TYPE} code while arrays
335use an @code{ARRAY_TYPE} code.)  However, pointers to member functions
336use the @code{RECORD_TYPE} code.  Therefore, when writing a
337@code{switch} statement that depends on the code associated with a
338particular type, you should take care to handle pointers to member
339functions under the @code{RECORD_TYPE} case label.
340
341The following functions and macros deal with cv-qualification of types:
342@ftable @code
343@item TYPE_MAIN_VARIANT
344This macro returns the unqualified version of a type.  It may be applied
345to an unqualified type, but it is not always the identity function in
346that case.
347@end ftable
348
349A few other macros and functions are usable with all types:
350@ftable @code
351@item TYPE_SIZE
352The number of bits required to represent the type, represented as an
353@code{INTEGER_CST}.  For an incomplete type, @code{TYPE_SIZE} will be
354@code{NULL_TREE}.
355
356@item TYPE_ALIGN
357The alignment of the type, in bits, represented as an @code{int}.
358
359@item TYPE_NAME
360This macro returns a declaration (in the form of a @code{TYPE_DECL}) for
361the type.  (Note this macro does @emph{not} return an
362@code{IDENTIFIER_NODE}, as you might expect, given its name!)  You can
363look at the @code{DECL_NAME} of the @code{TYPE_DECL} to obtain the
364actual name of the type.  The @code{TYPE_NAME} will be @code{NULL_TREE}
365for a type that is not a built-in type, the result of a typedef, or a
366named class type.
367
368@item TYPE_CANONICAL
369This macro returns the ``canonical'' type for the given type
370node. Canonical types are used to improve performance in the C++ and
371Objective-C++ front ends by allowing efficient comparison between two
372type nodes in @code{same_type_p}: if the @code{TYPE_CANONICAL} values
373of the types are equal, the types are equivalent; otherwise, the types
374are not equivalent. The notion of equivalence for canonical types is
375the same as the notion of type equivalence in the language itself. For
376instance,
377
378When @code{TYPE_CANONICAL} is @code{NULL_TREE}, there is no canonical
379type for the given type node. In this case, comparison between this
380type and any other type requires the compiler to perform a deep,
381``structural'' comparison to see if the two type nodes have the same
382form and properties.
383
384The canonical type for a node is always the most fundamental type in
385the equivalence class of types. For instance, @code{int} is its own
386canonical type. A typedef @code{I} of @code{int} will have @code{int}
387as its canonical type. Similarly, @code{I*}@ and a typedef @code{IP}@
388(defined to @code{I*}) will has @code{int*} as their canonical
389type. When building a new type node, be sure to set
390@code{TYPE_CANONICAL} to the appropriate canonical type. If the new
391type is a compound type (built from other types), and any of those
392other types require structural equality, use
393@code{SET_TYPE_STRUCTURAL_EQUALITY} to ensure that the new type also
394requires structural equality. Finally, if for some reason you cannot
395guarantee that @code{TYPE_CANONICAL} will point to the canonical type,
396use @code{SET_TYPE_STRUCTURAL_EQUALITY} to make sure that the new
397type--and any type constructed based on it--requires structural
398equality. If you suspect that the canonical type system is
399miscomparing types, pass @code{--param verify-canonical-types=1} to
400the compiler or configure with @code{--enable-checking} to force the
401compiler to verify its canonical-type comparisons against the
402structural comparisons; the compiler will then print any warnings if
403the canonical types miscompare.
404
405@item TYPE_STRUCTURAL_EQUALITY_P
406This predicate holds when the node requires structural equality
407checks, e.g., when @code{TYPE_CANONICAL} is @code{NULL_TREE}.
408
409@item SET_TYPE_STRUCTURAL_EQUALITY
410This macro states that the type node it is given requires structural
411equality checks, e.g., it sets @code{TYPE_CANONICAL} to
412@code{NULL_TREE}.
413
414@item same_type_p
415This predicate takes two types as input, and holds if they are the same
416type.  For example, if one type is a @code{typedef} for the other, or
417both are @code{typedef}s for the same type.  This predicate also holds if
418the two trees given as input are simply copies of one another; i.e.,
419there is no difference between them at the source level, but, for
420whatever reason, a duplicate has been made in the representation.  You
421should never use @code{==} (pointer equality) to compare types; always
422use @code{same_type_p} instead.
423@end ftable
424
425Detailed below are the various kinds of types, and the macros that can
426be used to access them.  Although other kinds of types are used
427elsewhere in G++, the types described here are the only ones that you
428will encounter while examining the intermediate representation.
429
430@table @code
431@item VOID_TYPE
432Used to represent the @code{void} type.
433
434@item INTEGER_TYPE
435Used to represent the various integral types, including @code{char},
436@code{short}, @code{int}, @code{long}, and @code{long long}.  This code
437is not used for enumeration types, nor for the @code{bool} type.
438The @code{TYPE_PRECISION} is the number of bits used in
439the representation, represented as an @code{unsigned int}.  (Note that
440in the general case this is not the same value as @code{TYPE_SIZE};
441suppose that there were a 24-bit integer type, but that alignment
442requirements for the ABI required 32-bit alignment.  Then,
443@code{TYPE_SIZE} would be an @code{INTEGER_CST} for 32, while
444@code{TYPE_PRECISION} would be 24.)  The integer type is unsigned if
445@code{TYPE_UNSIGNED} holds; otherwise, it is signed.
446
447The @code{TYPE_MIN_VALUE} is an @code{INTEGER_CST} for the smallest
448integer that may be represented by this type.  Similarly, the
449@code{TYPE_MAX_VALUE} is an @code{INTEGER_CST} for the largest integer
450that may be represented by this type.
451
452@item REAL_TYPE
453Used to represent the @code{float}, @code{double}, and @code{long
454double} types.  The number of bits in the floating-point representation
455is given by @code{TYPE_PRECISION}, as in the @code{INTEGER_TYPE} case.
456
457@item FIXED_POINT_TYPE
458Used to represent the @code{short _Fract}, @code{_Fract}, @code{long
459_Fract}, @code{long long _Fract}, @code{short _Accum}, @code{_Accum},
460@code{long _Accum}, and @code{long long _Accum} types.  The number of bits
461in the fixed-point representation is given by @code{TYPE_PRECISION},
462as in the @code{INTEGER_TYPE} case.  There may be padding bits, fractional
463bits and integral bits.  The number of fractional bits is given by
464@code{TYPE_FBIT}, and the number of integral bits is given by @code{TYPE_IBIT}.
465The fixed-point type is unsigned if @code{TYPE_UNSIGNED} holds; otherwise,
466it is signed.
467The fixed-point type is saturating if @code{TYPE_SATURATING} holds; otherwise,
468it is not saturating.
469
470@item COMPLEX_TYPE
471Used to represent GCC built-in @code{__complex__} data types.  The
472@code{TREE_TYPE} is the type of the real and imaginary parts.
473
474@item ENUMERAL_TYPE
475Used to represent an enumeration type.  The @code{TYPE_PRECISION} gives
476(as an @code{int}), the number of bits used to represent the type.  If
477there are no negative enumeration constants, @code{TYPE_UNSIGNED} will
478hold.  The minimum and maximum enumeration constants may be obtained
479with @code{TYPE_MIN_VALUE} and @code{TYPE_MAX_VALUE}, respectively; each
480of these macros returns an @code{INTEGER_CST}.
481
482The actual enumeration constants themselves may be obtained by looking
483at the @code{TYPE_VALUES}.  This macro will return a @code{TREE_LIST},
484containing the constants.  The @code{TREE_PURPOSE} of each node will be
485an @code{IDENTIFIER_NODE} giving the name of the constant; the
486@code{TREE_VALUE} will be an @code{INTEGER_CST} giving the value
487assigned to that constant.  These constants will appear in the order in
488which they were declared.  The @code{TREE_TYPE} of each of these
489constants will be the type of enumeration type itself.
490
491@item BOOLEAN_TYPE
492Used to represent the @code{bool} type.
493
494@item POINTER_TYPE
495Used to represent pointer types, and pointer to data member types.  The
496@code{TREE_TYPE} gives the type to which this type points.
497
498@item REFERENCE_TYPE
499Used to represent reference types.  The @code{TREE_TYPE} gives the type
500to which this type refers.
501
502@item FUNCTION_TYPE
503Used to represent the type of non-member functions and of static member
504functions.  The @code{TREE_TYPE} gives the return type of the function.
505The @code{TYPE_ARG_TYPES} are a @code{TREE_LIST} of the argument types.
506The @code{TREE_VALUE} of each node in this list is the type of the
507corresponding argument; the @code{TREE_PURPOSE} is an expression for the
508default argument value, if any.  If the last node in the list is
509@code{void_list_node} (a @code{TREE_LIST} node whose @code{TREE_VALUE}
510is the @code{void_type_node}), then functions of this type do not take
511variable arguments.  Otherwise, they do take a variable number of
512arguments.
513
514Note that in C (but not in C++) a function declared like @code{void f()}
515is an unprototyped function taking a variable number of arguments; the
516@code{TYPE_ARG_TYPES} of such a function will be @code{NULL}.
517
518@item METHOD_TYPE
519Used to represent the type of a non-static member function.  Like a
520@code{FUNCTION_TYPE}, the return type is given by the @code{TREE_TYPE}.
521The type of @code{*this}, i.e., the class of which functions of this
522type are a member, is given by the @code{TYPE_METHOD_BASETYPE}.  The
523@code{TYPE_ARG_TYPES} is the parameter list, as for a
524@code{FUNCTION_TYPE}, and includes the @code{this} argument.
525
526@item ARRAY_TYPE
527Used to represent array types.  The @code{TREE_TYPE} gives the type of
528the elements in the array.  If the array-bound is present in the type,
529the @code{TYPE_DOMAIN} is an @code{INTEGER_TYPE} whose
530@code{TYPE_MIN_VALUE} and @code{TYPE_MAX_VALUE} will be the lower and
531upper bounds of the array, respectively.  The @code{TYPE_MIN_VALUE} will
532always be an @code{INTEGER_CST} for zero, while the
533@code{TYPE_MAX_VALUE} will be one less than the number of elements in
534the array, i.e., the highest value which may be used to index an element
535in the array.
536
537@item RECORD_TYPE
538Used to represent @code{struct} and @code{class} types, as well as
539pointers to member functions and similar constructs in other languages.
540@code{TYPE_FIELDS} contains the items contained in this type, each of
541which can be a @code{FIELD_DECL}, @code{VAR_DECL}, @code{CONST_DECL}, or
542@code{TYPE_DECL}.  You may not make any assumptions about the ordering
543of the fields in the type or whether one or more of them overlap.
544
545@item UNION_TYPE
546Used to represent @code{union} types.  Similar to @code{RECORD_TYPE}
547except that all @code{FIELD_DECL} nodes in @code{TYPE_FIELD} start at
548bit position zero.
549
550@item QUAL_UNION_TYPE
551Used to represent part of a variant record in Ada.  Similar to
552@code{UNION_TYPE} except that each @code{FIELD_DECL} has a
553@code{DECL_QUALIFIER} field, which contains a boolean expression that
554indicates whether the field is present in the object.  The type will only
555have one field, so each field's @code{DECL_QUALIFIER} is only evaluated
556if none of the expressions in the previous fields in @code{TYPE_FIELDS}
557are nonzero.  Normally these expressions will reference a field in the
558outer object using a @code{PLACEHOLDER_EXPR}.
559
560@item LANG_TYPE
561This node is used to represent a language-specific type.  The front
562end must handle it.
563
564@item OFFSET_TYPE
565This node is used to represent a pointer-to-data member.  For a data
566member @code{X::m} the @code{TYPE_OFFSET_BASETYPE} is @code{X} and the
567@code{TREE_TYPE} is the type of @code{m}.
568
569@end table
570
571There are variables whose values represent some of the basic types.
572These include:
573@table @code
574@item void_type_node
575A node for @code{void}.
576
577@item integer_type_node
578A node for @code{int}.
579
580@item unsigned_type_node.
581A node for @code{unsigned int}.
582
583@item char_type_node.
584A node for @code{char}.
585@end table
586@noindent
587It may sometimes be useful to compare one of these variables with a type
588in hand, using @code{same_type_p}.
589
590@c ---------------------------------------------------------------------
591@c Declarations
592@c ---------------------------------------------------------------------
593
594@node Declarations
595@section Declarations
596@cindex declaration
597@cindex variable
598@cindex type declaration
599@tindex LABEL_DECL
600@tindex CONST_DECL
601@tindex TYPE_DECL
602@tindex VAR_DECL
603@tindex PARM_DECL
604@tindex DEBUG_EXPR_DECL
605@tindex FIELD_DECL
606@tindex NAMESPACE_DECL
607@tindex RESULT_DECL
608@tindex TEMPLATE_DECL
609@tindex THUNK_DECL
610@findex THUNK_DELTA
611@findex DECL_INITIAL
612@findex DECL_SIZE
613@findex DECL_ALIGN
614@findex DECL_EXTERNAL
615
616This section covers the various kinds of declarations that appear in the
617internal representation, except for declarations of functions
618(represented by @code{FUNCTION_DECL} nodes), which are described in
619@ref{Functions}.
620
621@menu
622* Working with declarations::  Macros and functions that work on
623declarations.
624* Internal structure:: How declaration nodes are represented.
625@end menu
626
627@node Working with declarations
628@subsection Working with declarations
629
630Some macros can be used with any kind of declaration.  These include:
631@ftable @code
632@item DECL_NAME
633This macro returns an @code{IDENTIFIER_NODE} giving the name of the
634entity.
635
636@item TREE_TYPE
637This macro returns the type of the entity declared.
638
639@item EXPR_FILENAME
640This macro returns the name of the file in which the entity was
641declared, as a @code{char*}.  For an entity declared implicitly by the
642compiler (like @code{__builtin_memcpy}), this will be the string
643@code{"<internal>"}.
644
645@item EXPR_LINENO
646This macro returns the line number at which the entity was declared, as
647an @code{int}.
648
649@item DECL_ARTIFICIAL
650This predicate holds if the declaration was implicitly generated by the
651compiler.  For example, this predicate will hold of an implicitly
652declared member function, or of the @code{TYPE_DECL} implicitly
653generated for a class type.  Recall that in C++ code like:
654@smallexample
655struct S @{@};
656@end smallexample
657@noindent
658is roughly equivalent to C code like:
659@smallexample
660struct S @{@};
661typedef struct S S;
662@end smallexample
663The implicitly generated @code{typedef} declaration is represented by a
664@code{TYPE_DECL} for which @code{DECL_ARTIFICIAL} holds.
665
666@end ftable
667
668The various kinds of declarations include:
669@table @code
670@item LABEL_DECL
671These nodes are used to represent labels in function bodies.  For more
672information, see @ref{Functions}.  These nodes only appear in block
673scopes.
674
675@item CONST_DECL
676These nodes are used to represent enumeration constants.  The value of
677the constant is given by @code{DECL_INITIAL} which will be an
678@code{INTEGER_CST} with the same type as the @code{TREE_TYPE} of the
679@code{CONST_DECL}, i.e., an @code{ENUMERAL_TYPE}.
680
681@item RESULT_DECL
682These nodes represent the value returned by a function.  When a value is
683assigned to a @code{RESULT_DECL}, that indicates that the value should
684be returned, via bitwise copy, by the function.  You can use
685@code{DECL_SIZE} and @code{DECL_ALIGN} on a @code{RESULT_DECL}, just as
686with a @code{VAR_DECL}.
687
688@item TYPE_DECL
689These nodes represent @code{typedef} declarations.  The @code{TREE_TYPE}
690is the type declared to have the name given by @code{DECL_NAME}.  In
691some cases, there is no associated name.
692
693@item VAR_DECL
694These nodes represent variables with namespace or block scope, as well
695as static data members.  The @code{DECL_SIZE} and @code{DECL_ALIGN} are
696analogous to @code{TYPE_SIZE} and @code{TYPE_ALIGN}.  For a declaration,
697you should always use the @code{DECL_SIZE} and @code{DECL_ALIGN} rather
698than the @code{TYPE_SIZE} and @code{TYPE_ALIGN} given by the
699@code{TREE_TYPE}, since special attributes may have been applied to the
700variable to give it a particular size and alignment.  You may use the
701predicates @code{DECL_THIS_STATIC} or @code{DECL_THIS_EXTERN} to test
702whether the storage class specifiers @code{static} or @code{extern} were
703used to declare a variable.
704
705If this variable is initialized (but does not require a constructor),
706the @code{DECL_INITIAL} will be an expression for the initializer.  The
707initializer should be evaluated, and a bitwise copy into the variable
708performed.  If the @code{DECL_INITIAL} is the @code{error_mark_node},
709there is an initializer, but it is given by an explicit statement later
710in the code; no bitwise copy is required.
711
712GCC provides an extension that allows either automatic variables, or
713global variables, to be placed in particular registers.  This extension
714is being used for a particular @code{VAR_DECL} if @code{DECL_REGISTER}
715holds for the @code{VAR_DECL}, and if @code{DECL_ASSEMBLER_NAME} is not
716equal to @code{DECL_NAME}.  In that case, @code{DECL_ASSEMBLER_NAME} is
717the name of the register into which the variable will be placed.
718
719@item PARM_DECL
720Used to represent a parameter to a function.  Treat these nodes
721similarly to @code{VAR_DECL} nodes.  These nodes only appear in the
722@code{DECL_ARGUMENTS} for a @code{FUNCTION_DECL}.
723
724The @code{DECL_ARG_TYPE} for a @code{PARM_DECL} is the type that will
725actually be used when a value is passed to this function.  It may be a
726wider type than the @code{TREE_TYPE} of the parameter; for example, the
727ordinary type might be @code{short} while the @code{DECL_ARG_TYPE} is
728@code{int}.
729
730@item DEBUG_EXPR_DECL
731Used to represent an anonymous debug-information temporary created to
732hold an expression as it is optimized away, so that its value can be
733referenced in debug bind statements.
734
735@item FIELD_DECL
736These nodes represent non-static data members.  The @code{DECL_SIZE} and
737@code{DECL_ALIGN} behave as for @code{VAR_DECL} nodes.
738The position of the field within the parent record is specified by a
739combination of three attributes.  @code{DECL_FIELD_OFFSET} is the position,
740counting in bytes, of the @code{DECL_OFFSET_ALIGN}-bit sized word containing
741the bit of the field closest to the beginning of the structure.
742@code{DECL_FIELD_BIT_OFFSET} is the bit offset of the first bit of the field
743within this word; this may be nonzero even for fields that are not bit-fields,
744since @code{DECL_OFFSET_ALIGN} may be greater than the natural alignment
745of the field's type.
746
747If @code{DECL_C_BIT_FIELD} holds, this field is a bit-field.  In a bit-field,
748@code{DECL_BIT_FIELD_TYPE} also contains the type that was originally
749specified for it, while DECL_TYPE may be a modified type with lesser precision,
750according to the size of the bit field.
751
752@item NAMESPACE_DECL
753Namespaces provide a name hierarchy for other declarations.  They
754appear in the @code{DECL_CONTEXT} of other @code{_DECL} nodes.
755
756@end table
757
758@node Internal structure
759@subsection Internal structure
760
761@code{DECL} nodes are represented internally as a hierarchy of
762structures.
763
764@menu
765* Current structure hierarchy::  The current DECL node structure
766hierarchy.
767* Adding new DECL node types:: How to add a new DECL node to a
768frontend.
769@end menu
770
771@node Current structure hierarchy
772@subsubsection Current structure hierarchy
773
774@table @code
775
776@item struct tree_decl_minimal
777This is the minimal structure to inherit from in order for common
778@code{DECL} macros to work.  The fields it contains are a unique ID,
779source location, context, and name.
780
781@item struct tree_decl_common
782This structure inherits from @code{struct tree_decl_minimal}.  It
783contains fields that most @code{DECL} nodes need, such as a field to
784store alignment, machine mode, size, and attributes.
785
786@item struct tree_field_decl
787This structure inherits from @code{struct tree_decl_common}.  It is
788used to represent @code{FIELD_DECL}.
789
790@item struct tree_label_decl
791This structure inherits from @code{struct tree_decl_common}.  It is
792used to represent @code{LABEL_DECL}.
793
794@item struct tree_translation_unit_decl
795This structure inherits from @code{struct tree_decl_common}.  It is
796used to represent @code{TRANSLATION_UNIT_DECL}.
797
798@item struct tree_decl_with_rtl
799This structure inherits from @code{struct tree_decl_common}.  It
800contains a field to store the low-level RTL associated with a
801@code{DECL} node.
802
803@item struct tree_result_decl
804This structure inherits from @code{struct tree_decl_with_rtl}.  It is
805used to represent @code{RESULT_DECL}.
806
807@item struct tree_const_decl
808This structure inherits from @code{struct tree_decl_with_rtl}.  It is
809used to represent @code{CONST_DECL}.
810
811@item struct tree_parm_decl
812This structure inherits from @code{struct tree_decl_with_rtl}.  It is
813used to represent @code{PARM_DECL}.
814
815@item struct tree_decl_with_vis
816This structure inherits from @code{struct tree_decl_with_rtl}.  It
817contains fields necessary to store visibility information, as well as
818a section name and assembler name.
819
820@item struct tree_var_decl
821This structure inherits from @code{struct tree_decl_with_vis}.  It is
822used to represent @code{VAR_DECL}.
823
824@item struct tree_function_decl
825This structure inherits from @code{struct tree_decl_with_vis}.  It is
826used to represent @code{FUNCTION_DECL}.
827
828@end table
829@node Adding new DECL node types
830@subsubsection Adding new DECL node types
831
832Adding a new @code{DECL} tree consists of the following steps
833
834@table @asis
835
836@item Add a new tree code for the @code{DECL} node
837For language specific @code{DECL} nodes, there is a @file{.def} file
838in each frontend directory where the tree code should be added.
839For @code{DECL} nodes that are part of the middle-end, the code should
840be added to @file{tree.def}.
841
842@item Create a new structure type for the @code{DECL} node
843These structures should inherit from one of the existing structures in
844the language hierarchy by using that structure as the first member.
845
846@smallexample
847struct tree_foo_decl
848@{
849   struct tree_decl_with_vis common;
850@}
851@end smallexample
852
853Would create a structure name @code{tree_foo_decl} that inherits from
854@code{struct tree_decl_with_vis}.
855
856For language specific @code{DECL} nodes, this new structure type
857should go in the appropriate @file{.h} file.
858For @code{DECL} nodes that are part of the middle-end, the structure
859type should go in @file{tree.h}.
860
861@item Add a member to the tree structure enumerator for the node
862For garbage collection and dynamic checking purposes, each @code{DECL}
863node structure type is required to have a unique enumerator value
864specified with it.
865For language specific @code{DECL} nodes, this new enumerator value
866should go in the appropriate @file{.def} file.
867For @code{DECL} nodes that are part of the middle-end, the enumerator
868values are specified in @file{treestruct.def}.
869
870@item Update @code{union tree_node}
871In order to make your new structure type usable, it must be added to
872@code{union tree_node}.
873For language specific @code{DECL} nodes, a new entry should be added
874to the appropriate @file{.h} file of the form
875@smallexample
876  struct tree_foo_decl GTY ((tag ("TS_VAR_DECL"))) foo_decl;
877@end smallexample
878For @code{DECL} nodes that are part of the middle-end, the additional
879member goes directly into @code{union tree_node} in @file{tree.h}.
880
881@item Update dynamic checking info
882In order to be able to check whether accessing a named portion of
883@code{union tree_node} is legal, and whether a certain @code{DECL} node
884contains one of the enumerated @code{DECL} node structures in the
885hierarchy, a simple lookup table is used.
886This lookup table needs to be kept up to date with the tree structure
887hierarchy, or else checking and containment macros will fail
888inappropriately.
889
890For language specific @code{DECL} nodes, their is an @code{init_ts}
891function in an appropriate @file{.c} file, which initializes the lookup
892table.
893Code setting up the table for new @code{DECL} nodes should be added
894there.
895For each @code{DECL} tree code and enumerator value representing a
896member of the inheritance  hierarchy, the table should contain 1 if
897that tree code inherits (directly or indirectly) from that member.
898Thus, a @code{FOO_DECL} node derived from @code{struct decl_with_rtl},
899and enumerator value @code{TS_FOO_DECL}, would be set up as follows
900@smallexample
901tree_contains_struct[FOO_DECL][TS_FOO_DECL] = 1;
902tree_contains_struct[FOO_DECL][TS_DECL_WRTL] = 1;
903tree_contains_struct[FOO_DECL][TS_DECL_COMMON] = 1;
904tree_contains_struct[FOO_DECL][TS_DECL_MINIMAL] = 1;
905@end smallexample
906
907For @code{DECL} nodes that are part of the middle-end, the setup code
908goes into @file{tree.c}.
909
910@item Add macros to access any new fields and flags
911
912Each added field or flag should have a macro that is used to access
913it, that performs appropriate checking to ensure only the right type of
914@code{DECL} nodes access the field.
915
916These macros generally take the following form
917@smallexample
918#define FOO_DECL_FIELDNAME(NODE) FOO_DECL_CHECK(NODE)->foo_decl.fieldname
919@end smallexample
920However, if the structure is simply a base class for further
921structures, something like the following should be used
922@smallexample
923#define BASE_STRUCT_CHECK(T) CONTAINS_STRUCT_CHECK(T, TS_BASE_STRUCT)
924#define BASE_STRUCT_FIELDNAME(NODE) \
925   (BASE_STRUCT_CHECK(NODE)->base_struct.fieldname
926@end smallexample
927
928@end table
929
930
931@c ---------------------------------------------------------------------
932@c Attributes
933@c ---------------------------------------------------------------------
934@node Attributes
935@section Attributes in trees
936@cindex attributes
937
938Attributes, as specified using the @code{__attribute__} keyword, are
939represented internally as a @code{TREE_LIST}.  The @code{TREE_PURPOSE}
940is the name of the attribute, as an @code{IDENTIFIER_NODE}.  The
941@code{TREE_VALUE} is a @code{TREE_LIST} of the arguments of the
942attribute, if any, or @code{NULL_TREE} if there are no arguments; the
943arguments are stored as the @code{TREE_VALUE} of successive entries in
944the list, and may be identifiers or expressions.  The @code{TREE_CHAIN}
945of the attribute is the next attribute in a list of attributes applying
946to the same declaration or type, or @code{NULL_TREE} if there are no
947further attributes in the list.
948
949Attributes may be attached to declarations and to types; these
950attributes may be accessed with the following macros.  All attributes
951are stored in this way, and many also cause other changes to the
952declaration or type or to other internal compiler data structures.
953
954@deftypefn {Tree Macro} tree DECL_ATTRIBUTES (tree @var{decl})
955This macro returns the attributes on the declaration @var{decl}.
956@end deftypefn
957
958@deftypefn {Tree Macro} tree TYPE_ATTRIBUTES (tree @var{type})
959This macro returns the attributes on the type @var{type}.
960@end deftypefn
961
962
963@c ---------------------------------------------------------------------
964@c Expressions
965@c ---------------------------------------------------------------------
966
967@node Expression trees
968@section Expressions
969@cindex expression
970@findex TREE_TYPE
971@findex TREE_OPERAND
972
973The internal representation for expressions is for the most part quite
974straightforward.  However, there are a few facts that one must bear in
975mind.  In particular, the expression ``tree'' is actually a directed
976acyclic graph.  (For example there may be many references to the integer
977constant zero throughout the source program; many of these will be
978represented by the same expression node.)  You should not rely on
979certain kinds of node being shared, nor should you rely on certain kinds of
980nodes being unshared.
981
982The following macros can be used with all expression nodes:
983
984@ftable @code
985@item TREE_TYPE
986Returns the type of the expression.  This value may not be precisely the
987same type that would be given the expression in the original program.
988@end ftable
989
990In what follows, some nodes that one might expect to always have type
991@code{bool} are documented to have either integral or boolean type.  At
992some point in the future, the C front end may also make use of this same
993intermediate representation, and at this point these nodes will
994certainly have integral type.  The previous sentence is not meant to
995imply that the C++ front end does not or will not give these nodes
996integral type.
997
998Below, we list the various kinds of expression nodes.  Except where
999noted otherwise, the operands to an expression are accessed using the
1000@code{TREE_OPERAND} macro.  For example, to access the first operand to
1001a binary plus expression @code{expr}, use:
1002
1003@smallexample
1004TREE_OPERAND (expr, 0)
1005@end smallexample
1006@noindent
1007
1008As this example indicates, the operands are zero-indexed.
1009
1010
1011@menu
1012* Constants: Constant expressions.
1013* Storage References::
1014* Unary and Binary Expressions::
1015* Vectors::
1016@end menu
1017
1018@node Constant expressions
1019@subsection Constant expressions
1020@tindex INTEGER_CST
1021@findex TREE_INT_CST_HIGH
1022@findex TREE_INT_CST_LOW
1023@findex tree_int_cst_lt
1024@findex tree_int_cst_equal
1025@tindex REAL_CST
1026@tindex FIXED_CST
1027@tindex COMPLEX_CST
1028@tindex VECTOR_CST
1029@tindex STRING_CST
1030@findex TREE_STRING_LENGTH
1031@findex TREE_STRING_POINTER
1032
1033The table below begins with constants, moves on to unary expressions,
1034then proceeds to binary expressions, and concludes with various other
1035kinds of expressions:
1036
1037@table @code
1038@item INTEGER_CST
1039These nodes represent integer constants.  Note that the type of these
1040constants is obtained with @code{TREE_TYPE}; they are not always of type
1041@code{int}.  In particular, @code{char} constants are represented with
1042@code{INTEGER_CST} nodes.  The value of the integer constant @code{e} is
1043given by
1044@smallexample
1045((TREE_INT_CST_HIGH (e) << HOST_BITS_PER_WIDE_INT)
1046+ TREE_INST_CST_LOW (e))
1047@end smallexample
1048@noindent
1049HOST_BITS_PER_WIDE_INT is at least thirty-two on all platforms.  Both
1050@code{TREE_INT_CST_HIGH} and @code{TREE_INT_CST_LOW} return a
1051@code{HOST_WIDE_INT}.  The value of an @code{INTEGER_CST} is interpreted
1052as a signed or unsigned quantity depending on the type of the constant.
1053In general, the expression given above will overflow, so it should not
1054be used to calculate the value of the constant.
1055
1056The variable @code{integer_zero_node} is an integer constant with value
1057zero.  Similarly, @code{integer_one_node} is an integer constant with
1058value one.  The @code{size_zero_node} and @code{size_one_node} variables
1059are analogous, but have type @code{size_t} rather than @code{int}.
1060
1061The function @code{tree_int_cst_lt} is a predicate which holds if its
1062first argument is less than its second.  Both constants are assumed to
1063have the same signedness (i.e., either both should be signed or both
1064should be unsigned.)  The full width of the constant is used when doing
1065the comparison; the usual rules about promotions and conversions are
1066ignored.  Similarly, @code{tree_int_cst_equal} holds if the two
1067constants are equal.  The @code{tree_int_cst_sgn} function returns the
1068sign of a constant.  The value is @code{1}, @code{0}, or @code{-1}
1069according on whether the constant is greater than, equal to, or less
1070than zero.  Again, the signedness of the constant's type is taken into
1071account; an unsigned constant is never less than zero, no matter what
1072its bit-pattern.
1073
1074@item REAL_CST
1075
1076FIXME: Talk about how to obtain representations of this constant, do
1077comparisons, and so forth.
1078
1079@item FIXED_CST
1080
1081These nodes represent fixed-point constants.  The type of these constants
1082is obtained with @code{TREE_TYPE}.  @code{TREE_FIXED_CST_PTR} points to
1083a @code{struct fixed_value};  @code{TREE_FIXED_CST} returns the structure
1084itself.  @code{struct fixed_value} contains @code{data} with the size of two
1085@code{HOST_BITS_PER_WIDE_INT} and @code{mode} as the associated fixed-point
1086machine mode for @code{data}.
1087
1088@item COMPLEX_CST
1089These nodes are used to represent complex number constants, that is a
1090@code{__complex__} whose parts are constant nodes.  The
1091@code{TREE_REALPART} and @code{TREE_IMAGPART} return the real and the
1092imaginary parts respectively.
1093
1094@item VECTOR_CST
1095These nodes are used to represent vector constants, whose parts are
1096constant nodes.  Each individual constant node is either an integer or a
1097double constant node.  The first operand is a @code{TREE_LIST} of the
1098constant nodes and is accessed through @code{TREE_VECTOR_CST_ELTS}.
1099
1100@item STRING_CST
1101These nodes represent string-constants.  The @code{TREE_STRING_LENGTH}
1102returns the length of the string, as an @code{int}.  The
1103@code{TREE_STRING_POINTER} is a @code{char*} containing the string
1104itself.  The string may not be @code{NUL}-terminated, and it may contain
1105embedded @code{NUL} characters.  Therefore, the
1106@code{TREE_STRING_LENGTH} includes the trailing @code{NUL} if it is
1107present.
1108
1109For wide string constants, the @code{TREE_STRING_LENGTH} is the number
1110of bytes in the string, and the @code{TREE_STRING_POINTER}
1111points to an array of the bytes of the string, as represented on the
1112target system (that is, as integers in the target endianness).  Wide and
1113non-wide string constants are distinguished only by the @code{TREE_TYPE}
1114of the @code{STRING_CST}.
1115
1116FIXME: The formats of string constants are not well-defined when the
1117target system bytes are not the same width as host system bytes.
1118
1119@end table
1120
1121@node Storage References
1122@subsection References to storage
1123@tindex ADDR_EXPR
1124@tindex INDIRECT_REF
1125@tindex ARRAY_REF
1126@tindex ARRAY_RANGE_REF
1127@tindex TARGET_MEM_REF
1128@tindex COMPONENT_REF
1129
1130@table @code
1131@item ARRAY_REF
1132These nodes represent array accesses.  The first operand is the array;
1133the second is the index.  To calculate the address of the memory
1134accessed, you must scale the index by the size of the type of the array
1135elements.  The type of these expressions must be the type of a component of
1136the array.  The third and fourth operands are used after gimplification
1137to represent the lower bound and component size but should not be used
1138directly; call @code{array_ref_low_bound} and @code{array_ref_element_size}
1139instead.
1140
1141@item ARRAY_RANGE_REF
1142These nodes represent access to a range (or ``slice'') of an array.  The
1143operands are the same as that for @code{ARRAY_REF} and have the same
1144meanings.  The type of these expressions must be an array whose component
1145type is the same as that of the first operand.  The range of that array
1146type determines the amount of data these expressions access.
1147
1148@item TARGET_MEM_REF
1149These nodes represent memory accesses whose address directly map to
1150an addressing mode of the target architecture.  The first argument
1151is @code{TMR_SYMBOL} and must be a @code{VAR_DECL} of an object with
1152a fixed address.  The second argument is @code{TMR_BASE} and the
1153third one is @code{TMR_INDEX}.  The fourth argument is
1154@code{TMR_STEP} and must be an @code{INTEGER_CST}.  The fifth
1155argument is @code{TMR_OFFSET} and must be an @code{INTEGER_CST}.
1156Any of the arguments may be NULL if the appropriate component
1157does not appear in the address.  Address of the @code{TARGET_MEM_REF}
1158is determined in the following way.
1159
1160@smallexample
1161&TMR_SYMBOL + TMR_BASE + TMR_INDEX * TMR_STEP + TMR_OFFSET
1162@end smallexample
1163
1164The sixth argument is the reference to the original memory access, which
1165is preserved for the purposes of the RTL alias analysis.  The seventh
1166argument is a tag representing the results of tree level alias analysis.
1167
1168@item ADDR_EXPR
1169These nodes are used to represent the address of an object.  (These
1170expressions will always have pointer or reference type.)  The operand may
1171be another expression, or it may be a declaration.
1172
1173As an extension, GCC allows users to take the address of a label.  In
1174this case, the operand of the @code{ADDR_EXPR} will be a
1175@code{LABEL_DECL}.  The type of such an expression is @code{void*}.
1176
1177If the object addressed is not an lvalue, a temporary is created, and
1178the address of the temporary is used.
1179
1180@item INDIRECT_REF
1181These nodes are used to represent the object pointed to by a pointer.
1182The operand is the pointer being dereferenced; it will always have
1183pointer or reference type.
1184
1185@item COMPONENT_REF
1186These nodes represent non-static data member accesses.  The first
1187operand is the object (rather than a pointer to it); the second operand
1188is the @code{FIELD_DECL} for the data member.  The third operand represents
1189the byte offset of the field, but should not be used directly; call
1190@code{component_ref_field_offset} instead.
1191
1192
1193@end table
1194
1195@node Unary and Binary Expressions
1196@subsection Unary and Binary Expressions
1197@tindex NEGATE_EXPR
1198@tindex ABS_EXPR
1199@tindex BIT_NOT_EXPR
1200@tindex TRUTH_NOT_EXPR
1201@tindex PREDECREMENT_EXPR
1202@tindex PREINCREMENT_EXPR
1203@tindex POSTDECREMENT_EXPR
1204@tindex POSTINCREMENT_EXPR
1205@tindex FIX_TRUNC_EXPR
1206@tindex FLOAT_EXPR
1207@tindex COMPLEX_EXPR
1208@tindex CONJ_EXPR
1209@tindex REALPART_EXPR
1210@tindex IMAGPART_EXPR
1211@tindex NON_LVALUE_EXPR
1212@tindex NOP_EXPR
1213@tindex CONVERT_EXPR
1214@tindex FIXED_CONVERT_EXPR
1215@tindex THROW_EXPR
1216@tindex LSHIFT_EXPR
1217@tindex RSHIFT_EXPR
1218@tindex BIT_IOR_EXPR
1219@tindex BIT_XOR_EXPR
1220@tindex BIT_AND_EXPR
1221@tindex TRUTH_ANDIF_EXPR
1222@tindex TRUTH_ORIF_EXPR
1223@tindex TRUTH_AND_EXPR
1224@tindex TRUTH_OR_EXPR
1225@tindex TRUTH_XOR_EXPR
1226@tindex POINTER_PLUS_EXPR
1227@tindex PLUS_EXPR
1228@tindex MINUS_EXPR
1229@tindex MULT_EXPR
1230@tindex RDIV_EXPR
1231@tindex TRUNC_DIV_EXPR
1232@tindex FLOOR_DIV_EXPR
1233@tindex CEIL_DIV_EXPR
1234@tindex ROUND_DIV_EXPR
1235@tindex TRUNC_MOD_EXPR
1236@tindex FLOOR_MOD_EXPR
1237@tindex CEIL_MOD_EXPR
1238@tindex ROUND_MOD_EXPR
1239@tindex EXACT_DIV_EXPR
1240@tindex LT_EXPR
1241@tindex LE_EXPR
1242@tindex GT_EXPR
1243@tindex GE_EXPR
1244@tindex EQ_EXPR
1245@tindex NE_EXPR
1246@tindex ORDERED_EXPR
1247@tindex UNORDERED_EXPR
1248@tindex UNLT_EXPR
1249@tindex UNLE_EXPR
1250@tindex UNGT_EXPR
1251@tindex UNGE_EXPR
1252@tindex UNEQ_EXPR
1253@tindex LTGT_EXPR
1254@tindex MODIFY_EXPR
1255@tindex INIT_EXPR
1256@tindex COMPOUND_EXPR
1257@tindex COND_EXPR
1258@tindex CALL_EXPR
1259@tindex STMT_EXPR
1260@tindex BIND_EXPR
1261@tindex LOOP_EXPR
1262@tindex EXIT_EXPR
1263@tindex CLEANUP_POINT_EXPR
1264@tindex CONSTRUCTOR
1265@tindex COMPOUND_LITERAL_EXPR
1266@tindex SAVE_EXPR
1267@tindex TARGET_EXPR
1268@tindex VA_ARG_EXPR
1269
1270@table @code
1271@item NEGATE_EXPR
1272These nodes represent unary negation of the single operand, for both
1273integer and floating-point types.  The type of negation can be
1274determined by looking at the type of the expression.
1275
1276The behavior of this operation on signed arithmetic overflow is
1277controlled by the @code{flag_wrapv} and @code{flag_trapv} variables.
1278
1279@item ABS_EXPR
1280These nodes represent the absolute value of the single operand, for
1281both integer and floating-point types.  This is typically used to
1282implement the @code{abs}, @code{labs} and @code{llabs} builtins for
1283integer types, and the @code{fabs}, @code{fabsf} and @code{fabsl}
1284builtins for floating point types.  The type of abs operation can
1285be determined by looking at the type of the expression.
1286
1287This node is not used for complex types.  To represent the modulus
1288or complex abs of a complex value, use the @code{BUILT_IN_CABS},
1289@code{BUILT_IN_CABSF} or @code{BUILT_IN_CABSL} builtins, as used
1290to implement the C99 @code{cabs}, @code{cabsf} and @code{cabsl}
1291built-in functions.
1292
1293@item BIT_NOT_EXPR
1294These nodes represent bitwise complement, and will always have integral
1295type.  The only operand is the value to be complemented.
1296
1297@item TRUTH_NOT_EXPR
1298These nodes represent logical negation, and will always have integral
1299(or boolean) type.  The operand is the value being negated.  The type
1300of the operand and that of the result are always of @code{BOOLEAN_TYPE}
1301or @code{INTEGER_TYPE}.
1302
1303@item PREDECREMENT_EXPR
1304@itemx PREINCREMENT_EXPR
1305@itemx POSTDECREMENT_EXPR
1306@itemx POSTINCREMENT_EXPR
1307These nodes represent increment and decrement expressions.  The value of
1308the single operand is computed, and the operand incremented or
1309decremented.  In the case of @code{PREDECREMENT_EXPR} and
1310@code{PREINCREMENT_EXPR}, the value of the expression is the value
1311resulting after the increment or decrement; in the case of
1312@code{POSTDECREMENT_EXPR} and @code{POSTINCREMENT_EXPR} is the value
1313before the increment or decrement occurs.  The type of the operand, like
1314that of the result, will be either integral, boolean, or floating-point.
1315
1316@item FIX_TRUNC_EXPR
1317These nodes represent conversion of a floating-point value to an
1318integer.  The single operand will have a floating-point type, while
1319the complete expression will have an integral (or boolean) type.  The
1320operand is rounded towards zero.
1321
1322@item FLOAT_EXPR
1323These nodes represent conversion of an integral (or boolean) value to a
1324floating-point value.  The single operand will have integral type, while
1325the complete expression will have a floating-point type.
1326
1327FIXME: How is the operand supposed to be rounded?  Is this dependent on
1328@option{-mieee}?
1329
1330@item COMPLEX_EXPR
1331These nodes are used to represent complex numbers constructed from two
1332expressions of the same (integer or real) type.  The first operand is the
1333real part and the second operand is the imaginary part.
1334
1335@item CONJ_EXPR
1336These nodes represent the conjugate of their operand.
1337
1338@item REALPART_EXPR
1339@itemx IMAGPART_EXPR
1340These nodes represent respectively the real and the imaginary parts
1341of complex numbers (their sole argument).
1342
1343@item NON_LVALUE_EXPR
1344These nodes indicate that their one and only operand is not an lvalue.
1345A back end can treat these identically to the single operand.
1346
1347@item NOP_EXPR
1348These nodes are used to represent conversions that do not require any
1349code-generation.  For example, conversion of a @code{char*} to an
1350@code{int*} does not require any code be generated; such a conversion is
1351represented by a @code{NOP_EXPR}.  The single operand is the expression
1352to be converted.  The conversion from a pointer to a reference is also
1353represented with a @code{NOP_EXPR}.
1354
1355@item CONVERT_EXPR
1356These nodes are similar to @code{NOP_EXPR}s, but are used in those
1357situations where code may need to be generated.  For example, if an
1358@code{int*} is converted to an @code{int} code may need to be generated
1359on some platforms.  These nodes are never used for C++-specific
1360conversions, like conversions between pointers to different classes in
1361an inheritance hierarchy.  Any adjustments that need to be made in such
1362cases are always indicated explicitly.  Similarly, a user-defined
1363conversion is never represented by a @code{CONVERT_EXPR}; instead, the
1364function calls are made explicit.
1365
1366@item FIXED_CONVERT_EXPR
1367These nodes are used to represent conversions that involve fixed-point
1368values.  For example, from a fixed-point value to another fixed-point value,
1369from an integer to a fixed-point value, from a fixed-point value to an
1370integer, from a floating-point value to a fixed-point value, or from
1371a fixed-point value to a floating-point value.
1372
1373@item LSHIFT_EXPR
1374@itemx RSHIFT_EXPR
1375These nodes represent left and right shifts, respectively.  The first
1376operand is the value to shift; it will always be of integral type.  The
1377second operand is an expression for the number of bits by which to
1378shift.  Right shift should be treated as arithmetic, i.e., the
1379high-order bits should be zero-filled when the expression has unsigned
1380type and filled with the sign bit when the expression has signed type.
1381Note that the result is undefined if the second operand is larger
1382than or equal to the first operand's type size.
1383
1384
1385@item BIT_IOR_EXPR
1386@itemx BIT_XOR_EXPR
1387@itemx BIT_AND_EXPR
1388These nodes represent bitwise inclusive or, bitwise exclusive or, and
1389bitwise and, respectively.  Both operands will always have integral
1390type.
1391
1392@item TRUTH_ANDIF_EXPR
1393@itemx TRUTH_ORIF_EXPR
1394These nodes represent logical ``and'' and logical ``or'', respectively.
1395These operators are not strict; i.e., the second operand is evaluated
1396only if the value of the expression is not determined by evaluation of
1397the first operand.  The type of the operands and that of the result are
1398always of @code{BOOLEAN_TYPE} or @code{INTEGER_TYPE}.
1399
1400@item TRUTH_AND_EXPR
1401@itemx TRUTH_OR_EXPR
1402@itemx TRUTH_XOR_EXPR
1403These nodes represent logical and, logical or, and logical exclusive or.
1404They are strict; both arguments are always evaluated.  There are no
1405corresponding operators in C or C++, but the front end will sometimes
1406generate these expressions anyhow, if it can tell that strictness does
1407not matter.  The type of the operands and that of the result are
1408always of @code{BOOLEAN_TYPE} or @code{INTEGER_TYPE}.
1409
1410@itemx POINTER_PLUS_EXPR
1411This node represents pointer arithmetic.  The first operand is always
1412a pointer/reference type.  The second operand is always an unsigned
1413integer type compatible with sizetype.  This is the only binary
1414arithmetic operand that can operate on pointer types.
1415
1416@itemx PLUS_EXPR
1417@itemx MINUS_EXPR
1418@itemx MULT_EXPR
1419These nodes represent various binary arithmetic operations.
1420Respectively, these operations are addition, subtraction (of the second
1421operand from the first) and multiplication.  Their operands may have
1422either integral or floating type, but there will never be case in which
1423one operand is of floating type and the other is of integral type.
1424
1425The behavior of these operations on signed arithmetic overflow is
1426controlled by the @code{flag_wrapv} and @code{flag_trapv} variables.
1427
1428@item RDIV_EXPR
1429This node represents a floating point division operation.
1430
1431@item TRUNC_DIV_EXPR
1432@itemx FLOOR_DIV_EXPR
1433@itemx CEIL_DIV_EXPR
1434@itemx ROUND_DIV_EXPR
1435These nodes represent integer division operations that return an integer
1436result.  @code{TRUNC_DIV_EXPR} rounds towards zero, @code{FLOOR_DIV_EXPR}
1437rounds towards negative infinity, @code{CEIL_DIV_EXPR} rounds towards
1438positive infinity and @code{ROUND_DIV_EXPR} rounds to the closest integer.
1439Integer division in C and C++ is truncating, i.e.@: @code{TRUNC_DIV_EXPR}.
1440
1441The behavior of these operations on signed arithmetic overflow, when
1442dividing the minimum signed integer by minus one, is controlled by the
1443@code{flag_wrapv} and @code{flag_trapv} variables.
1444
1445@item TRUNC_MOD_EXPR
1446@itemx FLOOR_MOD_EXPR
1447@itemx CEIL_MOD_EXPR
1448@itemx ROUND_MOD_EXPR
1449These nodes represent the integer remainder or modulus operation.
1450The integer modulus of two operands @code{a} and @code{b} is
1451defined as @code{a - (a/b)*b} where the division calculated using
1452the corresponding division operator.  Hence for @code{TRUNC_MOD_EXPR}
1453this definition assumes division using truncation towards zero, i.e.@:
1454@code{TRUNC_DIV_EXPR}.  Integer remainder in C and C++ uses truncating
1455division, i.e.@: @code{TRUNC_MOD_EXPR}.
1456
1457@item EXACT_DIV_EXPR
1458The @code{EXACT_DIV_EXPR} code is used to represent integer divisions where
1459the numerator is known to be an exact multiple of the denominator.  This
1460allows the backend to choose between the faster of @code{TRUNC_DIV_EXPR},
1461@code{CEIL_DIV_EXPR} and @code{FLOOR_DIV_EXPR} for the current target.
1462
1463@item LT_EXPR
1464@itemx LE_EXPR
1465@itemx GT_EXPR
1466@itemx GE_EXPR
1467@itemx EQ_EXPR
1468@itemx NE_EXPR
1469These nodes represent the less than, less than or equal to, greater
1470than, greater than or equal to, equal, and not equal comparison
1471operators.  The first and second operand with either be both of integral
1472type or both of floating type.  The result type of these expressions
1473will always be of integral or boolean type.  These operations return
1474the result type's zero value for false, and the result type's one value
1475for true.
1476
1477For floating point comparisons, if we honor IEEE NaNs and either operand
1478is NaN, then @code{NE_EXPR} always returns true and the remaining operators
1479always return false.  On some targets, comparisons against an IEEE NaN,
1480other than equality and inequality, may generate a floating point exception.
1481
1482@item ORDERED_EXPR
1483@itemx UNORDERED_EXPR
1484These nodes represent non-trapping ordered and unordered comparison
1485operators.  These operations take two floating point operands and
1486determine whether they are ordered or unordered relative to each other.
1487If either operand is an IEEE NaN, their comparison is defined to be
1488unordered, otherwise the comparison is defined to be ordered.  The
1489result type of these expressions will always be of integral or boolean
1490type.  These operations return the result type's zero value for false,
1491and the result type's one value for true.
1492
1493@item UNLT_EXPR
1494@itemx UNLE_EXPR
1495@itemx UNGT_EXPR
1496@itemx UNGE_EXPR
1497@itemx UNEQ_EXPR
1498@itemx LTGT_EXPR
1499These nodes represent the unordered comparison operators.
1500These operations take two floating point operands and determine whether
1501the operands are unordered or are less than, less than or equal to,
1502greater than, greater than or equal to, or equal respectively.  For
1503example, @code{UNLT_EXPR} returns true if either operand is an IEEE
1504NaN or the first operand is less than the second.  With the possible
1505exception of @code{LTGT_EXPR}, all of these operations are guaranteed
1506not to generate a floating point exception.  The result
1507type of these expressions will always be of integral or boolean type.
1508These operations return the result type's zero value for false,
1509and the result type's one value for true.
1510
1511@item MODIFY_EXPR
1512These nodes represent assignment.  The left-hand side is the first
1513operand; the right-hand side is the second operand.  The left-hand side
1514will be a @code{VAR_DECL}, @code{INDIRECT_REF}, @code{COMPONENT_REF}, or
1515other lvalue.
1516
1517These nodes are used to represent not only assignment with @samp{=} but
1518also compound assignments (like @samp{+=}), by reduction to @samp{=}
1519assignment.  In other words, the representation for @samp{i += 3} looks
1520just like that for @samp{i = i + 3}.
1521
1522@item INIT_EXPR
1523These nodes are just like @code{MODIFY_EXPR}, but are used only when a
1524variable is initialized, rather than assigned to subsequently.  This
1525means that we can assume that the target of the initialization is not
1526used in computing its own value; any reference to the lhs in computing
1527the rhs is undefined.
1528
1529@item COMPOUND_EXPR
1530These nodes represent comma-expressions.  The first operand is an
1531expression whose value is computed and thrown away prior to the
1532evaluation of the second operand.  The value of the entire expression is
1533the value of the second operand.
1534
1535@item COND_EXPR
1536These nodes represent @code{?:} expressions.  The first operand
1537is of boolean or integral type.  If it evaluates to a nonzero value,
1538the second operand should be evaluated, and returned as the value of the
1539expression.  Otherwise, the third operand is evaluated, and returned as
1540the value of the expression.
1541
1542The second operand must have the same type as the entire expression,
1543unless it unconditionally throws an exception or calls a noreturn
1544function, in which case it should have void type.  The same constraints
1545apply to the third operand.  This allows array bounds checks to be
1546represented conveniently as @code{(i >= 0 && i < 10) ? i : abort()}.
1547
1548As a GNU extension, the C language front-ends allow the second
1549operand of the @code{?:} operator may be omitted in the source.
1550For example, @code{x ? : 3} is equivalent to @code{x ? x : 3},
1551assuming that @code{x} is an expression without side-effects.
1552In the tree representation, however, the second operand is always
1553present, possibly protected by @code{SAVE_EXPR} if the first
1554argument does cause side-effects.
1555
1556@item CALL_EXPR
1557These nodes are used to represent calls to functions, including
1558non-static member functions.  @code{CALL_EXPR}s are implemented as
1559expression nodes with a variable number of operands.  Rather than using
1560@code{TREE_OPERAND} to extract them, it is preferable to use the
1561specialized accessor macros and functions that operate specifically on
1562@code{CALL_EXPR} nodes.
1563
1564@code{CALL_EXPR_FN} returns a pointer to the
1565function to call; it is always an expression whose type is a
1566@code{POINTER_TYPE}.
1567
1568The number of arguments to the call is returned by @code{call_expr_nargs},
1569while the arguments themselves can be accessed with the @code{CALL_EXPR_ARG}
1570macro.  The arguments are zero-indexed and numbered left-to-right.
1571You can iterate over the arguments using @code{FOR_EACH_CALL_EXPR_ARG}, as in:
1572
1573@smallexample
1574tree call, arg;
1575call_expr_arg_iterator iter;
1576FOR_EACH_CALL_EXPR_ARG (arg, iter, call)
1577  /* arg is bound to successive arguments of call.  */
1578  @dots{};
1579@end smallexample
1580
1581For non-static
1582member functions, there will be an operand corresponding to the
1583@code{this} pointer.  There will always be expressions corresponding to
1584all of the arguments, even if the function is declared with default
1585arguments and some arguments are not explicitly provided at the call
1586sites.
1587
1588@code{CALL_EXPR}s also have a @code{CALL_EXPR_STATIC_CHAIN} operand that
1589is used to implement nested functions.  This operand is otherwise null.
1590
1591@item CLEANUP_POINT_EXPR
1592These nodes represent full-expressions.  The single operand is an
1593expression to evaluate.  Any destructor calls engendered by the creation
1594of temporaries during the evaluation of that expression should be
1595performed immediately after the expression is evaluated.
1596
1597@item CONSTRUCTOR
1598These nodes represent the brace-enclosed initializers for a structure or
1599array.  The first operand is reserved for use by the back end.  The
1600second operand is a @code{TREE_LIST}.  If the @code{TREE_TYPE} of the
1601@code{CONSTRUCTOR} is a @code{RECORD_TYPE} or @code{UNION_TYPE}, then
1602the @code{TREE_PURPOSE} of each node in the @code{TREE_LIST} will be a
1603@code{FIELD_DECL} and the @code{TREE_VALUE} of each node will be the
1604expression used to initialize that field.
1605
1606If the @code{TREE_TYPE} of the @code{CONSTRUCTOR} is an
1607@code{ARRAY_TYPE}, then the @code{TREE_PURPOSE} of each element in the
1608@code{TREE_LIST} will be an @code{INTEGER_CST} or a @code{RANGE_EXPR} of
1609two @code{INTEGER_CST}s.  A single @code{INTEGER_CST} indicates which
1610element of the array (indexed from zero) is being assigned to.  A
1611@code{RANGE_EXPR} indicates an inclusive range of elements to
1612initialize.  In both cases the @code{TREE_VALUE} is the corresponding
1613initializer.  It is re-evaluated for each element of a
1614@code{RANGE_EXPR}.  If the @code{TREE_PURPOSE} is @code{NULL_TREE}, then
1615the initializer is for the next available array element.
1616
1617In the front end, you should not depend on the fields appearing in any
1618particular order.  However, in the middle end, fields must appear in
1619declaration order.  You should not assume that all fields will be
1620represented.  Unrepresented fields will be set to zero.
1621
1622@item COMPOUND_LITERAL_EXPR
1623@findex COMPOUND_LITERAL_EXPR_DECL_EXPR
1624@findex COMPOUND_LITERAL_EXPR_DECL
1625These nodes represent ISO C99 compound literals.  The
1626@code{COMPOUND_LITERAL_EXPR_DECL_EXPR} is a @code{DECL_EXPR}
1627containing an anonymous @code{VAR_DECL} for
1628the unnamed object represented by the compound literal; the
1629@code{DECL_INITIAL} of that @code{VAR_DECL} is a @code{CONSTRUCTOR}
1630representing the brace-enclosed list of initializers in the compound
1631literal.  That anonymous @code{VAR_DECL} can also be accessed directly
1632by the @code{COMPOUND_LITERAL_EXPR_DECL} macro.
1633
1634@item SAVE_EXPR
1635
1636A @code{SAVE_EXPR} represents an expression (possibly involving
1637side-effects) that is used more than once.  The side-effects should
1638occur only the first time the expression is evaluated.  Subsequent uses
1639should just reuse the computed value.  The first operand to the
1640@code{SAVE_EXPR} is the expression to evaluate.  The side-effects should
1641be executed where the @code{SAVE_EXPR} is first encountered in a
1642depth-first preorder traversal of the expression tree.
1643
1644@item TARGET_EXPR
1645A @code{TARGET_EXPR} represents a temporary object.  The first operand
1646is a @code{VAR_DECL} for the temporary variable.  The second operand is
1647the initializer for the temporary.  The initializer is evaluated and,
1648if non-void, copied (bitwise) into the temporary.  If the initializer
1649is void, that means that it will perform the initialization itself.
1650
1651Often, a @code{TARGET_EXPR} occurs on the right-hand side of an
1652assignment, or as the second operand to a comma-expression which is
1653itself the right-hand side of an assignment, etc.  In this case, we say
1654that the @code{TARGET_EXPR} is ``normal''; otherwise, we say it is
1655``orphaned''.  For a normal @code{TARGET_EXPR} the temporary variable
1656should be treated as an alias for the left-hand side of the assignment,
1657rather than as a new temporary variable.
1658
1659The third operand to the @code{TARGET_EXPR}, if present, is a
1660cleanup-expression (i.e., destructor call) for the temporary.  If this
1661expression is orphaned, then this expression must be executed when the
1662statement containing this expression is complete.  These cleanups must
1663always be executed in the order opposite to that in which they were
1664encountered.  Note that if a temporary is created on one branch of a
1665conditional operator (i.e., in the second or third operand to a
1666@code{COND_EXPR}), the cleanup must be run only if that branch is
1667actually executed.
1668
1669@item VA_ARG_EXPR
1670This node is used to implement support for the C/C++ variable argument-list
1671mechanism.  It represents expressions like @code{va_arg (ap, type)}.
1672Its @code{TREE_TYPE} yields the tree representation for @code{type} and
1673its sole argument yields the representation for @code{ap}.
1674
1675@end table
1676
1677@node Vectors
1678@subsection Vectors
1679@tindex VEC_LSHIFT_EXPR
1680@tindex VEC_RSHIFT_EXPR
1681@tindex VEC_WIDEN_MULT_HI_EXPR
1682@tindex VEC_WIDEN_MULT_LO_EXPR
1683@tindex VEC_UNPACK_HI_EXPR
1684@tindex VEC_UNPACK_LO_EXPR
1685@tindex VEC_UNPACK_FLOAT_HI_EXPR
1686@tindex VEC_UNPACK_FLOAT_LO_EXPR
1687@tindex VEC_PACK_TRUNC_EXPR
1688@tindex VEC_PACK_SAT_EXPR
1689@tindex VEC_PACK_FIX_TRUNC_EXPR
1690@tindex VEC_EXTRACT_EVEN_EXPR
1691@tindex VEC_EXTRACT_ODD_EXPR
1692@tindex VEC_INTERLEAVE_HIGH_EXPR
1693@tindex VEC_INTERLEAVE_LOW_EXPR
1694
1695@table @code
1696@item VEC_LSHIFT_EXPR
1697@itemx VEC_RSHIFT_EXPR
1698These nodes represent whole vector left and right shifts, respectively.
1699The first operand is the vector to shift; it will always be of vector type.
1700The second operand is an expression for the number of bits by which to
1701shift.  Note that the result is undefined if the second operand is larger
1702than or equal to the first operand's type size.
1703
1704@item VEC_WIDEN_MULT_HI_EXPR
1705@itemx VEC_WIDEN_MULT_LO_EXPR
1706These nodes represent widening vector multiplication of the high and low
1707parts of the two input vectors, respectively.  Their operands are vectors
1708that contain the same number of elements (@code{N}) of the same integral type.
1709The result is a vector that contains half as many elements, of an integral type
1710whose size is twice as wide.  In the case of @code{VEC_WIDEN_MULT_HI_EXPR} the
1711high @code{N/2} elements of the two vector are multiplied to produce the
1712vector of @code{N/2} products. In the case of @code{VEC_WIDEN_MULT_LO_EXPR} the
1713low @code{N/2} elements of the two vector are multiplied to produce the
1714vector of @code{N/2} products.
1715
1716@item VEC_UNPACK_HI_EXPR
1717@itemx VEC_UNPACK_LO_EXPR
1718These nodes represent unpacking of the high and low parts of the input vector,
1719respectively.  The single operand is a vector that contains @code{N} elements
1720of the same integral or floating point type.  The result is a vector
1721that contains half as many elements, of an integral or floating point type
1722whose size is twice as wide.  In the case of @code{VEC_UNPACK_HI_EXPR} the
1723high @code{N/2} elements of the vector are extracted and widened (promoted).
1724In the case of @code{VEC_UNPACK_LO_EXPR} the low @code{N/2} elements of the
1725vector are extracted and widened (promoted).
1726
1727@item VEC_UNPACK_FLOAT_HI_EXPR
1728@itemx VEC_UNPACK_FLOAT_LO_EXPR
1729These nodes represent unpacking of the high and low parts of the input vector,
1730where the values are converted from fixed point to floating point.  The
1731single operand is a vector that contains @code{N} elements of the same
1732integral type.  The result is a vector that contains half as many elements
1733of a floating point type whose size is twice as wide.  In the case of
1734@code{VEC_UNPACK_HI_EXPR} the high @code{N/2} elements of the vector are
1735extracted, converted and widened.  In the case of @code{VEC_UNPACK_LO_EXPR}
1736the low @code{N/2} elements of the vector are extracted, converted and widened.
1737
1738@item VEC_PACK_TRUNC_EXPR
1739This node represents packing of truncated elements of the two input vectors
1740into the output vector.  Input operands are vectors that contain the same
1741number of elements of the same integral or floating point type.  The result
1742is a vector that contains twice as many elements of an integral or floating
1743point type whose size is half as wide. The elements of the two vectors are
1744demoted and merged (concatenated) to form the output vector.
1745
1746@item VEC_PACK_SAT_EXPR
1747This node represents packing of elements of the two input vectors into the
1748output vector using saturation.  Input operands are vectors that contain
1749the same number of elements of the same integral type.  The result is a
1750vector that contains twice as many elements of an integral type whose size
1751is half as wide.  The elements of the two vectors are demoted and merged
1752(concatenated) to form the output vector.
1753
1754@item VEC_PACK_FIX_TRUNC_EXPR
1755This node represents packing of elements of the two input vectors into the
1756output vector, where the values are converted from floating point
1757to fixed point.  Input operands are vectors that contain the same number
1758of elements of a floating point type.  The result is a vector that contains
1759twice as many elements of an integral type whose size is half as wide.  The
1760elements of the two vectors are merged (concatenated) to form the output
1761vector.
1762
1763@item VEC_EXTRACT_EVEN_EXPR
1764@itemx VEC_EXTRACT_ODD_EXPR
1765These nodes represent extracting of the even/odd elements of the two input
1766vectors, respectively. Their operands and result are vectors that contain the
1767same number of elements of the same type.
1768
1769@item VEC_INTERLEAVE_HIGH_EXPR
1770@itemx VEC_INTERLEAVE_LOW_EXPR
1771These nodes represent merging and interleaving of the high/low elements of the
1772two input vectors, respectively. The operands and the result are vectors that
1773contain the same number of elements (@code{N}) of the same type.
1774In the case of @code{VEC_INTERLEAVE_HIGH_EXPR}, the high @code{N/2} elements of
1775the first input vector are interleaved with the high @code{N/2} elements of the
1776second input vector. In the case of @code{VEC_INTERLEAVE_LOW_EXPR}, the low
1777@code{N/2} elements of the first input vector are interleaved with the low
1778@code{N/2} elements of the second input vector.
1779
1780@end table
1781
1782
1783@c ---------------------------------------------------------------------
1784@c Statements
1785@c ---------------------------------------------------------------------
1786
1787@node Statements
1788@section Statements
1789@cindex Statements
1790
1791Most statements in GIMPLE are assignment statements, represented by
1792@code{GIMPLE_ASSIGN}.  No other C expressions can appear at statement level;
1793a reference to a volatile object is converted into a
1794@code{GIMPLE_ASSIGN}.
1795
1796There are also several varieties of complex statements.
1797
1798@menu
1799* Basic Statements::
1800* Blocks::
1801* Statement Sequences::
1802* Empty Statements::
1803* Jumps::
1804* Cleanups::
1805* OpenMP::
1806@end menu
1807
1808@node Basic Statements
1809@subsection Basic Statements
1810@cindex Basic Statements
1811
1812@table @code
1813@item ASM_EXPR
1814
1815Used to represent an inline assembly statement.  For an inline assembly
1816statement like:
1817@smallexample
1818asm ("mov x, y");
1819@end smallexample
1820The @code{ASM_STRING} macro will return a @code{STRING_CST} node for
1821@code{"mov x, y"}.  If the original statement made use of the
1822extended-assembly syntax, then @code{ASM_OUTPUTS},
1823@code{ASM_INPUTS}, and @code{ASM_CLOBBERS} will be the outputs, inputs,
1824and clobbers for the statement, represented as @code{STRING_CST} nodes.
1825The extended-assembly syntax looks like:
1826@smallexample
1827asm ("fsinx %1,%0" : "=f" (result) : "f" (angle));
1828@end smallexample
1829The first string is the @code{ASM_STRING}, containing the instruction
1830template.  The next two strings are the output and inputs, respectively;
1831this statement has no clobbers.  As this example indicates, ``plain''
1832assembly statements are merely a special case of extended assembly
1833statements; they have no cv-qualifiers, outputs, inputs, or clobbers.
1834All of the strings will be @code{NUL}-terminated, and will contain no
1835embedded @code{NUL}-characters.
1836
1837If the assembly statement is declared @code{volatile}, or if the
1838statement was not an extended assembly statement, and is therefore
1839implicitly volatile, then the predicate @code{ASM_VOLATILE_P} will hold
1840of the @code{ASM_EXPR}.
1841
1842@item DECL_EXPR
1843
1844Used to represent a local declaration.  The @code{DECL_EXPR_DECL} macro
1845can be used to obtain the entity declared.  This declaration may be a
1846@code{LABEL_DECL}, indicating that the label declared is a local label.
1847(As an extension, GCC allows the declaration of labels with scope.)  In
1848C, this declaration may be a @code{FUNCTION_DECL}, indicating the
1849use of the GCC nested function extension.  For more information,
1850@pxref{Functions}.
1851
1852@item LABEL_EXPR
1853
1854Used to represent a label.  The @code{LABEL_DECL} declared by this
1855statement can be obtained with the @code{LABEL_EXPR_LABEL} macro.  The
1856@code{IDENTIFIER_NODE} giving the name of the label can be obtained from
1857the @code{LABEL_DECL} with @code{DECL_NAME}.
1858
1859@item GOTO_EXPR
1860
1861Used to represent a @code{goto} statement.  The @code{GOTO_DESTINATION} will
1862usually be a @code{LABEL_DECL}.  However, if the ``computed goto'' extension
1863has been used, the @code{GOTO_DESTINATION} will be an arbitrary expression
1864indicating the destination.  This expression will always have pointer type.
1865
1866@item RETURN_EXPR
1867
1868Used to represent a @code{return} statement.  Operand 0 represents the
1869value to return.  It should either be the @code{RESULT_DECL} for the
1870containing function, or a @code{MODIFY_EXPR} or @code{INIT_EXPR}
1871setting the function's @code{RESULT_DECL}.  It will be
1872@code{NULL_TREE} if the statement was just
1873@smallexample
1874return;
1875@end smallexample
1876
1877@item LOOP_EXPR
1878These nodes represent ``infinite'' loops.  The @code{LOOP_EXPR_BODY}
1879represents the body of the loop.  It should be executed forever, unless
1880an @code{EXIT_EXPR} is encountered.
1881
1882@item EXIT_EXPR
1883These nodes represent conditional exits from the nearest enclosing
1884@code{LOOP_EXPR}.  The single operand is the condition; if it is
1885nonzero, then the loop should be exited.  An @code{EXIT_EXPR} will only
1886appear within a @code{LOOP_EXPR}.
1887
1888@item SWITCH_STMT
1889
1890Used to represent a @code{switch} statement.  The @code{SWITCH_STMT_COND}
1891is the expression on which the switch is occurring.  See the documentation
1892for an @code{IF_STMT} for more information on the representation used
1893for the condition.  The @code{SWITCH_STMT_BODY} is the body of the switch
1894statement.   The @code{SWITCH_STMT_TYPE} is the original type of switch
1895expression as given in the source, before any compiler conversions.
1896
1897@item CASE_LABEL_EXPR
1898
1899Use to represent a @code{case} label, range of @code{case} labels, or a
1900@code{default} label.  If @code{CASE_LOW} is @code{NULL_TREE}, then this is a
1901@code{default} label.  Otherwise, if @code{CASE_HIGH} is @code{NULL_TREE}, then
1902this is an ordinary @code{case} label.  In this case, @code{CASE_LOW} is
1903an expression giving the value of the label.  Both @code{CASE_LOW} and
1904@code{CASE_HIGH} are @code{INTEGER_CST} nodes.  These values will have
1905the same type as the condition expression in the switch statement.
1906
1907Otherwise, if both @code{CASE_LOW} and @code{CASE_HIGH} are defined, the
1908statement is a range of case labels.  Such statements originate with the
1909extension that allows users to write things of the form:
1910@smallexample
1911case 2 ... 5:
1912@end smallexample
1913The first value will be @code{CASE_LOW}, while the second will be
1914@code{CASE_HIGH}.
1915
1916@end table
1917
1918
1919@node Blocks
1920@subsection Blocks
1921@cindex Blocks
1922
1923Block scopes and the variables they declare in GENERIC are
1924expressed using the @code{BIND_EXPR} code, which in previous
1925versions of GCC was primarily used for the C statement-expression
1926extension.
1927
1928Variables in a block are collected into @code{BIND_EXPR_VARS} in
1929declaration order through their @code{TREE_CHAIN} field.  Any runtime
1930initialization is moved out of @code{DECL_INITIAL} and into a
1931statement in the controlled block.  When gimplifying from C or C++,
1932this initialization replaces the @code{DECL_STMT}.  These variables
1933will never require cleanups.  The scope of these variables is just the
1934body
1935
1936Variable-length arrays (VLAs) complicate this process, as their
1937size often refers to variables initialized earlier in the block.
1938To handle this, we currently split the block at that point, and
1939move the VLA into a new, inner @code{BIND_EXPR}.  This strategy
1940may change in the future.
1941
1942A C++ program will usually contain more @code{BIND_EXPR}s than
1943there are syntactic blocks in the source code, since several C++
1944constructs have implicit scopes associated with them.  On the
1945other hand, although the C++ front end uses pseudo-scopes to
1946handle cleanups for objects with destructors, these don't
1947translate into the GIMPLE form; multiple declarations at the same
1948level use the same @code{BIND_EXPR}.
1949
1950@node Statement Sequences
1951@subsection Statement Sequences
1952@cindex Statement Sequences
1953
1954Multiple statements at the same nesting level are collected into
1955a @code{STATEMENT_LIST}.  Statement lists are modified and
1956traversed using the interface in @samp{tree-iterator.h}.
1957
1958@node Empty Statements
1959@subsection Empty Statements
1960@cindex Empty Statements
1961
1962Whenever possible, statements with no effect are discarded.  But
1963if they are nested within another construct which cannot be
1964discarded for some reason, they are instead replaced with an
1965empty statement, generated by @code{build_empty_stmt}.
1966Initially, all empty statements were shared, after the pattern of
1967the Java front end, but this caused a lot of trouble in practice.
1968
1969An empty statement is represented as @code{(void)0}.
1970
1971@node Jumps
1972@subsection Jumps
1973@cindex Jumps
1974
1975Other jumps are expressed by either @code{GOTO_EXPR} or
1976@code{RETURN_EXPR}.
1977
1978The operand of a @code{GOTO_EXPR} must be either a label or a
1979variable containing the address to jump to.
1980
1981The operand of a @code{RETURN_EXPR} is either @code{NULL_TREE},
1982@code{RESULT_DECL}, or a @code{MODIFY_EXPR} which sets the return
1983value.  It would be nice to move the @code{MODIFY_EXPR} into a
1984separate statement, but the special return semantics in
1985@code{expand_return} make that difficult.  It may still happen in
1986the future, perhaps by moving most of that logic into
1987@code{expand_assignment}.
1988
1989@node Cleanups
1990@subsection Cleanups
1991@cindex Cleanups
1992
1993Destructors for local C++ objects and similar dynamic cleanups are
1994represented in GIMPLE by a @code{TRY_FINALLY_EXPR}.
1995@code{TRY_FINALLY_EXPR} has two operands, both of which are a sequence
1996of statements to execute.  The first sequence is executed.  When it
1997completes the second sequence is executed.
1998
1999The first sequence may complete in the following ways:
2000
2001@enumerate
2002
2003@item Execute the last statement in the sequence and fall off the
2004end.
2005
2006@item Execute a goto statement (@code{GOTO_EXPR}) to an ordinary
2007label outside the sequence.
2008
2009@item Execute a return statement (@code{RETURN_EXPR}).
2010
2011@item Throw an exception.  This is currently not explicitly represented in
2012GIMPLE.
2013
2014@end enumerate
2015
2016The second sequence is not executed if the first sequence completes by
2017calling @code{setjmp} or @code{exit} or any other function that does
2018not return.  The second sequence is also not executed if the first
2019sequence completes via a non-local goto or a computed goto (in general
2020the compiler does not know whether such a goto statement exits the
2021first sequence or not, so we assume that it doesn't).
2022
2023After the second sequence is executed, if it completes normally by
2024falling off the end, execution continues wherever the first sequence
2025would have continued, by falling off the end, or doing a goto, etc.
2026
2027@code{TRY_FINALLY_EXPR} complicates the flow graph, since the cleanup
2028needs to appear on every edge out of the controlled block; this
2029reduces the freedom to move code across these edges.  Therefore, the
2030EH lowering pass which runs before most of the optimization passes
2031eliminates these expressions by explicitly adding the cleanup to each
2032edge.  Rethrowing the exception is represented using @code{RESX_EXPR}.
2033
2034@node OpenMP
2035@subsection OpenMP
2036@tindex OMP_PARALLEL
2037@tindex OMP_FOR
2038@tindex OMP_SECTIONS
2039@tindex OMP_SINGLE
2040@tindex OMP_SECTION
2041@tindex OMP_MASTER
2042@tindex OMP_ORDERED
2043@tindex OMP_CRITICAL
2044@tindex OMP_RETURN
2045@tindex OMP_CONTINUE
2046@tindex OMP_ATOMIC
2047@tindex OMP_CLAUSE
2048
2049All the statements starting with @code{OMP_} represent directives and
2050clauses used by the OpenMP API @w{@uref{http://www.openmp.org/}}.
2051
2052@table @code
2053@item OMP_PARALLEL
2054
2055Represents @code{#pragma omp parallel [clause1 @dots{} clauseN]}. It
2056has four operands:
2057
2058Operand @code{OMP_PARALLEL_BODY} is valid while in GENERIC and
2059High GIMPLE forms.  It contains the body of code to be executed
2060by all the threads.  During GIMPLE lowering, this operand becomes
2061@code{NULL} and the body is emitted linearly after
2062@code{OMP_PARALLEL}.
2063
2064Operand @code{OMP_PARALLEL_CLAUSES} is the list of clauses
2065associated with the directive.
2066
2067Operand @code{OMP_PARALLEL_FN} is created by
2068@code{pass_lower_omp}, it contains the @code{FUNCTION_DECL}
2069for the function that will contain the body of the parallel
2070region.
2071
2072Operand @code{OMP_PARALLEL_DATA_ARG} is also created by
2073@code{pass_lower_omp}. If there are shared variables to be
2074communicated to the children threads, this operand will contain
2075the @code{VAR_DECL} that contains all the shared values and
2076variables.
2077
2078@item OMP_FOR
2079
2080Represents @code{#pragma omp for [clause1 @dots{} clauseN]}.  It
2081has 5 operands:
2082
2083Operand @code{OMP_FOR_BODY} contains the loop body.
2084
2085Operand @code{OMP_FOR_CLAUSES} is the list of clauses
2086associated with the directive.
2087
2088Operand @code{OMP_FOR_INIT} is the loop initialization code of
2089the form @code{VAR = N1}.
2090
2091Operand @code{OMP_FOR_COND} is the loop conditional expression
2092of the form @code{VAR @{<,>,<=,>=@} N2}.
2093
2094Operand @code{OMP_FOR_INCR} is the loop index increment of the
2095form @code{VAR @{+=,-=@} INCR}.
2096
2097Operand @code{OMP_FOR_PRE_BODY} contains side-effect code from
2098operands @code{OMP_FOR_INIT}, @code{OMP_FOR_COND} and
2099@code{OMP_FOR_INC}.  These side-effects are part of the
2100@code{OMP_FOR} block but must be evaluated before the start of
2101loop body.
2102
2103The loop index variable @code{VAR} must be a signed integer variable,
2104which is implicitly private to each thread.  Bounds
2105@code{N1} and @code{N2} and the increment expression
2106@code{INCR} are required to be loop invariant integer
2107expressions that are evaluated without any synchronization. The
2108evaluation order, frequency of evaluation and side-effects are
2109unspecified by the standard.
2110
2111@item OMP_SECTIONS
2112
2113Represents @code{#pragma omp sections [clause1 @dots{} clauseN]}.
2114
2115Operand @code{OMP_SECTIONS_BODY} contains the sections body,
2116which in turn contains a set of @code{OMP_SECTION} nodes for
2117each of the concurrent sections delimited by @code{#pragma omp
2118section}.
2119
2120Operand @code{OMP_SECTIONS_CLAUSES} is the list of clauses
2121associated with the directive.
2122
2123@item OMP_SECTION
2124
2125Section delimiter for @code{OMP_SECTIONS}.
2126
2127@item OMP_SINGLE
2128
2129Represents @code{#pragma omp single}.
2130
2131Operand @code{OMP_SINGLE_BODY} contains the body of code to be
2132executed by a single thread.
2133
2134Operand @code{OMP_SINGLE_CLAUSES} is the list of clauses
2135associated with the directive.
2136
2137@item OMP_MASTER
2138
2139Represents @code{#pragma omp master}.
2140
2141Operand @code{OMP_MASTER_BODY} contains the body of code to be
2142executed by the master thread.
2143
2144@item OMP_ORDERED
2145
2146Represents @code{#pragma omp ordered}.
2147
2148Operand @code{OMP_ORDERED_BODY} contains the body of code to be
2149executed in the sequential order dictated by the loop index
2150variable.
2151
2152@item OMP_CRITICAL
2153
2154Represents @code{#pragma omp critical [name]}.
2155
2156Operand @code{OMP_CRITICAL_BODY} is the critical section.
2157
2158Operand @code{OMP_CRITICAL_NAME} is an optional identifier to
2159label the critical section.
2160
2161@item OMP_RETURN
2162
2163This does not represent any OpenMP directive, it is an artificial
2164marker to indicate the end of the body of an OpenMP@. It is used
2165by the flow graph (@code{tree-cfg.c}) and OpenMP region
2166building code (@code{omp-low.c}).
2167
2168@item OMP_CONTINUE
2169
2170Similarly, this instruction does not represent an OpenMP
2171directive, it is used by @code{OMP_FOR} and
2172@code{OMP_SECTIONS} to mark the place where the code needs to
2173loop to the next iteration (in the case of @code{OMP_FOR}) or
2174the next section (in the case of @code{OMP_SECTIONS}).
2175
2176In some cases, @code{OMP_CONTINUE} is placed right before
2177@code{OMP_RETURN}.  But if there are cleanups that need to
2178occur right after the looping body, it will be emitted between
2179@code{OMP_CONTINUE} and @code{OMP_RETURN}.
2180
2181@item OMP_ATOMIC
2182
2183Represents @code{#pragma omp atomic}.
2184
2185Operand 0 is the address at which the atomic operation is to be
2186performed.
2187
2188Operand 1 is the expression to evaluate.  The gimplifier tries
2189three alternative code generation strategies.  Whenever possible,
2190an atomic update built-in is used.  If that fails, a
2191compare-and-swap loop is attempted.  If that also fails, a
2192regular critical section around the expression is used.
2193
2194@item OMP_CLAUSE
2195
2196Represents clauses associated with one of the @code{OMP_} directives.
2197Clauses are represented by separate sub-codes defined in
2198@file{tree.h}.  Clauses codes can be one of:
2199@code{OMP_CLAUSE_PRIVATE}, @code{OMP_CLAUSE_SHARED},
2200@code{OMP_CLAUSE_FIRSTPRIVATE},
2201@code{OMP_CLAUSE_LASTPRIVATE}, @code{OMP_CLAUSE_COPYIN},
2202@code{OMP_CLAUSE_COPYPRIVATE}, @code{OMP_CLAUSE_IF},
2203@code{OMP_CLAUSE_NUM_THREADS}, @code{OMP_CLAUSE_SCHEDULE},
2204@code{OMP_CLAUSE_NOWAIT}, @code{OMP_CLAUSE_ORDERED},
2205@code{OMP_CLAUSE_DEFAULT}, and @code{OMP_CLAUSE_REDUCTION}.  Each code
2206represents the corresponding OpenMP clause.
2207
2208Clauses associated with the same directive are chained together
2209via @code{OMP_CLAUSE_CHAIN}. Those clauses that accept a list
2210of variables are restricted to exactly one, accessed with
2211@code{OMP_CLAUSE_VAR}.  Therefore, multiple variables under the
2212same clause @code{C} need to be represented as multiple @code{C} clauses
2213chained together.  This facilitates adding new clauses during
2214compilation.
2215
2216@end table
2217
2218@c ---------------------------------------------------------------------
2219@c Functions
2220@c ---------------------------------------------------------------------
2221
2222@node Functions
2223@section Functions
2224@cindex function
2225@tindex FUNCTION_DECL
2226
2227A function is represented by a @code{FUNCTION_DECL} node.  It stores
2228the basic pieces of the function such as body, parameters, and return
2229type as well as information on the surrounding context, visibility,
2230and linkage.
2231
2232@menu
2233* Function Basics::     Function names, body, and parameters.
2234* Function Properties:: Context, linkage, etc.
2235@end menu
2236
2237@c ---------------------------------------------------------------------
2238@c Function Basics
2239@c ---------------------------------------------------------------------
2240
2241@node Function Basics
2242@subsection Function Basics
2243@findex DECL_NAME
2244@findex DECL_ASSEMBLER_NAME
2245@findex TREE_PUBLIC
2246@findex DECL_ARTIFICIAL
2247@findex DECL_FUNCTION_SPECIFIC_TARGET
2248@findex DECL_FUNCTION_SPECIFIC_OPTIMIZATION
2249
2250A function has four core parts: the name, the parameters, the result,
2251and the body.  The following macros and functions access these parts
2252of a @code{FUNCTION_DECL} as well as other basic features:
2253@ftable @code
2254@item DECL_NAME
2255This macro returns the unqualified name of the function, as an
2256@code{IDENTIFIER_NODE}.  For an instantiation of a function template,
2257the @code{DECL_NAME} is the unqualified name of the template, not
2258something like @code{f<int>}.  The value of @code{DECL_NAME} is
2259undefined when used on a constructor, destructor, overloaded operator,
2260or type-conversion operator, or any function that is implicitly
2261generated by the compiler.  See below for macros that can be used to
2262distinguish these cases.
2263
2264@item DECL_ASSEMBLER_NAME
2265This macro returns the mangled name of the function, also an
2266@code{IDENTIFIER_NODE}.  This name does not contain leading underscores
2267on systems that prefix all identifiers with underscores.  The mangled
2268name is computed in the same way on all platforms; if special processing
2269is required to deal with the object file format used on a particular
2270platform, it is the responsibility of the back end to perform those
2271modifications.  (Of course, the back end should not modify
2272@code{DECL_ASSEMBLER_NAME} itself.)
2273
2274Using @code{DECL_ASSEMBLER_NAME} will cause additional memory to be
2275allocated (for the mangled name of the entity) so it should be used
2276only when emitting assembly code.  It should not be used within the
2277optimizers to determine whether or not two declarations are the same,
2278even though some of the existing optimizers do use it in that way.
2279These uses will be removed over time.
2280
2281@item DECL_ARGUMENTS
2282This macro returns the @code{PARM_DECL} for the first argument to the
2283function.  Subsequent @code{PARM_DECL} nodes can be obtained by
2284following the @code{TREE_CHAIN} links.
2285
2286@item DECL_RESULT
2287This macro returns the @code{RESULT_DECL} for the function.
2288
2289@item DECL_SAVED_TREE
2290This macro returns the complete body of the function.
2291
2292@item TREE_TYPE
2293This macro returns the @code{FUNCTION_TYPE} or @code{METHOD_TYPE} for
2294the function.
2295
2296@item DECL_INITIAL
2297A function that has a definition in the current translation unit will
2298have a non-@code{NULL} @code{DECL_INITIAL}.  However, back ends should not make
2299use of the particular value given by @code{DECL_INITIAL}.
2300
2301It should contain a tree of @code{BLOCK} nodes that mirrors the scopes
2302that variables are bound in the function.  Each block contains a list
2303of decls declared in a basic block, a pointer to a chain of blocks at
2304the next lower scope level, then a pointer to the next block at the
2305same level and a backpointer to the parent @code{BLOCK} or
2306@code{FUNCTION_DECL}.  So given a function as follows:
2307
2308@smallexample
2309void foo()
2310@{
2311  int a;
2312  @{
2313    int b;
2314  @}
2315  int c;
2316@}
2317@end smallexample
2318
2319you would get the following:
2320
2321@smallexample
2322tree foo = FUNCTION_DECL;
2323tree decl_a = VAR_DECL;
2324tree decl_b = VAR_DECL;
2325tree decl_c = VAR_DECL;
2326tree block_a = BLOCK;
2327tree block_b = BLOCK;
2328tree block_c = BLOCK;
2329BLOCK_VARS(block_a) = decl_a;
2330BLOCK_SUBBLOCKS(block_a) = block_b;
2331BLOCK_CHAIN(block_a) = block_c;
2332BLOCK_SUPERCONTEXT(block_a) = foo;
2333BLOCK_VARS(block_b) = decl_b;
2334BLOCK_SUPERCONTEXT(block_b) = block_a;
2335BLOCK_VARS(block_c) = decl_c;
2336BLOCK_SUPERCONTEXT(block_c) = foo;
2337DECL_INITIAL(foo) = block_a;
2338@end smallexample
2339
2340@end ftable
2341
2342@c ---------------------------------------------------------------------
2343@c Function Properties
2344@c ---------------------------------------------------------------------
2345
2346@node Function Properties
2347@subsection Function Properties
2348@cindex function properties
2349@cindex statements
2350
2351To determine the scope of a function, you can use the
2352@code{DECL_CONTEXT} macro.  This macro will return the class
2353(either a @code{RECORD_TYPE} or a @code{UNION_TYPE}) or namespace (a
2354@code{NAMESPACE_DECL}) of which the function is a member.  For a virtual
2355function, this macro returns the class in which the function was
2356actually defined, not the base class in which the virtual declaration
2357occurred.
2358
2359In C, the @code{DECL_CONTEXT} for a function maybe another function.
2360This representation indicates that the GNU nested function extension
2361is in use.  For details on the semantics of nested functions, see the
2362GCC Manual.  The nested function can refer to local variables in its
2363containing function.  Such references are not explicitly marked in the
2364tree structure; back ends must look at the @code{DECL_CONTEXT} for the
2365referenced @code{VAR_DECL}.  If the @code{DECL_CONTEXT} for the
2366referenced @code{VAR_DECL} is not the same as the function currently
2367being processed, and neither @code{DECL_EXTERNAL} nor
2368@code{TREE_STATIC} hold, then the reference is to a local variable in
2369a containing function, and the back end must take appropriate action.
2370
2371@ftable @code
2372@item DECL_EXTERNAL
2373This predicate holds if the function is undefined.
2374
2375@item TREE_PUBLIC
2376This predicate holds if the function has external linkage.
2377
2378@item TREE_STATIC
2379This predicate holds if the function has been defined.
2380
2381@item TREE_THIS_VOLATILE
2382This predicate holds if the function does not return normally.
2383
2384@item TREE_READONLY
2385This predicate holds if the function can only read its arguments.
2386
2387@item DECL_PURE_P
2388This predicate holds if the function can only read its arguments, but
2389may also read global memory.
2390
2391@item DECL_VIRTUAL_P
2392This predicate holds if the function is virtual.
2393
2394@item DECL_ARTIFICIAL
2395This macro holds if the function was implicitly generated by the
2396compiler, rather than explicitly declared.  In addition to implicitly
2397generated class member functions, this macro holds for the special
2398functions created to implement static initialization and destruction, to
2399compute run-time type information, and so forth.
2400
2401@item DECL_FUNCTION_SPECIFIC_TARGET
2402This macro returns a tree node that holds the target options that are
2403to be used to compile this particular function or @code{NULL_TREE} if
2404the function is to be compiled with the target options specified on
2405the command line.
2406
2407@item DECL_FUNCTION_SPECIFIC_OPTIMIZATION
2408This macro returns a tree node that holds the optimization options
2409that are to be used to compile this particular function or
2410@code{NULL_TREE} if the function is to be compiled with the
2411optimization options specified on the command line.
2412
2413@end ftable
2414
2415@subsubsection Statements
2416
2417There are tree nodes corresponding to all of the source-level
2418statement constructs, used within the C and C++ frontends.  These are
2419enumerated here, together with a list of the various macros that can
2420be used to obtain information about them.  There are a few macros that
2421can be used with all statements:
2422
2423@c ---------------------------------------------------------------------
2424@c Language-dependent trees
2425@c ---------------------------------------------------------------------
2426
2427@node Language-dependent trees
2428@section Language-dependent trees
2429@cindex language-dependent trees
2430
2431Front ends may wish to keep some state associated with various GENERIC
2432trees while parsing.  To support this, trees provide a set of flags
2433that may be used by the front end.  They are accessed using
2434@code{TREE_LANG_FLAG_n} where @samp{n} is currently 0 through 6.
2435
2436If necessary, a front end can use some language-dependent tree
2437codes in its GENERIC representation, so long as it provides a
2438hook for converting them to GIMPLE and doesn't expect them to
2439work with any (hypothetical) optimizers that run before the
2440conversion to GIMPLE@. The intermediate representation used while
2441parsing C and C++ looks very little like GENERIC, but the C and
2442C++ gimplifier hooks are perfectly happy to take it as input and
2443spit out GIMPLE@.
2444
2445
2446
2447@node C and C++ Trees
2448@section C and C++ Trees
2449
2450This section documents the internal representation used by GCC to
2451represent C and C++ source programs.  When presented with a C or C++
2452source program, GCC parses the program, performs semantic analysis
2453(including the generation of error messages), and then produces the
2454internal representation described here.  This representation contains a
2455complete representation for the entire translation unit provided as
2456input to the front end.  This representation is then typically processed
2457by a code-generator in order to produce machine code, but could also be
2458used in the creation of source browsers, intelligent editors, automatic
2459documentation generators, interpreters, and any other programs needing
2460the ability to process C or C++ code.
2461
2462This section explains the internal representation.  In particular, it
2463documents the internal representation for C and C++ source
2464constructs, and the macros, functions, and variables that can be used to
2465access these constructs.  The C++ representation is largely a superset
2466of the representation used in the C front end.  There is only one
2467construct used in C that does not appear in the C++ front end and that
2468is the GNU ``nested function'' extension.  Many of the macros documented
2469here do not apply in C because the corresponding language constructs do
2470not appear in C@.
2471
2472The C and C++ front ends generate a mix of GENERIC trees and ones
2473specific to C and C++.  These language-specific trees are higher-level
2474constructs than the ones in GENERIC to make the parser's job easier.
2475This section describes those trees that aren't part of GENERIC as well
2476as aspects of GENERIC trees that are treated in a language-specific
2477manner.
2478
2479If you are developing a ``back end'', be it is a code-generator or some
2480other tool, that uses this representation, you may occasionally find
2481that you need to ask questions not easily answered by the functions and
2482macros available here.  If that situation occurs, it is quite likely
2483that GCC already supports the functionality you desire, but that the
2484interface is simply not documented here.  In that case, you should ask
2485the GCC maintainers (via mail to @email{gcc@@gcc.gnu.org}) about
2486documenting the functionality you require.  Similarly, if you find
2487yourself writing functions that do not deal directly with your back end,
2488but instead might be useful to other people using the GCC front end, you
2489should submit your patches for inclusion in GCC@.
2490
2491@menu
2492* Types for C++::               Fundamental and aggregate types.
2493* Namespaces::                  Namespaces.
2494* Classes::                     Classes.
2495* Functions for C++::           Overloading and accessors for C++.
2496* Statements for C++::          Statements specific to C and C++.
2497* C++ Expressions::    From @code{typeid} to @code{throw}.
2498@end menu
2499
2500@node Types for C++
2501@subsection Types for C++
2502@tindex UNKNOWN_TYPE
2503@tindex TYPENAME_TYPE
2504@tindex TYPEOF_TYPE
2505@findex CP_TYPE_QUALS
2506@findex TYPE_UNQUALIFIED
2507@findex TYPE_QUAL_CONST
2508@findex TYPE_QUAL_VOLATILE
2509@findex TYPE_QUAL_RESTRICT
2510@findex TYPE_MAIN_VARIANT
2511@cindex qualified type
2512@findex TYPE_SIZE
2513@findex TYPE_ALIGN
2514@findex TYPE_PRECISION
2515@findex TYPE_ARG_TYPES
2516@findex TYPE_METHOD_BASETYPE
2517@findex TYPE_PTRMEM_P
2518@findex TYPE_OFFSET_BASETYPE
2519@findex TREE_TYPE
2520@findex TYPE_CONTEXT
2521@findex TYPE_NAME
2522@findex TYPENAME_TYPE_FULLNAME
2523@findex TYPE_FIELDS
2524@findex TYPE_PTROBV_P
2525
2526In C++, an array type is not qualified; rather the type of the array
2527elements is qualified.  This situation is reflected in the intermediate
2528representation.  The macros described here will always examine the
2529qualification of the underlying element type when applied to an array
2530type.  (If the element type is itself an array, then the recursion
2531continues until a non-array type is found, and the qualification of this
2532type is examined.)  So, for example, @code{CP_TYPE_CONST_P} will hold of
2533the type @code{const int ()[7]}, denoting an array of seven @code{int}s.
2534
2535The following functions and macros deal with cv-qualification of types:
2536@ftable @code
2537@item CP_TYPE_QUALS
2538This macro returns the set of type qualifiers applied to this type.
2539This value is @code{TYPE_UNQUALIFIED} if no qualifiers have been
2540applied.  The @code{TYPE_QUAL_CONST} bit is set if the type is
2541@code{const}-qualified.  The @code{TYPE_QUAL_VOLATILE} bit is set if the
2542type is @code{volatile}-qualified.  The @code{TYPE_QUAL_RESTRICT} bit is
2543set if the type is @code{restrict}-qualified.
2544
2545@item CP_TYPE_CONST_P
2546This macro holds if the type is @code{const}-qualified.
2547
2548@item CP_TYPE_VOLATILE_P
2549This macro holds if the type is @code{volatile}-qualified.
2550
2551@item CP_TYPE_RESTRICT_P
2552This macro holds if the type is @code{restrict}-qualified.
2553
2554@item CP_TYPE_CONST_NON_VOLATILE_P
2555This predicate holds for a type that is @code{const}-qualified, but
2556@emph{not} @code{volatile}-qualified; other cv-qualifiers are ignored as
2557well: only the @code{const}-ness is tested.
2558
2559@end ftable
2560
2561A few other macros and functions are usable with all types:
2562@ftable @code
2563@item TYPE_SIZE
2564The number of bits required to represent the type, represented as an
2565@code{INTEGER_CST}.  For an incomplete type, @code{TYPE_SIZE} will be
2566@code{NULL_TREE}.
2567
2568@item TYPE_ALIGN
2569The alignment of the type, in bits, represented as an @code{int}.
2570
2571@item TYPE_NAME
2572This macro returns a declaration (in the form of a @code{TYPE_DECL}) for
2573the type.  (Note this macro does @emph{not} return an
2574@code{IDENTIFIER_NODE}, as you might expect, given its name!)  You can
2575look at the @code{DECL_NAME} of the @code{TYPE_DECL} to obtain the
2576actual name of the type.  The @code{TYPE_NAME} will be @code{NULL_TREE}
2577for a type that is not a built-in type, the result of a typedef, or a
2578named class type.
2579
2580@item CP_INTEGRAL_TYPE
2581This predicate holds if the type is an integral type.  Notice that in
2582C++, enumerations are @emph{not} integral types.
2583
2584@item ARITHMETIC_TYPE_P
2585This predicate holds if the type is an integral type (in the C++ sense)
2586or a floating point type.
2587
2588@item CLASS_TYPE_P
2589This predicate holds for a class-type.
2590
2591@item TYPE_BUILT_IN
2592This predicate holds for a built-in type.
2593
2594@item TYPE_PTRMEM_P
2595This predicate holds if the type is a pointer to data member.
2596
2597@item TYPE_PTR_P
2598This predicate holds if the type is a pointer type, and the pointee is
2599not a data member.
2600
2601@item TYPE_PTRFN_P
2602This predicate holds for a pointer to function type.
2603
2604@item TYPE_PTROB_P
2605This predicate holds for a pointer to object type.  Note however that it
2606does not hold for the generic pointer to object type @code{void *}.  You
2607may use @code{TYPE_PTROBV_P} to test for a pointer to object type as
2608well as @code{void *}.
2609
2610@end ftable
2611
2612The table below describes types specific to C and C++ as well as
2613language-dependent info about GENERIC types.
2614
2615@table @code
2616
2617@item POINTER_TYPE
2618Used to represent pointer types, and pointer to data member types.  If
2619@code{TREE_TYPE}
2620is a pointer to data member type, then @code{TYPE_PTRMEM_P} will hold.
2621For a pointer to data member type of the form @samp{T X::*},
2622@code{TYPE_PTRMEM_CLASS_TYPE} will be the type @code{X}, while
2623@code{TYPE_PTRMEM_POINTED_TO_TYPE} will be the type @code{T}.
2624
2625@item RECORD_TYPE
2626Used to represent @code{struct} and @code{class} types in C and C++.  If
2627@code{TYPE_PTRMEMFUNC_P} holds, then this type is a pointer-to-member
2628type.  In that case, the @code{TYPE_PTRMEMFUNC_FN_TYPE} is a
2629@code{POINTER_TYPE} pointing to a @code{METHOD_TYPE}.  The
2630@code{METHOD_TYPE} is the type of a function pointed to by the
2631pointer-to-member function.  If @code{TYPE_PTRMEMFUNC_P} does not hold,
2632this type is a class type.  For more information, see @pxref{Classes}.
2633
2634@item UNKNOWN_TYPE
2635This node is used to represent a type the knowledge of which is
2636insufficient for a sound processing.
2637
2638@item TYPENAME_TYPE
2639Used to represent a construct of the form @code{typename T::A}.  The
2640@code{TYPE_CONTEXT} is @code{T}; the @code{TYPE_NAME} is an
2641@code{IDENTIFIER_NODE} for @code{A}.  If the type is specified via a
2642template-id, then @code{TYPENAME_TYPE_FULLNAME} yields a
2643@code{TEMPLATE_ID_EXPR}.  The @code{TREE_TYPE} is non-@code{NULL} if the
2644node is implicitly generated in support for the implicit typename
2645extension; in which case the @code{TREE_TYPE} is a type node for the
2646base-class.
2647
2648@item TYPEOF_TYPE
2649Used to represent the @code{__typeof__} extension.  The
2650@code{TYPE_FIELDS} is the expression the type of which is being
2651represented.
2652
2653@end table
2654
2655
2656@c ---------------------------------------------------------------------
2657@c Namespaces
2658@c ---------------------------------------------------------------------
2659
2660@node Namespaces
2661@subsection Namespaces
2662@cindex namespace, scope
2663@tindex NAMESPACE_DECL
2664
2665The root of the entire intermediate representation is the variable
2666@code{global_namespace}.  This is the namespace specified with @code{::}
2667in C++ source code.  All other namespaces, types, variables, functions,
2668and so forth can be found starting with this namespace.
2669
2670However, except for the fact that it is distinguished as the root of the
2671representation, the global namespace is no different from any other
2672namespace.  Thus, in what follows, we describe namespaces generally,
2673rather than the global namespace in particular.
2674
2675A namespace is represented by a @code{NAMESPACE_DECL} node.
2676
2677The following macros and functions can be used on a @code{NAMESPACE_DECL}:
2678
2679@ftable @code
2680@item DECL_NAME
2681This macro is used to obtain the @code{IDENTIFIER_NODE} corresponding to
2682the unqualified name of the name of the namespace (@pxref{Identifiers}).
2683The name of the global namespace is @samp{::}, even though in C++ the
2684global namespace is unnamed.  However, you should use comparison with
2685@code{global_namespace}, rather than @code{DECL_NAME} to determine
2686whether or not a namespace is the global one.  An unnamed namespace
2687will have a @code{DECL_NAME} equal to @code{anonymous_namespace_name}.
2688Within a single translation unit, all unnamed namespaces will have the
2689same name.
2690
2691@item DECL_CONTEXT
2692This macro returns the enclosing namespace.  The @code{DECL_CONTEXT} for
2693the @code{global_namespace} is @code{NULL_TREE}.
2694
2695@item DECL_NAMESPACE_ALIAS
2696If this declaration is for a namespace alias, then
2697@code{DECL_NAMESPACE_ALIAS} is the namespace for which this one is an
2698alias.
2699
2700Do not attempt to use @code{cp_namespace_decls} for a namespace which is
2701an alias.  Instead, follow @code{DECL_NAMESPACE_ALIAS} links until you
2702reach an ordinary, non-alias, namespace, and call
2703@code{cp_namespace_decls} there.
2704
2705@item DECL_NAMESPACE_STD_P
2706This predicate holds if the namespace is the special @code{::std}
2707namespace.
2708
2709@item cp_namespace_decls
2710This function will return the declarations contained in the namespace,
2711including types, overloaded functions, other namespaces, and so forth.
2712If there are no declarations, this function will return
2713@code{NULL_TREE}.  The declarations are connected through their
2714@code{TREE_CHAIN} fields.
2715
2716Although most entries on this list will be declarations,
2717@code{TREE_LIST} nodes may also appear.  In this case, the
2718@code{TREE_VALUE} will be an @code{OVERLOAD}.  The value of the
2719@code{TREE_PURPOSE} is unspecified; back ends should ignore this value.
2720As with the other kinds of declarations returned by
2721@code{cp_namespace_decls}, the @code{TREE_CHAIN} will point to the next
2722declaration in this list.
2723
2724For more information on the kinds of declarations that can occur on this
2725list, @xref{Declarations}.  Some declarations will not appear on this
2726list.  In particular, no @code{FIELD_DECL}, @code{LABEL_DECL}, or
2727@code{PARM_DECL} nodes will appear here.
2728
2729This function cannot be used with namespaces that have
2730@code{DECL_NAMESPACE_ALIAS} set.
2731
2732@end ftable
2733
2734@c ---------------------------------------------------------------------
2735@c Classes
2736@c ---------------------------------------------------------------------
2737
2738@node Classes
2739@subsection Classes
2740@cindex class, scope
2741@tindex RECORD_TYPE
2742@tindex UNION_TYPE
2743@findex CLASSTYPE_DECLARED_CLASS
2744@findex TYPE_BINFO
2745@findex BINFO_TYPE
2746@findex TYPE_FIELDS
2747@findex TYPE_VFIELD
2748@findex TYPE_METHODS
2749
2750Besides namespaces, the other high-level scoping construct in C++ is the
2751class.  (Throughout this manual the term @dfn{class} is used to mean the
2752types referred to in the ANSI/ISO C++ Standard as classes; these include
2753types defined with the @code{class}, @code{struct}, and @code{union}
2754keywords.)
2755
2756A class type is represented by either a @code{RECORD_TYPE} or a
2757@code{UNION_TYPE}.  A class declared with the @code{union} tag is
2758represented by a @code{UNION_TYPE}, while classes declared with either
2759the @code{struct} or the @code{class} tag are represented by
2760@code{RECORD_TYPE}s.  You can use the @code{CLASSTYPE_DECLARED_CLASS}
2761macro to discern whether or not a particular type is a @code{class} as
2762opposed to a @code{struct}.  This macro will be true only for classes
2763declared with the @code{class} tag.
2764
2765Almost all non-function members are available on the @code{TYPE_FIELDS}
2766list.  Given one member, the next can be found by following the
2767@code{TREE_CHAIN}.  You should not depend in any way on the order in
2768which fields appear on this list.  All nodes on this list will be
2769@samp{DECL} nodes.  A @code{FIELD_DECL} is used to represent a non-static
2770data member, a @code{VAR_DECL} is used to represent a static data
2771member, and a @code{TYPE_DECL} is used to represent a type.  Note that
2772the @code{CONST_DECL} for an enumeration constant will appear on this
2773list, if the enumeration type was declared in the class.  (Of course,
2774the @code{TYPE_DECL} for the enumeration type will appear here as well.)
2775There are no entries for base classes on this list.  In particular,
2776there is no @code{FIELD_DECL} for the ``base-class portion'' of an
2777object.
2778
2779The @code{TYPE_VFIELD} is a compiler-generated field used to point to
2780virtual function tables.  It may or may not appear on the
2781@code{TYPE_FIELDS} list.  However, back ends should handle the
2782@code{TYPE_VFIELD} just like all the entries on the @code{TYPE_FIELDS}
2783list.
2784
2785The function members are available on the @code{TYPE_METHODS} list.
2786Again, subsequent members are found by following the @code{TREE_CHAIN}
2787field.  If a function is overloaded, each of the overloaded functions
2788appears; no @code{OVERLOAD} nodes appear on the @code{TYPE_METHODS}
2789list.  Implicitly declared functions (including default constructors,
2790copy constructors, assignment operators, and destructors) will appear on
2791this list as well.
2792
2793Every class has an associated @dfn{binfo}, which can be obtained with
2794@code{TYPE_BINFO}.  Binfos are used to represent base-classes.  The
2795binfo given by @code{TYPE_BINFO} is the degenerate case, whereby every
2796class is considered to be its own base-class.  The base binfos for a
2797particular binfo are held in a vector, whose length is obtained with
2798@code{BINFO_N_BASE_BINFOS}.  The base binfos themselves are obtained
2799with @code{BINFO_BASE_BINFO} and @code{BINFO_BASE_ITERATE}.  To add a
2800new binfo, use @code{BINFO_BASE_APPEND}.  The vector of base binfos can
2801be obtained with @code{BINFO_BASE_BINFOS}, but normally you do not need
2802to use that.  The class type associated with a binfo is given by
2803@code{BINFO_TYPE}.  It is not always the case that @code{BINFO_TYPE
2804(TYPE_BINFO (x))}, because of typedefs and qualified types.  Neither is
2805it the case that @code{TYPE_BINFO (BINFO_TYPE (y))} is the same binfo as
2806@code{y}.  The reason is that if @code{y} is a binfo representing a
2807base-class @code{B} of a derived class @code{D}, then @code{BINFO_TYPE
2808(y)} will be @code{B}, and @code{TYPE_BINFO (BINFO_TYPE (y))} will be
2809@code{B} as its own base-class, rather than as a base-class of @code{D}.
2810
2811The access to a base type can be found with @code{BINFO_BASE_ACCESS}.
2812This will produce @code{access_public_node}, @code{access_private_node}
2813or @code{access_protected_node}.  If bases are always public,
2814@code{BINFO_BASE_ACCESSES} may be @code{NULL}.
2815
2816@code{BINFO_VIRTUAL_P} is used to specify whether the binfo is inherited
2817virtually or not.  The other flags, @code{BINFO_MARKED_P} and
2818@code{BINFO_FLAG_1} to @code{BINFO_FLAG_6} can be used for language
2819specific use.
2820
2821The following macros can be used on a tree node representing a class-type.
2822
2823@ftable @code
2824@item LOCAL_CLASS_P
2825This predicate holds if the class is local class @emph{i.e.}@: declared
2826inside a function body.
2827
2828@item TYPE_POLYMORPHIC_P
2829This predicate holds if the class has at least one virtual function
2830(declared or inherited).
2831
2832@item TYPE_HAS_DEFAULT_CONSTRUCTOR
2833This predicate holds whenever its argument represents a class-type with
2834default constructor.
2835
2836@item CLASSTYPE_HAS_MUTABLE
2837@itemx TYPE_HAS_MUTABLE_P
2838These predicates hold for a class-type having a mutable data member.
2839
2840@item CLASSTYPE_NON_POD_P
2841This predicate holds only for class-types that are not PODs.
2842
2843@item TYPE_HAS_NEW_OPERATOR
2844This predicate holds for a class-type that defines
2845@code{operator new}.
2846
2847@item TYPE_HAS_ARRAY_NEW_OPERATOR
2848This predicate holds for a class-type for which
2849@code{operator new[]} is defined.
2850
2851@item TYPE_OVERLOADS_CALL_EXPR
2852This predicate holds for class-type for which the function call
2853@code{operator()} is overloaded.
2854
2855@item TYPE_OVERLOADS_ARRAY_REF
2856This predicate holds for a class-type that overloads
2857@code{operator[]}
2858
2859@item TYPE_OVERLOADS_ARROW
2860This predicate holds for a class-type for which @code{operator->} is
2861overloaded.
2862
2863@end ftable
2864
2865@node Functions for C++
2866@subsection Functions for C++
2867@cindex function
2868@tindex FUNCTION_DECL
2869@tindex OVERLOAD
2870@findex OVL_CURRENT
2871@findex OVL_NEXT
2872
2873A function is represented by a @code{FUNCTION_DECL} node.  A set of
2874overloaded functions is sometimes represented by an @code{OVERLOAD} node.
2875
2876An @code{OVERLOAD} node is not a declaration, so none of the
2877@samp{DECL_} macros should be used on an @code{OVERLOAD}.  An
2878@code{OVERLOAD} node is similar to a @code{TREE_LIST}.  Use
2879@code{OVL_CURRENT} to get the function associated with an
2880@code{OVERLOAD} node; use @code{OVL_NEXT} to get the next
2881@code{OVERLOAD} node in the list of overloaded functions.  The macros
2882@code{OVL_CURRENT} and @code{OVL_NEXT} are actually polymorphic; you can
2883use them to work with @code{FUNCTION_DECL} nodes as well as with
2884overloads.  In the case of a @code{FUNCTION_DECL}, @code{OVL_CURRENT}
2885will always return the function itself, and @code{OVL_NEXT} will always
2886be @code{NULL_TREE}.
2887
2888To determine the scope of a function, you can use the
2889@code{DECL_CONTEXT} macro.  This macro will return the class
2890(either a @code{RECORD_TYPE} or a @code{UNION_TYPE}) or namespace (a
2891@code{NAMESPACE_DECL}) of which the function is a member.  For a virtual
2892function, this macro returns the class in which the function was
2893actually defined, not the base class in which the virtual declaration
2894occurred.
2895
2896If a friend function is defined in a class scope, the
2897@code{DECL_FRIEND_CONTEXT} macro can be used to determine the class in
2898which it was defined.  For example, in
2899@smallexample
2900class C @{ friend void f() @{@} @};
2901@end smallexample
2902@noindent
2903the @code{DECL_CONTEXT} for @code{f} will be the
2904@code{global_namespace}, but the @code{DECL_FRIEND_CONTEXT} will be the
2905@code{RECORD_TYPE} for @code{C}.
2906
2907
2908The following macros and functions can be used on a @code{FUNCTION_DECL}:
2909@ftable @code
2910@item DECL_MAIN_P
2911This predicate holds for a function that is the program entry point
2912@code{::code}.
2913
2914@item DECL_LOCAL_FUNCTION_P
2915This predicate holds if the function was declared at block scope, even
2916though it has a global scope.
2917
2918@item DECL_ANTICIPATED
2919This predicate holds if the function is a built-in function but its
2920prototype is not yet explicitly declared.
2921
2922@item DECL_EXTERN_C_FUNCTION_P
2923This predicate holds if the function is declared as an
2924`@code{extern "C"}' function.
2925
2926@item DECL_LINKONCE_P
2927This macro holds if multiple copies of this function may be emitted in
2928various translation units.  It is the responsibility of the linker to
2929merge the various copies.  Template instantiations are the most common
2930example of functions for which @code{DECL_LINKONCE_P} holds; G++
2931instantiates needed templates in all translation units which require them,
2932and then relies on the linker to remove duplicate instantiations.
2933
2934FIXME: This macro is not yet implemented.
2935
2936@item DECL_FUNCTION_MEMBER_P
2937This macro holds if the function is a member of a class, rather than a
2938member of a namespace.
2939
2940@item DECL_STATIC_FUNCTION_P
2941This predicate holds if the function a static member function.
2942
2943@item DECL_NONSTATIC_MEMBER_FUNCTION_P
2944This macro holds for a non-static member function.
2945
2946@item DECL_CONST_MEMFUNC_P
2947This predicate holds for a @code{const}-member function.
2948
2949@item DECL_VOLATILE_MEMFUNC_P
2950This predicate holds for a @code{volatile}-member function.
2951
2952@item DECL_CONSTRUCTOR_P
2953This macro holds if the function is a constructor.
2954
2955@item DECL_NONCONVERTING_P
2956This predicate holds if the constructor is a non-converting constructor.
2957
2958@item DECL_COMPLETE_CONSTRUCTOR_P
2959This predicate holds for a function which is a constructor for an object
2960of a complete type.
2961
2962@item DECL_BASE_CONSTRUCTOR_P
2963This predicate holds for a function which is a constructor for a base
2964class sub-object.
2965
2966@item DECL_COPY_CONSTRUCTOR_P
2967This predicate holds for a function which is a copy-constructor.
2968
2969@item DECL_DESTRUCTOR_P
2970This macro holds if the function is a destructor.
2971
2972@item DECL_COMPLETE_DESTRUCTOR_P
2973This predicate holds if the function is the destructor for an object a
2974complete type.
2975
2976@item DECL_OVERLOADED_OPERATOR_P
2977This macro holds if the function is an overloaded operator.
2978
2979@item DECL_CONV_FN_P
2980This macro holds if the function is a type-conversion operator.
2981
2982@item DECL_GLOBAL_CTOR_P
2983This predicate holds if the function is a file-scope initialization
2984function.
2985
2986@item DECL_GLOBAL_DTOR_P
2987This predicate holds if the function is a file-scope finalization
2988function.
2989
2990@item DECL_THUNK_P
2991This predicate holds if the function is a thunk.
2992
2993These functions represent stub code that adjusts the @code{this} pointer
2994and then jumps to another function.  When the jumped-to function
2995returns, control is transferred directly to the caller, without
2996returning to the thunk.  The first parameter to the thunk is always the
2997@code{this} pointer; the thunk should add @code{THUNK_DELTA} to this
2998value.  (The @code{THUNK_DELTA} is an @code{int}, not an
2999@code{INTEGER_CST}.)
3000
3001Then, if @code{THUNK_VCALL_OFFSET} (an @code{INTEGER_CST}) is nonzero
3002the adjusted @code{this} pointer must be adjusted again.  The complete
3003calculation is given by the following pseudo-code:
3004
3005@smallexample
3006this += THUNK_DELTA
3007if (THUNK_VCALL_OFFSET)
3008  this += (*((ptrdiff_t **) this))[THUNK_VCALL_OFFSET]
3009@end smallexample
3010
3011Finally, the thunk should jump to the location given
3012by @code{DECL_INITIAL}; this will always be an expression for the
3013address of a function.
3014
3015@item DECL_NON_THUNK_FUNCTION_P
3016This predicate holds if the function is @emph{not} a thunk function.
3017
3018@item GLOBAL_INIT_PRIORITY
3019If either @code{DECL_GLOBAL_CTOR_P} or @code{DECL_GLOBAL_DTOR_P} holds,
3020then this gives the initialization priority for the function.  The
3021linker will arrange that all functions for which
3022@code{DECL_GLOBAL_CTOR_P} holds are run in increasing order of priority
3023before @code{main} is called.  When the program exits, all functions for
3024which @code{DECL_GLOBAL_DTOR_P} holds are run in the reverse order.
3025
3026@item TYPE_RAISES_EXCEPTIONS
3027This macro returns the list of exceptions that a (member-)function can
3028raise.  The returned list, if non @code{NULL}, is comprised of nodes
3029whose @code{TREE_VALUE} represents a type.
3030
3031@item TYPE_NOTHROW_P
3032This predicate holds when the exception-specification of its arguments
3033is of the form `@code{()}'.
3034
3035@item DECL_ARRAY_DELETE_OPERATOR_P
3036This predicate holds if the function an overloaded
3037@code{operator delete[]}.
3038
3039@end ftable
3040
3041@c ---------------------------------------------------------------------
3042@c Function Bodies
3043@c ---------------------------------------------------------------------
3044
3045@node Statements for C++
3046@subsection Statements for C++
3047@cindex statements
3048@tindex BREAK_STMT
3049@tindex CLEANUP_STMT
3050@findex CLEANUP_DECL
3051@findex CLEANUP_EXPR
3052@tindex CONTINUE_STMT
3053@tindex DECL_STMT
3054@findex DECL_STMT_DECL
3055@tindex DO_STMT
3056@findex DO_BODY
3057@findex DO_COND
3058@tindex EMPTY_CLASS_EXPR
3059@tindex EXPR_STMT
3060@findex EXPR_STMT_EXPR
3061@tindex FOR_STMT
3062@findex FOR_INIT_STMT
3063@findex FOR_COND
3064@findex FOR_EXPR
3065@findex FOR_BODY
3066@tindex HANDLER
3067@tindex IF_STMT
3068@findex IF_COND
3069@findex THEN_CLAUSE
3070@findex ELSE_CLAUSE
3071@tindex RETURN_STMT
3072@findex RETURN_EXPR
3073@tindex SUBOBJECT
3074@findex SUBOBJECT_CLEANUP
3075@tindex SWITCH_STMT
3076@findex SWITCH_COND
3077@findex SWITCH_BODY
3078@tindex TRY_BLOCK
3079@findex TRY_STMTS
3080@findex TRY_HANDLERS
3081@findex HANDLER_PARMS
3082@findex HANDLER_BODY
3083@findex USING_STMT
3084@tindex WHILE_STMT
3085@findex WHILE_BODY
3086@findex WHILE_COND
3087
3088A function that has a definition in the current translation unit will
3089have a non-@code{NULL} @code{DECL_INITIAL}.  However, back ends should not make
3090use of the particular value given by @code{DECL_INITIAL}.
3091
3092The @code{DECL_SAVED_TREE} macro will give the complete body of the
3093function.
3094
3095@subsubsection Statements
3096
3097There are tree nodes corresponding to all of the source-level
3098statement constructs, used within the C and C++ frontends.  These are
3099enumerated here, together with a list of the various macros that can
3100be used to obtain information about them.  There are a few macros that
3101can be used with all statements:
3102
3103@ftable @code
3104@item STMT_IS_FULL_EXPR_P
3105In C++, statements normally constitute ``full expressions''; temporaries
3106created during a statement are destroyed when the statement is complete.
3107However, G++ sometimes represents expressions by statements; these
3108statements will not have @code{STMT_IS_FULL_EXPR_P} set.  Temporaries
3109created during such statements should be destroyed when the innermost
3110enclosing statement with @code{STMT_IS_FULL_EXPR_P} set is exited.
3111
3112@end ftable
3113
3114Here is the list of the various statement nodes, and the macros used to
3115access them.  This documentation describes the use of these nodes in
3116non-template functions (including instantiations of template functions).
3117In template functions, the same nodes are used, but sometimes in
3118slightly different ways.
3119
3120Many of the statements have substatements.  For example, a @code{while}
3121loop will have a body, which is itself a statement.  If the substatement
3122is @code{NULL_TREE}, it is considered equivalent to a statement
3123consisting of a single @code{;}, i.e., an expression statement in which
3124the expression has been omitted.  A substatement may in fact be a list
3125of statements, connected via their @code{TREE_CHAIN}s.  So, you should
3126always process the statement tree by looping over substatements, like
3127this:
3128@smallexample
3129void process_stmt (stmt)
3130     tree stmt;
3131@{
3132  while (stmt)
3133    @{
3134      switch (TREE_CODE (stmt))
3135        @{
3136        case IF_STMT:
3137          process_stmt (THEN_CLAUSE (stmt));
3138          /* @r{More processing here.}  */
3139          break;
3140
3141        @dots{}
3142        @}
3143
3144      stmt = TREE_CHAIN (stmt);
3145    @}
3146@}
3147@end smallexample
3148In other words, while the @code{then} clause of an @code{if} statement
3149in C++ can be only one statement (although that one statement may be a
3150compound statement), the intermediate representation will sometimes use
3151several statements chained together.
3152
3153@table @code
3154@item BREAK_STMT
3155
3156Used to represent a @code{break} statement.  There are no additional
3157fields.
3158
3159@item CLEANUP_STMT
3160
3161Used to represent an action that should take place upon exit from the
3162enclosing scope.  Typically, these actions are calls to destructors for
3163local objects, but back ends cannot rely on this fact.  If these nodes
3164are in fact representing such destructors, @code{CLEANUP_DECL} will be
3165the @code{VAR_DECL} destroyed.  Otherwise, @code{CLEANUP_DECL} will be
3166@code{NULL_TREE}.  In any case, the @code{CLEANUP_EXPR} is the
3167expression to execute.  The cleanups executed on exit from a scope
3168should be run in the reverse order of the order in which the associated
3169@code{CLEANUP_STMT}s were encountered.
3170
3171@item CONTINUE_STMT
3172
3173Used to represent a @code{continue} statement.  There are no additional
3174fields.
3175
3176@item CTOR_STMT
3177
3178Used to mark the beginning (if @code{CTOR_BEGIN_P} holds) or end (if
3179@code{CTOR_END_P} holds of the main body of a constructor.  See also
3180@code{SUBOBJECT} for more information on how to use these nodes.
3181
3182@item DO_STMT
3183
3184Used to represent a @code{do} loop.  The body of the loop is given by
3185@code{DO_BODY} while the termination condition for the loop is given by
3186@code{DO_COND}.  The condition for a @code{do}-statement is always an
3187expression.
3188
3189@item EMPTY_CLASS_EXPR
3190
3191Used to represent a temporary object of a class with no data whose
3192address is never taken.  (All such objects are interchangeable.)  The
3193@code{TREE_TYPE} represents the type of the object.
3194
3195@item EXPR_STMT
3196
3197Used to represent an expression statement.  Use @code{EXPR_STMT_EXPR} to
3198obtain the expression.
3199
3200@item FOR_STMT
3201
3202Used to represent a @code{for} statement.  The @code{FOR_INIT_STMT} is
3203the initialization statement for the loop.  The @code{FOR_COND} is the
3204termination condition.  The @code{FOR_EXPR} is the expression executed
3205right before the @code{FOR_COND} on each loop iteration; often, this
3206expression increments a counter.  The body of the loop is given by
3207@code{FOR_BODY}.  Note that @code{FOR_INIT_STMT} and @code{FOR_BODY}
3208return statements, while @code{FOR_COND} and @code{FOR_EXPR} return
3209expressions.
3210
3211@item HANDLER
3212
3213Used to represent a C++ @code{catch} block.  The @code{HANDLER_TYPE}
3214is the type of exception that will be caught by this handler; it is
3215equal (by pointer equality) to @code{NULL} if this handler is for all
3216types.  @code{HANDLER_PARMS} is the @code{DECL_STMT} for the catch
3217parameter, and @code{HANDLER_BODY} is the code for the block itself.
3218
3219@item IF_STMT
3220
3221Used to represent an @code{if} statement.  The @code{IF_COND} is the
3222expression.
3223
3224If the condition is a @code{TREE_LIST}, then the @code{TREE_PURPOSE} is
3225a statement (usually a @code{DECL_STMT}).  Each time the condition is
3226evaluated, the statement should be executed.  Then, the
3227@code{TREE_VALUE} should be used as the conditional expression itself.
3228This representation is used to handle C++ code like this:
3229
3230C++ distinguishes between this and @code{COND_EXPR} for handling templates.
3231
3232@smallexample
3233if (int i = 7) @dots{}
3234@end smallexample
3235
3236where there is a new local variable (or variables) declared within the
3237condition.
3238
3239The @code{THEN_CLAUSE} represents the statement given by the @code{then}
3240condition, while the @code{ELSE_CLAUSE} represents the statement given
3241by the @code{else} condition.
3242
3243@item SUBOBJECT
3244
3245In a constructor, these nodes are used to mark the point at which a
3246subobject of @code{this} is fully constructed.  If, after this point, an
3247exception is thrown before a @code{CTOR_STMT} with @code{CTOR_END_P} set
3248is encountered, the @code{SUBOBJECT_CLEANUP} must be executed.  The
3249cleanups must be executed in the reverse order in which they appear.
3250
3251@item SWITCH_STMT
3252
3253Used to represent a @code{switch} statement.  The @code{SWITCH_STMT_COND}
3254is the expression on which the switch is occurring.  See the documentation
3255for an @code{IF_STMT} for more information on the representation used
3256for the condition.  The @code{SWITCH_STMT_BODY} is the body of the switch
3257statement.   The @code{SWITCH_STMT_TYPE} is the original type of switch
3258expression as given in the source, before any compiler conversions.
3259
3260@item TRY_BLOCK
3261Used to represent a @code{try} block.  The body of the try block is
3262given by @code{TRY_STMTS}.  Each of the catch blocks is a @code{HANDLER}
3263node.  The first handler is given by @code{TRY_HANDLERS}.  Subsequent
3264handlers are obtained by following the @code{TREE_CHAIN} link from one
3265handler to the next.  The body of the handler is given by
3266@code{HANDLER_BODY}.
3267
3268If @code{CLEANUP_P} holds of the @code{TRY_BLOCK}, then the
3269@code{TRY_HANDLERS} will not be a @code{HANDLER} node.  Instead, it will
3270be an expression that should be executed if an exception is thrown in
3271the try block.  It must rethrow the exception after executing that code.
3272And, if an exception is thrown while the expression is executing,
3273@code{terminate} must be called.
3274
3275@item USING_STMT
3276Used to represent a @code{using} directive.  The namespace is given by
3277@code{USING_STMT_NAMESPACE}, which will be a NAMESPACE_DECL@.  This node
3278is needed inside template functions, to implement using directives
3279during instantiation.
3280
3281@item WHILE_STMT
3282
3283Used to represent a @code{while} loop.  The @code{WHILE_COND} is the
3284termination condition for the loop.  See the documentation for an
3285@code{IF_STMT} for more information on the representation used for the
3286condition.
3287
3288The @code{WHILE_BODY} is the body of the loop.
3289
3290@end table
3291
3292@node C++ Expressions
3293@subsection C++ Expressions
3294
3295This section describes expressions specific to the C and C++ front
3296ends.
3297
3298@table @code
3299@item TYPEID_EXPR
3300
3301Used to represent a @code{typeid} expression.
3302
3303@item NEW_EXPR
3304@itemx VEC_NEW_EXPR
3305
3306Used to represent a call to @code{new} and @code{new[]} respectively.
3307
3308@item DELETE_EXPR
3309@itemx VEC_DELETE_EXPR
3310
3311Used to represent a call to @code{delete} and @code{delete[]} respectively.
3312
3313@item MEMBER_REF
3314
3315Represents a reference to a member of a class.
3316
3317@item THROW_EXPR
3318
3319Represents an instance of @code{throw} in the program.  Operand 0,
3320which is the expression to throw, may be @code{NULL_TREE}.
3321
3322
3323@item AGGR_INIT_EXPR
3324An @code{AGGR_INIT_EXPR} represents the initialization as the return
3325value of a function call, or as the result of a constructor.  An
3326@code{AGGR_INIT_EXPR} will only appear as a full-expression, or as the
3327second operand of a @code{TARGET_EXPR}.  @code{AGGR_INIT_EXPR}s have
3328a representation similar to that of @code{CALL_EXPR}s.  You can use
3329the @code{AGGR_INIT_EXPR_FN} and @code{AGGR_INIT_EXPR_ARG} macros to access
3330the function to call and the arguments to pass.
3331
3332If @code{AGGR_INIT_VIA_CTOR_P} holds of the @code{AGGR_INIT_EXPR}, then
3333the initialization is via a constructor call.  The address of the
3334@code{AGGR_INIT_EXPR_SLOT} operand, which is always a @code{VAR_DECL},
3335is taken, and this value replaces the first argument in the argument
3336list.
3337
3338In either case, the expression is void.
3339
3340
3341@end table
3342
3343
3344@node Java Trees
3345@section Java Trees
3346