xref: /netbsd-src/external/gpl3/gcc.old/dist/gcc/doc/generic.texi (revision 212397c69a103ae7e5eafa8731ddfae671d2dee7)
1@c Copyright (C) 2004-2013 Free Software Foundation, Inc.
2@c This is part of the GCC manual.
3@c For copying conditions, see the file gcc.texi.
4
5@c ---------------------------------------------------------------------
6@c GENERIC
7@c ---------------------------------------------------------------------
8
9@node GENERIC
10@chapter GENERIC
11@cindex GENERIC
12
13The purpose of GENERIC is simply to provide a
14language-independent way of representing an entire function in
15trees.  To this end, it was necessary to add a few new tree codes
16to the back end, but most everything was already there.  If you
17can express it with the codes in @code{gcc/tree.def}, it's
18GENERIC@.
19
20Early on, there was a great deal of debate about how to think
21about statements in a tree IL@.  In GENERIC, a statement is
22defined as any expression whose value, if any, is ignored.  A
23statement will always have @code{TREE_SIDE_EFFECTS} set (or it
24will be discarded), but a non-statement expression may also have
25side effects.  A @code{CALL_EXPR}, for instance.
26
27It would be possible for some local optimizations to work on the
28GENERIC form of a function; indeed, the adapted tree inliner
29works fine on GENERIC, but the current compiler performs inlining
30after lowering to GIMPLE (a restricted form described in the next
31section). Indeed, currently the frontends perform this lowering
32before handing off to @code{tree_rest_of_compilation}, but this
33seems inelegant.
34
35@menu
36* Deficiencies::                Topics net yet covered in this document.
37* Tree overview::               All about @code{tree}s.
38* Types::                       Fundamental and aggregate types.
39* Declarations::                Type declarations and variables.
40* Attributes::                  Declaration and type attributes.
41* Expressions: Expression trees.            Operating on data.
42* Statements::                  Control flow and related trees.
43* Functions::           	Function bodies, linkage, and other aspects.
44* Language-dependent trees::    Topics and trees specific to language front ends.
45* C and C++ Trees::     	Trees specific to C and C++.
46* Java Trees:: 	                Trees specific to Java.
47@end menu
48
49@c ---------------------------------------------------------------------
50@c Deficiencies
51@c ---------------------------------------------------------------------
52
53@node Deficiencies
54@section Deficiencies
55
56There are many places in which this document is incomplet and incorrekt.
57It is, as of yet, only @emph{preliminary} documentation.
58
59@c ---------------------------------------------------------------------
60@c Overview
61@c ---------------------------------------------------------------------
62
63@node Tree overview
64@section Overview
65@cindex tree
66@findex TREE_CODE
67
68The central data structure used by the internal representation is the
69@code{tree}.  These nodes, while all of the C type @code{tree}, are of
70many varieties.  A @code{tree} is a pointer type, but the object to
71which it points may be of a variety of types.  From this point forward,
72we will refer to trees in ordinary type, rather than in @code{this
73font}, except when talking about the actual C type @code{tree}.
74
75You can tell what kind of node a particular tree is by using the
76@code{TREE_CODE} macro.  Many, many macros take trees as input and
77return trees as output.  However, most macros require a certain kind of
78tree node as input.  In other words, there is a type-system for trees,
79but it is not reflected in the C type-system.
80
81For safety, it is useful to configure GCC with @option{--enable-checking}.
82Although this results in a significant performance penalty (since all
83tree types are checked at run-time), and is therefore inappropriate in a
84release version, it is extremely helpful during the development process.
85
86Many macros behave as predicates.  Many, although not all, of these
87predicates end in @samp{_P}.  Do not rely on the result type of these
88macros being of any particular type.  You may, however, rely on the fact
89that the type can be compared to @code{0}, so that statements like
90@smallexample
91if (TEST_P (t) && !TEST_P (y))
92  x = 1;
93@end smallexample
94@noindent
95and
96@smallexample
97int i = (TEST_P (t) != 0);
98@end smallexample
99@noindent
100are legal.  Macros that return @code{int} values now may be changed to
101return @code{tree} values, or other pointers in the future.  Even those
102that continue to return @code{int} may return multiple nonzero codes
103where previously they returned only zero and one.  Therefore, you should
104not write code like
105@smallexample
106if (TEST_P (t) == 1)
107@end smallexample
108@noindent
109as this code is not guaranteed to work correctly in the future.
110
111You should not take the address of values returned by the macros or
112functions described here.  In particular, no guarantee is given that the
113values are lvalues.
114
115In general, the names of macros are all in uppercase, while the names of
116functions are entirely in lowercase.  There are rare exceptions to this
117rule.  You should assume that any macro or function whose name is made
118up entirely of uppercase letters may evaluate its arguments more than
119once.  You may assume that a macro or function whose name is made up
120entirely of lowercase letters will evaluate its arguments only once.
121
122The @code{error_mark_node} is a special tree.  Its tree code is
123@code{ERROR_MARK}, but since there is only ever one node with that code,
124the usual practice is to compare the tree against
125@code{error_mark_node}.  (This test is just a test for pointer
126equality.)  If an error has occurred during front-end processing the
127flag @code{errorcount} will be set.  If the front end has encountered
128code it cannot handle, it will issue a message to the user and set
129@code{sorrycount}.  When these flags are set, any macro or function
130which normally returns a tree of a particular kind may instead return
131the @code{error_mark_node}.  Thus, if you intend to do any processing of
132erroneous code, you must be prepared to deal with the
133@code{error_mark_node}.
134
135Occasionally, a particular tree slot (like an operand to an expression,
136or a particular field in a declaration) will be referred to as
137``reserved for the back end''.  These slots are used to store RTL when
138the tree is converted to RTL for use by the GCC back end.  However, if
139that process is not taking place (e.g., if the front end is being hooked
140up to an intelligent editor), then those slots may be used by the
141back end presently in use.
142
143If you encounter situations that do not match this documentation, such
144as tree nodes of types not mentioned here, or macros documented to
145return entities of a particular kind that instead return entities of
146some different kind, you have found a bug, either in the front end or in
147the documentation.  Please report these bugs as you would any other
148bug.
149
150@menu
151* Macros and Functions::Macros and functions that can be used with all trees.
152* Identifiers::         The names of things.
153* Containers::          Lists and vectors.
154@end menu
155
156@c ---------------------------------------------------------------------
157@c Trees
158@c ---------------------------------------------------------------------
159
160@node Macros and Functions
161@subsection Trees
162@cindex tree
163@findex TREE_CHAIN
164@findex TREE_TYPE
165
166All GENERIC trees have two fields in common.  First, @code{TREE_CHAIN}
167is a pointer that can be used as a singly-linked list to other trees.
168The other is @code{TREE_TYPE}.  Many trees store the type of an
169expression or declaration in this field.
170
171These are some other functions for handling trees:
172
173@ftable @code
174
175@item tree_size
176Return the number of bytes a tree takes.
177
178@item build0
179@itemx build1
180@itemx build2
181@itemx build3
182@itemx build4
183@itemx build5
184@itemx build6
185
186These functions build a tree and supply values to put in each
187parameter.  The basic signature is @samp{@w{code, type, [operands]}}.
188@code{code} is the @code{TREE_CODE}, and @code{type} is a tree
189representing the @code{TREE_TYPE}.  These are followed by the
190operands, each of which is also a tree.
191
192@end ftable
193
194
195@c ---------------------------------------------------------------------
196@c Identifiers
197@c ---------------------------------------------------------------------
198
199@node Identifiers
200@subsection Identifiers
201@cindex identifier
202@cindex name
203@tindex IDENTIFIER_NODE
204
205An @code{IDENTIFIER_NODE} represents a slightly more general concept
206that the standard C or C++ concept of identifier.  In particular, an
207@code{IDENTIFIER_NODE} may contain a @samp{$}, or other extraordinary
208characters.
209
210There are never two distinct @code{IDENTIFIER_NODE}s representing the
211same identifier.  Therefore, you may use pointer equality to compare
212@code{IDENTIFIER_NODE}s, rather than using a routine like
213@code{strcmp}.  Use @code{get_identifier} to obtain the unique
214@code{IDENTIFIER_NODE} for a supplied string.
215
216You can use the following macros to access identifiers:
217@ftable @code
218@item IDENTIFIER_POINTER
219The string represented by the identifier, represented as a
220@code{char*}.  This string is always @code{NUL}-terminated, and contains
221no embedded @code{NUL} characters.
222
223@item IDENTIFIER_LENGTH
224The length of the string returned by @code{IDENTIFIER_POINTER}, not
225including the trailing @code{NUL}.  This value of
226@code{IDENTIFIER_LENGTH (x)} is always the same as @code{strlen
227(IDENTIFIER_POINTER (x))}.
228
229@item IDENTIFIER_OPNAME_P
230This predicate holds if the identifier represents the name of an
231overloaded operator.  In this case, you should not depend on the
232contents of either the @code{IDENTIFIER_POINTER} or the
233@code{IDENTIFIER_LENGTH}.
234
235@item IDENTIFIER_TYPENAME_P
236This predicate holds if the identifier represents the name of a
237user-defined conversion operator.  In this case, the @code{TREE_TYPE} of
238the @code{IDENTIFIER_NODE} holds the type to which the conversion
239operator converts.
240
241@end ftable
242
243@c ---------------------------------------------------------------------
244@c Containers
245@c ---------------------------------------------------------------------
246
247@node Containers
248@subsection Containers
249@cindex container
250@cindex list
251@cindex vector
252@tindex TREE_LIST
253@tindex TREE_VEC
254@findex TREE_PURPOSE
255@findex TREE_VALUE
256@findex TREE_VEC_LENGTH
257@findex TREE_VEC_ELT
258
259Two common container data structures can be represented directly with
260tree nodes.  A @code{TREE_LIST} is a singly linked list containing two
261trees per node.  These are the @code{TREE_PURPOSE} and @code{TREE_VALUE}
262of each node.  (Often, the @code{TREE_PURPOSE} contains some kind of
263tag, or additional information, while the @code{TREE_VALUE} contains the
264majority of the payload.  In other cases, the @code{TREE_PURPOSE} is
265simply @code{NULL_TREE}, while in still others both the
266@code{TREE_PURPOSE} and @code{TREE_VALUE} are of equal stature.)  Given
267one @code{TREE_LIST} node, the next node is found by following the
268@code{TREE_CHAIN}.  If the @code{TREE_CHAIN} is @code{NULL_TREE}, then
269you have reached the end of the list.
270
271A @code{TREE_VEC} is a simple vector.  The @code{TREE_VEC_LENGTH} is an
272integer (not a tree) giving the number of nodes in the vector.  The
273nodes themselves are accessed using the @code{TREE_VEC_ELT} macro, which
274takes two arguments.  The first is the @code{TREE_VEC} in question; the
275second is an integer indicating which element in the vector is desired.
276The elements are indexed from zero.
277
278@c ---------------------------------------------------------------------
279@c Types
280@c ---------------------------------------------------------------------
281
282@node Types
283@section Types
284@cindex type
285@cindex pointer
286@cindex reference
287@cindex fundamental type
288@cindex array
289@tindex VOID_TYPE
290@tindex INTEGER_TYPE
291@tindex TYPE_MIN_VALUE
292@tindex TYPE_MAX_VALUE
293@tindex REAL_TYPE
294@tindex FIXED_POINT_TYPE
295@tindex COMPLEX_TYPE
296@tindex ENUMERAL_TYPE
297@tindex BOOLEAN_TYPE
298@tindex POINTER_TYPE
299@tindex REFERENCE_TYPE
300@tindex FUNCTION_TYPE
301@tindex METHOD_TYPE
302@tindex ARRAY_TYPE
303@tindex RECORD_TYPE
304@tindex UNION_TYPE
305@tindex UNKNOWN_TYPE
306@tindex OFFSET_TYPE
307@findex TYPE_UNQUALIFIED
308@findex TYPE_QUAL_CONST
309@findex TYPE_QUAL_VOLATILE
310@findex TYPE_QUAL_RESTRICT
311@findex TYPE_MAIN_VARIANT
312@cindex qualified type
313@findex TYPE_SIZE
314@findex TYPE_ALIGN
315@findex TYPE_PRECISION
316@findex TYPE_ARG_TYPES
317@findex TYPE_METHOD_BASETYPE
318@findex TYPE_OFFSET_BASETYPE
319@findex TREE_TYPE
320@findex TYPE_CONTEXT
321@findex TYPE_NAME
322@findex TYPENAME_TYPE_FULLNAME
323@findex TYPE_FIELDS
324@findex TYPE_CANONICAL
325@findex TYPE_STRUCTURAL_EQUALITY_P
326@findex SET_TYPE_STRUCTURAL_EQUALITY
327
328All types have corresponding tree nodes.  However, you should not assume
329that there is exactly one tree node corresponding to each type.  There
330are often multiple nodes corresponding to the same type.
331
332For the most part, different kinds of types have different tree codes.
333(For example, pointer types use a @code{POINTER_TYPE} code while arrays
334use an @code{ARRAY_TYPE} code.)  However, pointers to member functions
335use the @code{RECORD_TYPE} code.  Therefore, when writing a
336@code{switch} statement that depends on the code associated with a
337particular type, you should take care to handle pointers to member
338functions under the @code{RECORD_TYPE} case label.
339
340The following functions and macros deal with cv-qualification of types:
341@ftable @code
342@item TYPE_MAIN_VARIANT
343This macro returns the unqualified version of a type.  It may be applied
344to an unqualified type, but it is not always the identity function in
345that case.
346@end ftable
347
348A few other macros and functions are usable with all types:
349@ftable @code
350@item TYPE_SIZE
351The number of bits required to represent the type, represented as an
352@code{INTEGER_CST}.  For an incomplete type, @code{TYPE_SIZE} will be
353@code{NULL_TREE}.
354
355@item TYPE_ALIGN
356The alignment of the type, in bits, represented as an @code{int}.
357
358@item TYPE_NAME
359This macro returns a declaration (in the form of a @code{TYPE_DECL}) for
360the type.  (Note this macro does @emph{not} return an
361@code{IDENTIFIER_NODE}, as you might expect, given its name!)  You can
362look at the @code{DECL_NAME} of the @code{TYPE_DECL} to obtain the
363actual name of the type.  The @code{TYPE_NAME} will be @code{NULL_TREE}
364for a type that is not a built-in type, the result of a typedef, or a
365named class type.
366
367@item TYPE_CANONICAL
368This macro returns the ``canonical'' type for the given type
369node. Canonical types are used to improve performance in the C++ and
370Objective-C++ front ends by allowing efficient comparison between two
371type nodes in @code{same_type_p}: if the @code{TYPE_CANONICAL} values
372of the types are equal, the types are equivalent; otherwise, the types
373are not equivalent. The notion of equivalence for canonical types is
374the same as the notion of type equivalence in the language itself. For
375instance,
376
377When @code{TYPE_CANONICAL} is @code{NULL_TREE}, there is no canonical
378type for the given type node. In this case, comparison between this
379type and any other type requires the compiler to perform a deep,
380``structural'' comparison to see if the two type nodes have the same
381form and properties.
382
383The canonical type for a node is always the most fundamental type in
384the equivalence class of types. For instance, @code{int} is its own
385canonical type. A typedef @code{I} of @code{int} will have @code{int}
386as its canonical type. Similarly, @code{I*}@ and a typedef @code{IP}@
387(defined to @code{I*}) will has @code{int*} as their canonical
388type. When building a new type node, be sure to set
389@code{TYPE_CANONICAL} to the appropriate canonical type. If the new
390type is a compound type (built from other types), and any of those
391other types require structural equality, use
392@code{SET_TYPE_STRUCTURAL_EQUALITY} to ensure that the new type also
393requires structural equality. Finally, if for some reason you cannot
394guarantee that @code{TYPE_CANONICAL} will point to the canonical type,
395use @code{SET_TYPE_STRUCTURAL_EQUALITY} to make sure that the new
396type--and any type constructed based on it--requires structural
397equality. If you suspect that the canonical type system is
398miscomparing types, pass @code{--param verify-canonical-types=1} to
399the compiler or configure with @code{--enable-checking} to force the
400compiler to verify its canonical-type comparisons against the
401structural comparisons; the compiler will then print any warnings if
402the canonical types miscompare.
403
404@item TYPE_STRUCTURAL_EQUALITY_P
405This predicate holds when the node requires structural equality
406checks, e.g., when @code{TYPE_CANONICAL} is @code{NULL_TREE}.
407
408@item SET_TYPE_STRUCTURAL_EQUALITY
409This macro states that the type node it is given requires structural
410equality checks, e.g., it sets @code{TYPE_CANONICAL} to
411@code{NULL_TREE}.
412
413@item same_type_p
414This predicate takes two types as input, and holds if they are the same
415type.  For example, if one type is a @code{typedef} for the other, or
416both are @code{typedef}s for the same type.  This predicate also holds if
417the two trees given as input are simply copies of one another; i.e.,
418there is no difference between them at the source level, but, for
419whatever reason, a duplicate has been made in the representation.  You
420should never use @code{==} (pointer equality) to compare types; always
421use @code{same_type_p} instead.
422@end ftable
423
424Detailed below are the various kinds of types, and the macros that can
425be used to access them.  Although other kinds of types are used
426elsewhere in G++, the types described here are the only ones that you
427will encounter while examining the intermediate representation.
428
429@table @code
430@item VOID_TYPE
431Used to represent the @code{void} type.
432
433@item INTEGER_TYPE
434Used to represent the various integral types, including @code{char},
435@code{short}, @code{int}, @code{long}, and @code{long long}.  This code
436is not used for enumeration types, nor for the @code{bool} type.
437The @code{TYPE_PRECISION} is the number of bits used in
438the representation, represented as an @code{unsigned int}.  (Note that
439in the general case this is not the same value as @code{TYPE_SIZE};
440suppose that there were a 24-bit integer type, but that alignment
441requirements for the ABI required 32-bit alignment.  Then,
442@code{TYPE_SIZE} would be an @code{INTEGER_CST} for 32, while
443@code{TYPE_PRECISION} would be 24.)  The integer type is unsigned if
444@code{TYPE_UNSIGNED} holds; otherwise, it is signed.
445
446The @code{TYPE_MIN_VALUE} is an @code{INTEGER_CST} for the smallest
447integer that may be represented by this type.  Similarly, the
448@code{TYPE_MAX_VALUE} is an @code{INTEGER_CST} for the largest integer
449that may be represented by this type.
450
451@item REAL_TYPE
452Used to represent the @code{float}, @code{double}, and @code{long
453double} types.  The number of bits in the floating-point representation
454is given by @code{TYPE_PRECISION}, as in the @code{INTEGER_TYPE} case.
455
456@item FIXED_POINT_TYPE
457Used to represent the @code{short _Fract}, @code{_Fract}, @code{long
458_Fract}, @code{long long _Fract}, @code{short _Accum}, @code{_Accum},
459@code{long _Accum}, and @code{long long _Accum} types.  The number of bits
460in the fixed-point representation is given by @code{TYPE_PRECISION},
461as in the @code{INTEGER_TYPE} case.  There may be padding bits, fractional
462bits and integral bits.  The number of fractional bits is given by
463@code{TYPE_FBIT}, and the number of integral bits is given by @code{TYPE_IBIT}.
464The fixed-point type is unsigned if @code{TYPE_UNSIGNED} holds; otherwise,
465it is signed.
466The fixed-point type is saturating if @code{TYPE_SATURATING} holds; otherwise,
467it is not saturating.
468
469@item COMPLEX_TYPE
470Used to represent GCC built-in @code{__complex__} data types.  The
471@code{TREE_TYPE} is the type of the real and imaginary parts.
472
473@item ENUMERAL_TYPE
474Used to represent an enumeration type.  The @code{TYPE_PRECISION} gives
475(as an @code{int}), the number of bits used to represent the type.  If
476there are no negative enumeration constants, @code{TYPE_UNSIGNED} will
477hold.  The minimum and maximum enumeration constants may be obtained
478with @code{TYPE_MIN_VALUE} and @code{TYPE_MAX_VALUE}, respectively; each
479of these macros returns an @code{INTEGER_CST}.
480
481The actual enumeration constants themselves may be obtained by looking
482at the @code{TYPE_VALUES}.  This macro will return a @code{TREE_LIST},
483containing the constants.  The @code{TREE_PURPOSE} of each node will be
484an @code{IDENTIFIER_NODE} giving the name of the constant; the
485@code{TREE_VALUE} will be an @code{INTEGER_CST} giving the value
486assigned to that constant.  These constants will appear in the order in
487which they were declared.  The @code{TREE_TYPE} of each of these
488constants will be the type of enumeration type itself.
489
490@item BOOLEAN_TYPE
491Used to represent the @code{bool} type.
492
493@item POINTER_TYPE
494Used to represent pointer types, and pointer to data member types.  The
495@code{TREE_TYPE} gives the type to which this type points.
496
497@item REFERENCE_TYPE
498Used to represent reference types.  The @code{TREE_TYPE} gives the type
499to which this type refers.
500
501@item FUNCTION_TYPE
502Used to represent the type of non-member functions and of static member
503functions.  The @code{TREE_TYPE} gives the return type of the function.
504The @code{TYPE_ARG_TYPES} are a @code{TREE_LIST} of the argument types.
505The @code{TREE_VALUE} of each node in this list is the type of the
506corresponding argument; the @code{TREE_PURPOSE} is an expression for the
507default argument value, if any.  If the last node in the list is
508@code{void_list_node} (a @code{TREE_LIST} node whose @code{TREE_VALUE}
509is the @code{void_type_node}), then functions of this type do not take
510variable arguments.  Otherwise, they do take a variable number of
511arguments.
512
513Note that in C (but not in C++) a function declared like @code{void f()}
514is an unprototyped function taking a variable number of arguments; the
515@code{TYPE_ARG_TYPES} of such a function will be @code{NULL}.
516
517@item METHOD_TYPE
518Used to represent the type of a non-static member function.  Like a
519@code{FUNCTION_TYPE}, the return type is given by the @code{TREE_TYPE}.
520The type of @code{*this}, i.e., the class of which functions of this
521type are a member, is given by the @code{TYPE_METHOD_BASETYPE}.  The
522@code{TYPE_ARG_TYPES} is the parameter list, as for a
523@code{FUNCTION_TYPE}, and includes the @code{this} argument.
524
525@item ARRAY_TYPE
526Used to represent array types.  The @code{TREE_TYPE} gives the type of
527the elements in the array.  If the array-bound is present in the type,
528the @code{TYPE_DOMAIN} is an @code{INTEGER_TYPE} whose
529@code{TYPE_MIN_VALUE} and @code{TYPE_MAX_VALUE} will be the lower and
530upper bounds of the array, respectively.  The @code{TYPE_MIN_VALUE} will
531always be an @code{INTEGER_CST} for zero, while the
532@code{TYPE_MAX_VALUE} will be one less than the number of elements in
533the array, i.e., the highest value which may be used to index an element
534in the array.
535
536@item RECORD_TYPE
537Used to represent @code{struct} and @code{class} types, as well as
538pointers to member functions and similar constructs in other languages.
539@code{TYPE_FIELDS} contains the items contained in this type, each of
540which can be a @code{FIELD_DECL}, @code{VAR_DECL}, @code{CONST_DECL}, or
541@code{TYPE_DECL}.  You may not make any assumptions about the ordering
542of the fields in the type or whether one or more of them overlap.
543
544@item UNION_TYPE
545Used to represent @code{union} types.  Similar to @code{RECORD_TYPE}
546except that all @code{FIELD_DECL} nodes in @code{TYPE_FIELD} start at
547bit position zero.
548
549@item QUAL_UNION_TYPE
550Used to represent part of a variant record in Ada.  Similar to
551@code{UNION_TYPE} except that each @code{FIELD_DECL} has a
552@code{DECL_QUALIFIER} field, which contains a boolean expression that
553indicates whether the field is present in the object.  The type will only
554have one field, so each field's @code{DECL_QUALIFIER} is only evaluated
555if none of the expressions in the previous fields in @code{TYPE_FIELDS}
556are nonzero.  Normally these expressions will reference a field in the
557outer object using a @code{PLACEHOLDER_EXPR}.
558
559@item LANG_TYPE
560This node is used to represent a language-specific type.  The front
561end must handle it.
562
563@item OFFSET_TYPE
564This node is used to represent a pointer-to-data member.  For a data
565member @code{X::m} the @code{TYPE_OFFSET_BASETYPE} is @code{X} and the
566@code{TREE_TYPE} is the type of @code{m}.
567
568@end table
569
570There are variables whose values represent some of the basic types.
571These include:
572@table @code
573@item void_type_node
574A node for @code{void}.
575
576@item integer_type_node
577A node for @code{int}.
578
579@item unsigned_type_node.
580A node for @code{unsigned int}.
581
582@item char_type_node.
583A node for @code{char}.
584@end table
585@noindent
586It may sometimes be useful to compare one of these variables with a type
587in hand, using @code{same_type_p}.
588
589@c ---------------------------------------------------------------------
590@c Declarations
591@c ---------------------------------------------------------------------
592
593@node Declarations
594@section Declarations
595@cindex declaration
596@cindex variable
597@cindex type declaration
598@tindex LABEL_DECL
599@tindex CONST_DECL
600@tindex TYPE_DECL
601@tindex VAR_DECL
602@tindex PARM_DECL
603@tindex DEBUG_EXPR_DECL
604@tindex FIELD_DECL
605@tindex NAMESPACE_DECL
606@tindex RESULT_DECL
607@tindex TEMPLATE_DECL
608@tindex THUNK_DECL
609@findex THUNK_DELTA
610@findex DECL_INITIAL
611@findex DECL_SIZE
612@findex DECL_ALIGN
613@findex DECL_EXTERNAL
614
615This section covers the various kinds of declarations that appear in the
616internal representation, except for declarations of functions
617(represented by @code{FUNCTION_DECL} nodes), which are described in
618@ref{Functions}.
619
620@menu
621* Working with declarations::  Macros and functions that work on
622declarations.
623* Internal structure:: How declaration nodes are represented.
624@end menu
625
626@node Working with declarations
627@subsection Working with declarations
628
629Some macros can be used with any kind of declaration.  These include:
630@ftable @code
631@item DECL_NAME
632This macro returns an @code{IDENTIFIER_NODE} giving the name of the
633entity.
634
635@item TREE_TYPE
636This macro returns the type of the entity declared.
637
638@item EXPR_FILENAME
639This macro returns the name of the file in which the entity was
640declared, as a @code{char*}.  For an entity declared implicitly by the
641compiler (like @code{__builtin_memcpy}), this will be the string
642@code{"<internal>"}.
643
644@item EXPR_LINENO
645This macro returns the line number at which the entity was declared, as
646an @code{int}.
647
648@item DECL_ARTIFICIAL
649This predicate holds if the declaration was implicitly generated by the
650compiler.  For example, this predicate will hold of an implicitly
651declared member function, or of the @code{TYPE_DECL} implicitly
652generated for a class type.  Recall that in C++ code like:
653@smallexample
654struct S @{@};
655@end smallexample
656@noindent
657is roughly equivalent to C code like:
658@smallexample
659struct S @{@};
660typedef struct S S;
661@end smallexample
662The implicitly generated @code{typedef} declaration is represented by a
663@code{TYPE_DECL} for which @code{DECL_ARTIFICIAL} holds.
664
665@end ftable
666
667The various kinds of declarations include:
668@table @code
669@item LABEL_DECL
670These nodes are used to represent labels in function bodies.  For more
671information, see @ref{Functions}.  These nodes only appear in block
672scopes.
673
674@item CONST_DECL
675These nodes are used to represent enumeration constants.  The value of
676the constant is given by @code{DECL_INITIAL} which will be an
677@code{INTEGER_CST} with the same type as the @code{TREE_TYPE} of the
678@code{CONST_DECL}, i.e., an @code{ENUMERAL_TYPE}.
679
680@item RESULT_DECL
681These nodes represent the value returned by a function.  When a value is
682assigned to a @code{RESULT_DECL}, that indicates that the value should
683be returned, via bitwise copy, by the function.  You can use
684@code{DECL_SIZE} and @code{DECL_ALIGN} on a @code{RESULT_DECL}, just as
685with a @code{VAR_DECL}.
686
687@item TYPE_DECL
688These nodes represent @code{typedef} declarations.  The @code{TREE_TYPE}
689is the type declared to have the name given by @code{DECL_NAME}.  In
690some cases, there is no associated name.
691
692@item VAR_DECL
693These nodes represent variables with namespace or block scope, as well
694as static data members.  The @code{DECL_SIZE} and @code{DECL_ALIGN} are
695analogous to @code{TYPE_SIZE} and @code{TYPE_ALIGN}.  For a declaration,
696you should always use the @code{DECL_SIZE} and @code{DECL_ALIGN} rather
697than the @code{TYPE_SIZE} and @code{TYPE_ALIGN} given by the
698@code{TREE_TYPE}, since special attributes may have been applied to the
699variable to give it a particular size and alignment.  You may use the
700predicates @code{DECL_THIS_STATIC} or @code{DECL_THIS_EXTERN} to test
701whether the storage class specifiers @code{static} or @code{extern} were
702used to declare a variable.
703
704If this variable is initialized (but does not require a constructor),
705the @code{DECL_INITIAL} will be an expression for the initializer.  The
706initializer should be evaluated, and a bitwise copy into the variable
707performed.  If the @code{DECL_INITIAL} is the @code{error_mark_node},
708there is an initializer, but it is given by an explicit statement later
709in the code; no bitwise copy is required.
710
711GCC provides an extension that allows either automatic variables, or
712global variables, to be placed in particular registers.  This extension
713is being used for a particular @code{VAR_DECL} if @code{DECL_REGISTER}
714holds for the @code{VAR_DECL}, and if @code{DECL_ASSEMBLER_NAME} is not
715equal to @code{DECL_NAME}.  In that case, @code{DECL_ASSEMBLER_NAME} is
716the name of the register into which the variable will be placed.
717
718@item PARM_DECL
719Used to represent a parameter to a function.  Treat these nodes
720similarly to @code{VAR_DECL} nodes.  These nodes only appear in the
721@code{DECL_ARGUMENTS} for a @code{FUNCTION_DECL}.
722
723The @code{DECL_ARG_TYPE} for a @code{PARM_DECL} is the type that will
724actually be used when a value is passed to this function.  It may be a
725wider type than the @code{TREE_TYPE} of the parameter; for example, the
726ordinary type might be @code{short} while the @code{DECL_ARG_TYPE} is
727@code{int}.
728
729@item DEBUG_EXPR_DECL
730Used to represent an anonymous debug-information temporary created to
731hold an expression as it is optimized away, so that its value can be
732referenced in debug bind statements.
733
734@item FIELD_DECL
735These nodes represent non-static data members.  The @code{DECL_SIZE} and
736@code{DECL_ALIGN} behave as for @code{VAR_DECL} nodes.
737The position of the field within the parent record is specified by a
738combination of three attributes.  @code{DECL_FIELD_OFFSET} is the position,
739counting in bytes, of the @code{DECL_OFFSET_ALIGN}-bit sized word containing
740the bit of the field closest to the beginning of the structure.
741@code{DECL_FIELD_BIT_OFFSET} is the bit offset of the first bit of the field
742within this word; this may be nonzero even for fields that are not bit-fields,
743since @code{DECL_OFFSET_ALIGN} may be greater than the natural alignment
744of the field's type.
745
746If @code{DECL_C_BIT_FIELD} holds, this field is a bit-field.  In a bit-field,
747@code{DECL_BIT_FIELD_TYPE} also contains the type that was originally
748specified for it, while DECL_TYPE may be a modified type with lesser precision,
749according to the size of the bit field.
750
751@item NAMESPACE_DECL
752Namespaces provide a name hierarchy for other declarations.  They
753appear in the @code{DECL_CONTEXT} of other @code{_DECL} nodes.
754
755@end table
756
757@node Internal structure
758@subsection Internal structure
759
760@code{DECL} nodes are represented internally as a hierarchy of
761structures.
762
763@menu
764* Current structure hierarchy::  The current DECL node structure
765hierarchy.
766* Adding new DECL node types:: How to add a new DECL node to a
767frontend.
768@end menu
769
770@node Current structure hierarchy
771@subsubsection Current structure hierarchy
772
773@table @code
774
775@item struct tree_decl_minimal
776This is the minimal structure to inherit from in order for common
777@code{DECL} macros to work.  The fields it contains are a unique ID,
778source location, context, and name.
779
780@item struct tree_decl_common
781This structure inherits from @code{struct tree_decl_minimal}.  It
782contains fields that most @code{DECL} nodes need, such as a field to
783store alignment, machine mode, size, and attributes.
784
785@item struct tree_field_decl
786This structure inherits from @code{struct tree_decl_common}.  It is
787used to represent @code{FIELD_DECL}.
788
789@item struct tree_label_decl
790This structure inherits from @code{struct tree_decl_common}.  It is
791used to represent @code{LABEL_DECL}.
792
793@item struct tree_translation_unit_decl
794This structure inherits from @code{struct tree_decl_common}.  It is
795used to represent @code{TRANSLATION_UNIT_DECL}.
796
797@item struct tree_decl_with_rtl
798This structure inherits from @code{struct tree_decl_common}.  It
799contains a field to store the low-level RTL associated with a
800@code{DECL} node.
801
802@item struct tree_result_decl
803This structure inherits from @code{struct tree_decl_with_rtl}.  It is
804used to represent @code{RESULT_DECL}.
805
806@item struct tree_const_decl
807This structure inherits from @code{struct tree_decl_with_rtl}.  It is
808used to represent @code{CONST_DECL}.
809
810@item struct tree_parm_decl
811This structure inherits from @code{struct tree_decl_with_rtl}.  It is
812used to represent @code{PARM_DECL}.
813
814@item struct tree_decl_with_vis
815This structure inherits from @code{struct tree_decl_with_rtl}.  It
816contains fields necessary to store visibility information, as well as
817a section name and assembler name.
818
819@item struct tree_var_decl
820This structure inherits from @code{struct tree_decl_with_vis}.  It is
821used to represent @code{VAR_DECL}.
822
823@item struct tree_function_decl
824This structure inherits from @code{struct tree_decl_with_vis}.  It is
825used to represent @code{FUNCTION_DECL}.
826
827@end table
828@node Adding new DECL node types
829@subsubsection Adding new DECL node types
830
831Adding a new @code{DECL} tree consists of the following steps
832
833@table @asis
834
835@item Add a new tree code for the @code{DECL} node
836For language specific @code{DECL} nodes, there is a @file{.def} file
837in each frontend directory where the tree code should be added.
838For @code{DECL} nodes that are part of the middle-end, the code should
839be added to @file{tree.def}.
840
841@item Create a new structure type for the @code{DECL} node
842These structures should inherit from one of the existing structures in
843the language hierarchy by using that structure as the first member.
844
845@smallexample
846struct tree_foo_decl
847@{
848   struct tree_decl_with_vis common;
849@}
850@end smallexample
851
852Would create a structure name @code{tree_foo_decl} that inherits from
853@code{struct tree_decl_with_vis}.
854
855For language specific @code{DECL} nodes, this new structure type
856should go in the appropriate @file{.h} file.
857For @code{DECL} nodes that are part of the middle-end, the structure
858type should go in @file{tree.h}.
859
860@item Add a member to the tree structure enumerator for the node
861For garbage collection and dynamic checking purposes, each @code{DECL}
862node structure type is required to have a unique enumerator value
863specified with it.
864For language specific @code{DECL} nodes, this new enumerator value
865should go in the appropriate @file{.def} file.
866For @code{DECL} nodes that are part of the middle-end, the enumerator
867values are specified in @file{treestruct.def}.
868
869@item Update @code{union tree_node}
870In order to make your new structure type usable, it must be added to
871@code{union tree_node}.
872For language specific @code{DECL} nodes, a new entry should be added
873to the appropriate @file{.h} file of the form
874@smallexample
875  struct tree_foo_decl GTY ((tag ("TS_VAR_DECL"))) foo_decl;
876@end smallexample
877For @code{DECL} nodes that are part of the middle-end, the additional
878member goes directly into @code{union tree_node} in @file{tree.h}.
879
880@item Update dynamic checking info
881In order to be able to check whether accessing a named portion of
882@code{union tree_node} is legal, and whether a certain @code{DECL} node
883contains one of the enumerated @code{DECL} node structures in the
884hierarchy, a simple lookup table is used.
885This lookup table needs to be kept up to date with the tree structure
886hierarchy, or else checking and containment macros will fail
887inappropriately.
888
889For language specific @code{DECL} nodes, their is an @code{init_ts}
890function in an appropriate @file{.c} file, which initializes the lookup
891table.
892Code setting up the table for new @code{DECL} nodes should be added
893there.
894For each @code{DECL} tree code and enumerator value representing a
895member of the inheritance  hierarchy, the table should contain 1 if
896that tree code inherits (directly or indirectly) from that member.
897Thus, a @code{FOO_DECL} node derived from @code{struct decl_with_rtl},
898and enumerator value @code{TS_FOO_DECL}, would be set up as follows
899@smallexample
900tree_contains_struct[FOO_DECL][TS_FOO_DECL] = 1;
901tree_contains_struct[FOO_DECL][TS_DECL_WRTL] = 1;
902tree_contains_struct[FOO_DECL][TS_DECL_COMMON] = 1;
903tree_contains_struct[FOO_DECL][TS_DECL_MINIMAL] = 1;
904@end smallexample
905
906For @code{DECL} nodes that are part of the middle-end, the setup code
907goes into @file{tree.c}.
908
909@item Add macros to access any new fields and flags
910
911Each added field or flag should have a macro that is used to access
912it, that performs appropriate checking to ensure only the right type of
913@code{DECL} nodes access the field.
914
915These macros generally take the following form
916@smallexample
917#define FOO_DECL_FIELDNAME(NODE) FOO_DECL_CHECK(NODE)->foo_decl.fieldname
918@end smallexample
919However, if the structure is simply a base class for further
920structures, something like the following should be used
921@smallexample
922#define BASE_STRUCT_CHECK(T) CONTAINS_STRUCT_CHECK(T, TS_BASE_STRUCT)
923#define BASE_STRUCT_FIELDNAME(NODE) \
924   (BASE_STRUCT_CHECK(NODE)->base_struct.fieldname
925@end smallexample
926
927@end table
928
929
930@c ---------------------------------------------------------------------
931@c Attributes
932@c ---------------------------------------------------------------------
933@node Attributes
934@section Attributes in trees
935@cindex attributes
936
937Attributes, as specified using the @code{__attribute__} keyword, are
938represented internally as a @code{TREE_LIST}.  The @code{TREE_PURPOSE}
939is the name of the attribute, as an @code{IDENTIFIER_NODE}.  The
940@code{TREE_VALUE} is a @code{TREE_LIST} of the arguments of the
941attribute, if any, or @code{NULL_TREE} if there are no arguments; the
942arguments are stored as the @code{TREE_VALUE} of successive entries in
943the list, and may be identifiers or expressions.  The @code{TREE_CHAIN}
944of the attribute is the next attribute in a list of attributes applying
945to the same declaration or type, or @code{NULL_TREE} if there are no
946further attributes in the list.
947
948Attributes may be attached to declarations and to types; these
949attributes may be accessed with the following macros.  All attributes
950are stored in this way, and many also cause other changes to the
951declaration or type or to other internal compiler data structures.
952
953@deftypefn {Tree Macro} tree DECL_ATTRIBUTES (tree @var{decl})
954This macro returns the attributes on the declaration @var{decl}.
955@end deftypefn
956
957@deftypefn {Tree Macro} tree TYPE_ATTRIBUTES (tree @var{type})
958This macro returns the attributes on the type @var{type}.
959@end deftypefn
960
961
962@c ---------------------------------------------------------------------
963@c Expressions
964@c ---------------------------------------------------------------------
965
966@node Expression trees
967@section Expressions
968@cindex expression
969@findex TREE_TYPE
970@findex TREE_OPERAND
971
972The internal representation for expressions is for the most part quite
973straightforward.  However, there are a few facts that one must bear in
974mind.  In particular, the expression ``tree'' is actually a directed
975acyclic graph.  (For example there may be many references to the integer
976constant zero throughout the source program; many of these will be
977represented by the same expression node.)  You should not rely on
978certain kinds of node being shared, nor should you rely on certain kinds of
979nodes being unshared.
980
981The following macros can be used with all expression nodes:
982
983@ftable @code
984@item TREE_TYPE
985Returns the type of the expression.  This value may not be precisely the
986same type that would be given the expression in the original program.
987@end ftable
988
989In what follows, some nodes that one might expect to always have type
990@code{bool} are documented to have either integral or boolean type.  At
991some point in the future, the C front end may also make use of this same
992intermediate representation, and at this point these nodes will
993certainly have integral type.  The previous sentence is not meant to
994imply that the C++ front end does not or will not give these nodes
995integral type.
996
997Below, we list the various kinds of expression nodes.  Except where
998noted otherwise, the operands to an expression are accessed using the
999@code{TREE_OPERAND} macro.  For example, to access the first operand to
1000a binary plus expression @code{expr}, use:
1001
1002@smallexample
1003TREE_OPERAND (expr, 0)
1004@end smallexample
1005@noindent
1006
1007As this example indicates, the operands are zero-indexed.
1008
1009
1010@menu
1011* Constants: Constant expressions.
1012* Storage References::
1013* Unary and Binary Expressions::
1014* Vectors::
1015@end menu
1016
1017@node Constant expressions
1018@subsection Constant expressions
1019@tindex INTEGER_CST
1020@findex TREE_INT_CST_HIGH
1021@findex TREE_INT_CST_LOW
1022@findex tree_int_cst_lt
1023@findex tree_int_cst_equal
1024@tindex REAL_CST
1025@tindex FIXED_CST
1026@tindex COMPLEX_CST
1027@tindex VECTOR_CST
1028@tindex STRING_CST
1029@findex TREE_STRING_LENGTH
1030@findex TREE_STRING_POINTER
1031
1032The table below begins with constants, moves on to unary expressions,
1033then proceeds to binary expressions, and concludes with various other
1034kinds of expressions:
1035
1036@table @code
1037@item INTEGER_CST
1038These nodes represent integer constants.  Note that the type of these
1039constants is obtained with @code{TREE_TYPE}; they are not always of type
1040@code{int}.  In particular, @code{char} constants are represented with
1041@code{INTEGER_CST} nodes.  The value of the integer constant @code{e} is
1042given by
1043@smallexample
1044((TREE_INT_CST_HIGH (e) << HOST_BITS_PER_WIDE_INT)
1045+ TREE_INST_CST_LOW (e))
1046@end smallexample
1047@noindent
1048HOST_BITS_PER_WIDE_INT is at least thirty-two on all platforms.  Both
1049@code{TREE_INT_CST_HIGH} and @code{TREE_INT_CST_LOW} return a
1050@code{HOST_WIDE_INT}.  The value of an @code{INTEGER_CST} is interpreted
1051as a signed or unsigned quantity depending on the type of the constant.
1052In general, the expression given above will overflow, so it should not
1053be used to calculate the value of the constant.
1054
1055The variable @code{integer_zero_node} is an integer constant with value
1056zero.  Similarly, @code{integer_one_node} is an integer constant with
1057value one.  The @code{size_zero_node} and @code{size_one_node} variables
1058are analogous, but have type @code{size_t} rather than @code{int}.
1059
1060The function @code{tree_int_cst_lt} is a predicate which holds if its
1061first argument is less than its second.  Both constants are assumed to
1062have the same signedness (i.e., either both should be signed or both
1063should be unsigned.)  The full width of the constant is used when doing
1064the comparison; the usual rules about promotions and conversions are
1065ignored.  Similarly, @code{tree_int_cst_equal} holds if the two
1066constants are equal.  The @code{tree_int_cst_sgn} function returns the
1067sign of a constant.  The value is @code{1}, @code{0}, or @code{-1}
1068according on whether the constant is greater than, equal to, or less
1069than zero.  Again, the signedness of the constant's type is taken into
1070account; an unsigned constant is never less than zero, no matter what
1071its bit-pattern.
1072
1073@item REAL_CST
1074
1075FIXME: Talk about how to obtain representations of this constant, do
1076comparisons, and so forth.
1077
1078@item FIXED_CST
1079
1080These nodes represent fixed-point constants.  The type of these constants
1081is obtained with @code{TREE_TYPE}.  @code{TREE_FIXED_CST_PTR} points to
1082a @code{struct fixed_value};  @code{TREE_FIXED_CST} returns the structure
1083itself.  @code{struct fixed_value} contains @code{data} with the size of two
1084@code{HOST_BITS_PER_WIDE_INT} and @code{mode} as the associated fixed-point
1085machine mode for @code{data}.
1086
1087@item COMPLEX_CST
1088These nodes are used to represent complex number constants, that is a
1089@code{__complex__} whose parts are constant nodes.  The
1090@code{TREE_REALPART} and @code{TREE_IMAGPART} return the real and the
1091imaginary parts respectively.
1092
1093@item VECTOR_CST
1094These nodes are used to represent vector constants, whose parts are
1095constant nodes.  Each individual constant node is either an integer or a
1096double constant node.  The first operand is a @code{TREE_LIST} of the
1097constant nodes and is accessed through @code{TREE_VECTOR_CST_ELTS}.
1098
1099@item STRING_CST
1100These nodes represent string-constants.  The @code{TREE_STRING_LENGTH}
1101returns the length of the string, as an @code{int}.  The
1102@code{TREE_STRING_POINTER} is a @code{char*} containing the string
1103itself.  The string may not be @code{NUL}-terminated, and it may contain
1104embedded @code{NUL} characters.  Therefore, the
1105@code{TREE_STRING_LENGTH} includes the trailing @code{NUL} if it is
1106present.
1107
1108For wide string constants, the @code{TREE_STRING_LENGTH} is the number
1109of bytes in the string, and the @code{TREE_STRING_POINTER}
1110points to an array of the bytes of the string, as represented on the
1111target system (that is, as integers in the target endianness).  Wide and
1112non-wide string constants are distinguished only by the @code{TREE_TYPE}
1113of the @code{STRING_CST}.
1114
1115FIXME: The formats of string constants are not well-defined when the
1116target system bytes are not the same width as host system bytes.
1117
1118@end table
1119
1120@node Storage References
1121@subsection References to storage
1122@tindex ADDR_EXPR
1123@tindex INDIRECT_REF
1124@tindex MEM_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 MEM_REF
1186These nodes are used to represent the object pointed to by a pointer
1187offset by a constant.
1188The first operand is the pointer being dereferenced; it will always have
1189pointer or reference type.  The second operand is a pointer constant.
1190Its type is specifying the type to be used for type-based alias analysis.
1191
1192@item COMPONENT_REF
1193These nodes represent non-static data member accesses.  The first
1194operand is the object (rather than a pointer to it); the second operand
1195is the @code{FIELD_DECL} for the data member.  The third operand represents
1196the byte offset of the field, but should not be used directly; call
1197@code{component_ref_field_offset} instead.
1198
1199
1200@end table
1201
1202@node Unary and Binary Expressions
1203@subsection Unary and Binary Expressions
1204@tindex NEGATE_EXPR
1205@tindex ABS_EXPR
1206@tindex BIT_NOT_EXPR
1207@tindex TRUTH_NOT_EXPR
1208@tindex PREDECREMENT_EXPR
1209@tindex PREINCREMENT_EXPR
1210@tindex POSTDECREMENT_EXPR
1211@tindex POSTINCREMENT_EXPR
1212@tindex FIX_TRUNC_EXPR
1213@tindex FLOAT_EXPR
1214@tindex COMPLEX_EXPR
1215@tindex CONJ_EXPR
1216@tindex REALPART_EXPR
1217@tindex IMAGPART_EXPR
1218@tindex NON_LVALUE_EXPR
1219@tindex NOP_EXPR
1220@tindex CONVERT_EXPR
1221@tindex FIXED_CONVERT_EXPR
1222@tindex THROW_EXPR
1223@tindex LSHIFT_EXPR
1224@tindex RSHIFT_EXPR
1225@tindex BIT_IOR_EXPR
1226@tindex BIT_XOR_EXPR
1227@tindex BIT_AND_EXPR
1228@tindex TRUTH_ANDIF_EXPR
1229@tindex TRUTH_ORIF_EXPR
1230@tindex TRUTH_AND_EXPR
1231@tindex TRUTH_OR_EXPR
1232@tindex TRUTH_XOR_EXPR
1233@tindex POINTER_PLUS_EXPR
1234@tindex PLUS_EXPR
1235@tindex MINUS_EXPR
1236@tindex MULT_EXPR
1237@tindex MULT_HIGHPART_EXPR
1238@tindex RDIV_EXPR
1239@tindex TRUNC_DIV_EXPR
1240@tindex FLOOR_DIV_EXPR
1241@tindex CEIL_DIV_EXPR
1242@tindex ROUND_DIV_EXPR
1243@tindex TRUNC_MOD_EXPR
1244@tindex FLOOR_MOD_EXPR
1245@tindex CEIL_MOD_EXPR
1246@tindex ROUND_MOD_EXPR
1247@tindex EXACT_DIV_EXPR
1248@tindex LT_EXPR
1249@tindex LE_EXPR
1250@tindex GT_EXPR
1251@tindex GE_EXPR
1252@tindex EQ_EXPR
1253@tindex NE_EXPR
1254@tindex ORDERED_EXPR
1255@tindex UNORDERED_EXPR
1256@tindex UNLT_EXPR
1257@tindex UNLE_EXPR
1258@tindex UNGT_EXPR
1259@tindex UNGE_EXPR
1260@tindex UNEQ_EXPR
1261@tindex LTGT_EXPR
1262@tindex MODIFY_EXPR
1263@tindex INIT_EXPR
1264@tindex COMPOUND_EXPR
1265@tindex COND_EXPR
1266@tindex CALL_EXPR
1267@tindex STMT_EXPR
1268@tindex BIND_EXPR
1269@tindex LOOP_EXPR
1270@tindex EXIT_EXPR
1271@tindex CLEANUP_POINT_EXPR
1272@tindex CONSTRUCTOR
1273@tindex COMPOUND_LITERAL_EXPR
1274@tindex SAVE_EXPR
1275@tindex TARGET_EXPR
1276@tindex VA_ARG_EXPR
1277
1278@table @code
1279@item NEGATE_EXPR
1280These nodes represent unary negation of the single operand, for both
1281integer and floating-point types.  The type of negation can be
1282determined by looking at the type of the expression.
1283
1284The behavior of this operation on signed arithmetic overflow is
1285controlled by the @code{flag_wrapv} and @code{flag_trapv} variables.
1286
1287@item ABS_EXPR
1288These nodes represent the absolute value of the single operand, for
1289both integer and floating-point types.  This is typically used to
1290implement the @code{abs}, @code{labs} and @code{llabs} builtins for
1291integer types, and the @code{fabs}, @code{fabsf} and @code{fabsl}
1292builtins for floating point types.  The type of abs operation can
1293be determined by looking at the type of the expression.
1294
1295This node is not used for complex types.  To represent the modulus
1296or complex abs of a complex value, use the @code{BUILT_IN_CABS},
1297@code{BUILT_IN_CABSF} or @code{BUILT_IN_CABSL} builtins, as used
1298to implement the C99 @code{cabs}, @code{cabsf} and @code{cabsl}
1299built-in functions.
1300
1301@item BIT_NOT_EXPR
1302These nodes represent bitwise complement, and will always have integral
1303type.  The only operand is the value to be complemented.
1304
1305@item TRUTH_NOT_EXPR
1306These nodes represent logical negation, and will always have integral
1307(or boolean) type.  The operand is the value being negated.  The type
1308of the operand and that of the result are always of @code{BOOLEAN_TYPE}
1309or @code{INTEGER_TYPE}.
1310
1311@item PREDECREMENT_EXPR
1312@itemx PREINCREMENT_EXPR
1313@itemx POSTDECREMENT_EXPR
1314@itemx POSTINCREMENT_EXPR
1315These nodes represent increment and decrement expressions.  The value of
1316the single operand is computed, and the operand incremented or
1317decremented.  In the case of @code{PREDECREMENT_EXPR} and
1318@code{PREINCREMENT_EXPR}, the value of the expression is the value
1319resulting after the increment or decrement; in the case of
1320@code{POSTDECREMENT_EXPR} and @code{POSTINCREMENT_EXPR} is the value
1321before the increment or decrement occurs.  The type of the operand, like
1322that of the result, will be either integral, boolean, or floating-point.
1323
1324@item FIX_TRUNC_EXPR
1325These nodes represent conversion of a floating-point value to an
1326integer.  The single operand will have a floating-point type, while
1327the complete expression will have an integral (or boolean) type.  The
1328operand is rounded towards zero.
1329
1330@item FLOAT_EXPR
1331These nodes represent conversion of an integral (or boolean) value to a
1332floating-point value.  The single operand will have integral type, while
1333the complete expression will have a floating-point type.
1334
1335FIXME: How is the operand supposed to be rounded?  Is this dependent on
1336@option{-mieee}?
1337
1338@item COMPLEX_EXPR
1339These nodes are used to represent complex numbers constructed from two
1340expressions of the same (integer or real) type.  The first operand is the
1341real part and the second operand is the imaginary part.
1342
1343@item CONJ_EXPR
1344These nodes represent the conjugate of their operand.
1345
1346@item REALPART_EXPR
1347@itemx IMAGPART_EXPR
1348These nodes represent respectively the real and the imaginary parts
1349of complex numbers (their sole argument).
1350
1351@item NON_LVALUE_EXPR
1352These nodes indicate that their one and only operand is not an lvalue.
1353A back end can treat these identically to the single operand.
1354
1355@item NOP_EXPR
1356These nodes are used to represent conversions that do not require any
1357code-generation.  For example, conversion of a @code{char*} to an
1358@code{int*} does not require any code be generated; such a conversion is
1359represented by a @code{NOP_EXPR}.  The single operand is the expression
1360to be converted.  The conversion from a pointer to a reference is also
1361represented with a @code{NOP_EXPR}.
1362
1363@item CONVERT_EXPR
1364These nodes are similar to @code{NOP_EXPR}s, but are used in those
1365situations where code may need to be generated.  For example, if an
1366@code{int*} is converted to an @code{int} code may need to be generated
1367on some platforms.  These nodes are never used for C++-specific
1368conversions, like conversions between pointers to different classes in
1369an inheritance hierarchy.  Any adjustments that need to be made in such
1370cases are always indicated explicitly.  Similarly, a user-defined
1371conversion is never represented by a @code{CONVERT_EXPR}; instead, the
1372function calls are made explicit.
1373
1374@item FIXED_CONVERT_EXPR
1375These nodes are used to represent conversions that involve fixed-point
1376values.  For example, from a fixed-point value to another fixed-point value,
1377from an integer to a fixed-point value, from a fixed-point value to an
1378integer, from a floating-point value to a fixed-point value, or from
1379a fixed-point value to a floating-point value.
1380
1381@item LSHIFT_EXPR
1382@itemx RSHIFT_EXPR
1383These nodes represent left and right shifts, respectively.  The first
1384operand is the value to shift; it will always be of integral type.  The
1385second operand is an expression for the number of bits by which to
1386shift.  Right shift should be treated as arithmetic, i.e., the
1387high-order bits should be zero-filled when the expression has unsigned
1388type and filled with the sign bit when the expression has signed type.
1389Note that the result is undefined if the second operand is larger
1390than or equal to the first operand's type size. Unlike most nodes, these
1391can have a vector as first operand and a scalar as second operand.
1392
1393
1394@item BIT_IOR_EXPR
1395@itemx BIT_XOR_EXPR
1396@itemx BIT_AND_EXPR
1397These nodes represent bitwise inclusive or, bitwise exclusive or, and
1398bitwise and, respectively.  Both operands will always have integral
1399type.
1400
1401@item TRUTH_ANDIF_EXPR
1402@itemx TRUTH_ORIF_EXPR
1403These nodes represent logical ``and'' and logical ``or'', respectively.
1404These operators are not strict; i.e., the second operand is evaluated
1405only if the value of the expression is not determined by evaluation of
1406the first operand.  The type of the operands and that of the result are
1407always of @code{BOOLEAN_TYPE} or @code{INTEGER_TYPE}.
1408
1409@item TRUTH_AND_EXPR
1410@itemx TRUTH_OR_EXPR
1411@itemx TRUTH_XOR_EXPR
1412These nodes represent logical and, logical or, and logical exclusive or.
1413They are strict; both arguments are always evaluated.  There are no
1414corresponding operators in C or C++, but the front end will sometimes
1415generate these expressions anyhow, if it can tell that strictness does
1416not matter.  The type of the operands and that of the result are
1417always of @code{BOOLEAN_TYPE} or @code{INTEGER_TYPE}.
1418
1419@item POINTER_PLUS_EXPR
1420This node represents pointer arithmetic.  The first operand is always
1421a pointer/reference type.  The second operand is always an unsigned
1422integer type compatible with sizetype.  This is the only binary
1423arithmetic operand that can operate on pointer types.
1424
1425@item PLUS_EXPR
1426@itemx MINUS_EXPR
1427@itemx MULT_EXPR
1428These nodes represent various binary arithmetic operations.
1429Respectively, these operations are addition, subtraction (of the second
1430operand from the first) and multiplication.  Their operands may have
1431either integral or floating type, but there will never be case in which
1432one operand is of floating type and the other is of integral type.
1433
1434The behavior of these operations on signed arithmetic overflow is
1435controlled by the @code{flag_wrapv} and @code{flag_trapv} variables.
1436
1437@item MULT_HIGHPART_EXPR
1438This node represents the ``high-part'' of a widening multiplication.
1439For an integral type with @var{b} bits of precision, the result is
1440the most significant @var{b} bits of the full @math{2@var{b}} product.
1441
1442@item RDIV_EXPR
1443This node represents a floating point division operation.
1444
1445@item TRUNC_DIV_EXPR
1446@itemx FLOOR_DIV_EXPR
1447@itemx CEIL_DIV_EXPR
1448@itemx ROUND_DIV_EXPR
1449These nodes represent integer division operations that return an integer
1450result.  @code{TRUNC_DIV_EXPR} rounds towards zero, @code{FLOOR_DIV_EXPR}
1451rounds towards negative infinity, @code{CEIL_DIV_EXPR} rounds towards
1452positive infinity and @code{ROUND_DIV_EXPR} rounds to the closest integer.
1453Integer division in C and C++ is truncating, i.e.@: @code{TRUNC_DIV_EXPR}.
1454
1455The behavior of these operations on signed arithmetic overflow, when
1456dividing the minimum signed integer by minus one, is controlled by the
1457@code{flag_wrapv} and @code{flag_trapv} variables.
1458
1459@item TRUNC_MOD_EXPR
1460@itemx FLOOR_MOD_EXPR
1461@itemx CEIL_MOD_EXPR
1462@itemx ROUND_MOD_EXPR
1463These nodes represent the integer remainder or modulus operation.
1464The integer modulus of two operands @code{a} and @code{b} is
1465defined as @code{a - (a/b)*b} where the division calculated using
1466the corresponding division operator.  Hence for @code{TRUNC_MOD_EXPR}
1467this definition assumes division using truncation towards zero, i.e.@:
1468@code{TRUNC_DIV_EXPR}.  Integer remainder in C and C++ uses truncating
1469division, i.e.@: @code{TRUNC_MOD_EXPR}.
1470
1471@item EXACT_DIV_EXPR
1472The @code{EXACT_DIV_EXPR} code is used to represent integer divisions where
1473the numerator is known to be an exact multiple of the denominator.  This
1474allows the backend to choose between the faster of @code{TRUNC_DIV_EXPR},
1475@code{CEIL_DIV_EXPR} and @code{FLOOR_DIV_EXPR} for the current target.
1476
1477@item LT_EXPR
1478@itemx LE_EXPR
1479@itemx GT_EXPR
1480@itemx GE_EXPR
1481@itemx EQ_EXPR
1482@itemx NE_EXPR
1483These nodes represent the less than, less than or equal to, greater
1484than, greater than or equal to, equal, and not equal comparison
1485operators.  The first and second operands will either be both of integral
1486type, both of floating type or both of vector type.  The result type of
1487these expressions will always be of integral, boolean or signed integral
1488vector type.  These operations return the result type's zero value for
1489false, the result type's one value for true, and a vector whose elements
1490are zero (false) or minus one (true) for vectors.
1491
1492For floating point comparisons, if we honor IEEE NaNs and either operand
1493is NaN, then @code{NE_EXPR} always returns true and the remaining operators
1494always return false.  On some targets, comparisons against an IEEE NaN,
1495other than equality and inequality, may generate a floating point exception.
1496
1497@item ORDERED_EXPR
1498@itemx UNORDERED_EXPR
1499These nodes represent non-trapping ordered and unordered comparison
1500operators.  These operations take two floating point operands and
1501determine whether they are ordered or unordered relative to each other.
1502If either operand is an IEEE NaN, their comparison is defined to be
1503unordered, otherwise the comparison is defined to be ordered.  The
1504result type of these expressions will always be of integral or boolean
1505type.  These operations return the result type's zero value for false,
1506and the result type's one value for true.
1507
1508@item UNLT_EXPR
1509@itemx UNLE_EXPR
1510@itemx UNGT_EXPR
1511@itemx UNGE_EXPR
1512@itemx UNEQ_EXPR
1513@itemx LTGT_EXPR
1514These nodes represent the unordered comparison operators.
1515These operations take two floating point operands and determine whether
1516the operands are unordered or are less than, less than or equal to,
1517greater than, greater than or equal to, or equal respectively.  For
1518example, @code{UNLT_EXPR} returns true if either operand is an IEEE
1519NaN or the first operand is less than the second.  With the possible
1520exception of @code{LTGT_EXPR}, all of these operations are guaranteed
1521not to generate a floating point exception.  The result
1522type of these expressions will always be of integral or boolean type.
1523These operations return the result type's zero value for false,
1524and the result type's one value for true.
1525
1526@item MODIFY_EXPR
1527These nodes represent assignment.  The left-hand side is the first
1528operand; the right-hand side is the second operand.  The left-hand side
1529will be a @code{VAR_DECL}, @code{INDIRECT_REF}, @code{COMPONENT_REF}, or
1530other lvalue.
1531
1532These nodes are used to represent not only assignment with @samp{=} but
1533also compound assignments (like @samp{+=}), by reduction to @samp{=}
1534assignment.  In other words, the representation for @samp{i += 3} looks
1535just like that for @samp{i = i + 3}.
1536
1537@item INIT_EXPR
1538These nodes are just like @code{MODIFY_EXPR}, but are used only when a
1539variable is initialized, rather than assigned to subsequently.  This
1540means that we can assume that the target of the initialization is not
1541used in computing its own value; any reference to the lhs in computing
1542the rhs is undefined.
1543
1544@item COMPOUND_EXPR
1545These nodes represent comma-expressions.  The first operand is an
1546expression whose value is computed and thrown away prior to the
1547evaluation of the second operand.  The value of the entire expression is
1548the value of the second operand.
1549
1550@item COND_EXPR
1551These nodes represent @code{?:} expressions.  The first operand
1552is of boolean or integral type.  If it evaluates to a nonzero value,
1553the second operand should be evaluated, and returned as the value of the
1554expression.  Otherwise, the third operand is evaluated, and returned as
1555the value of the expression.
1556
1557The second operand must have the same type as the entire expression,
1558unless it unconditionally throws an exception or calls a noreturn
1559function, in which case it should have void type.  The same constraints
1560apply to the third operand.  This allows array bounds checks to be
1561represented conveniently as @code{(i >= 0 && i < 10) ? i : abort()}.
1562
1563As a GNU extension, the C language front-ends allow the second
1564operand of the @code{?:} operator may be omitted in the source.
1565For example, @code{x ? : 3} is equivalent to @code{x ? x : 3},
1566assuming that @code{x} is an expression without side-effects.
1567In the tree representation, however, the second operand is always
1568present, possibly protected by @code{SAVE_EXPR} if the first
1569argument does cause side-effects.
1570
1571@item CALL_EXPR
1572These nodes are used to represent calls to functions, including
1573non-static member functions.  @code{CALL_EXPR}s are implemented as
1574expression nodes with a variable number of operands.  Rather than using
1575@code{TREE_OPERAND} to extract them, it is preferable to use the
1576specialized accessor macros and functions that operate specifically on
1577@code{CALL_EXPR} nodes.
1578
1579@code{CALL_EXPR_FN} returns a pointer to the
1580function to call; it is always an expression whose type is a
1581@code{POINTER_TYPE}.
1582
1583The number of arguments to the call is returned by @code{call_expr_nargs},
1584while the arguments themselves can be accessed with the @code{CALL_EXPR_ARG}
1585macro.  The arguments are zero-indexed and numbered left-to-right.
1586You can iterate over the arguments using @code{FOR_EACH_CALL_EXPR_ARG}, as in:
1587
1588@smallexample
1589tree call, arg;
1590call_expr_arg_iterator iter;
1591FOR_EACH_CALL_EXPR_ARG (arg, iter, call)
1592  /* arg is bound to successive arguments of call.  */
1593  @dots{};
1594@end smallexample
1595
1596For non-static
1597member functions, there will be an operand corresponding to the
1598@code{this} pointer.  There will always be expressions corresponding to
1599all of the arguments, even if the function is declared with default
1600arguments and some arguments are not explicitly provided at the call
1601sites.
1602
1603@code{CALL_EXPR}s also have a @code{CALL_EXPR_STATIC_CHAIN} operand that
1604is used to implement nested functions.  This operand is otherwise null.
1605
1606@item CLEANUP_POINT_EXPR
1607These nodes represent full-expressions.  The single operand is an
1608expression to evaluate.  Any destructor calls engendered by the creation
1609of temporaries during the evaluation of that expression should be
1610performed immediately after the expression is evaluated.
1611
1612@item CONSTRUCTOR
1613These nodes represent the brace-enclosed initializers for a structure or
1614array.  The first operand is reserved for use by the back end.  The
1615second operand is a @code{TREE_LIST}.  If the @code{TREE_TYPE} of the
1616@code{CONSTRUCTOR} is a @code{RECORD_TYPE} or @code{UNION_TYPE}, then
1617the @code{TREE_PURPOSE} of each node in the @code{TREE_LIST} will be a
1618@code{FIELD_DECL} and the @code{TREE_VALUE} of each node will be the
1619expression used to initialize that field.
1620
1621If the @code{TREE_TYPE} of the @code{CONSTRUCTOR} is an
1622@code{ARRAY_TYPE}, then the @code{TREE_PURPOSE} of each element in the
1623@code{TREE_LIST} will be an @code{INTEGER_CST} or a @code{RANGE_EXPR} of
1624two @code{INTEGER_CST}s.  A single @code{INTEGER_CST} indicates which
1625element of the array (indexed from zero) is being assigned to.  A
1626@code{RANGE_EXPR} indicates an inclusive range of elements to
1627initialize.  In both cases the @code{TREE_VALUE} is the corresponding
1628initializer.  It is re-evaluated for each element of a
1629@code{RANGE_EXPR}.  If the @code{TREE_PURPOSE} is @code{NULL_TREE}, then
1630the initializer is for the next available array element.
1631
1632In the front end, you should not depend on the fields appearing in any
1633particular order.  However, in the middle end, fields must appear in
1634declaration order.  You should not assume that all fields will be
1635represented.  Unrepresented fields will be set to zero.
1636
1637@item COMPOUND_LITERAL_EXPR
1638@findex COMPOUND_LITERAL_EXPR_DECL_EXPR
1639@findex COMPOUND_LITERAL_EXPR_DECL
1640These nodes represent ISO C99 compound literals.  The
1641@code{COMPOUND_LITERAL_EXPR_DECL_EXPR} is a @code{DECL_EXPR}
1642containing an anonymous @code{VAR_DECL} for
1643the unnamed object represented by the compound literal; the
1644@code{DECL_INITIAL} of that @code{VAR_DECL} is a @code{CONSTRUCTOR}
1645representing the brace-enclosed list of initializers in the compound
1646literal.  That anonymous @code{VAR_DECL} can also be accessed directly
1647by the @code{COMPOUND_LITERAL_EXPR_DECL} macro.
1648
1649@item SAVE_EXPR
1650
1651A @code{SAVE_EXPR} represents an expression (possibly involving
1652side-effects) that is used more than once.  The side-effects should
1653occur only the first time the expression is evaluated.  Subsequent uses
1654should just reuse the computed value.  The first operand to the
1655@code{SAVE_EXPR} is the expression to evaluate.  The side-effects should
1656be executed where the @code{SAVE_EXPR} is first encountered in a
1657depth-first preorder traversal of the expression tree.
1658
1659@item TARGET_EXPR
1660A @code{TARGET_EXPR} represents a temporary object.  The first operand
1661is a @code{VAR_DECL} for the temporary variable.  The second operand is
1662the initializer for the temporary.  The initializer is evaluated and,
1663if non-void, copied (bitwise) into the temporary.  If the initializer
1664is void, that means that it will perform the initialization itself.
1665
1666Often, a @code{TARGET_EXPR} occurs on the right-hand side of an
1667assignment, or as the second operand to a comma-expression which is
1668itself the right-hand side of an assignment, etc.  In this case, we say
1669that the @code{TARGET_EXPR} is ``normal''; otherwise, we say it is
1670``orphaned''.  For a normal @code{TARGET_EXPR} the temporary variable
1671should be treated as an alias for the left-hand side of the assignment,
1672rather than as a new temporary variable.
1673
1674The third operand to the @code{TARGET_EXPR}, if present, is a
1675cleanup-expression (i.e., destructor call) for the temporary.  If this
1676expression is orphaned, then this expression must be executed when the
1677statement containing this expression is complete.  These cleanups must
1678always be executed in the order opposite to that in which they were
1679encountered.  Note that if a temporary is created on one branch of a
1680conditional operator (i.e., in the second or third operand to a
1681@code{COND_EXPR}), the cleanup must be run only if that branch is
1682actually executed.
1683
1684@item VA_ARG_EXPR
1685This node is used to implement support for the C/C++ variable argument-list
1686mechanism.  It represents expressions like @code{va_arg (ap, type)}.
1687Its @code{TREE_TYPE} yields the tree representation for @code{type} and
1688its sole argument yields the representation for @code{ap}.
1689
1690@end table
1691
1692@node Vectors
1693@subsection Vectors
1694@tindex VEC_LSHIFT_EXPR
1695@tindex VEC_RSHIFT_EXPR
1696@tindex VEC_WIDEN_MULT_HI_EXPR
1697@tindex VEC_WIDEN_MULT_LO_EXPR
1698@tindex VEC_UNPACK_HI_EXPR
1699@tindex VEC_UNPACK_LO_EXPR
1700@tindex VEC_UNPACK_FLOAT_HI_EXPR
1701@tindex VEC_UNPACK_FLOAT_LO_EXPR
1702@tindex VEC_PACK_TRUNC_EXPR
1703@tindex VEC_PACK_SAT_EXPR
1704@tindex VEC_PACK_FIX_TRUNC_EXPR
1705
1706@table @code
1707@item VEC_LSHIFT_EXPR
1708@itemx VEC_RSHIFT_EXPR
1709These nodes represent whole vector left and right shifts, respectively.
1710The first operand is the vector to shift; it will always be of vector type.
1711The second operand is an expression for the number of bits by which to
1712shift.  Note that the result is undefined if the second operand is larger
1713than or equal to the first operand's type size.
1714
1715@item VEC_WIDEN_MULT_HI_EXPR
1716@itemx VEC_WIDEN_MULT_LO_EXPR
1717These nodes represent widening vector multiplication of the high and low
1718parts of the two input vectors, respectively.  Their operands are vectors
1719that contain the same number of elements (@code{N}) of the same integral type.
1720The result is a vector that contains half as many elements, of an integral type
1721whose size is twice as wide.  In the case of @code{VEC_WIDEN_MULT_HI_EXPR} the
1722high @code{N/2} elements of the two vector are multiplied to produce the
1723vector of @code{N/2} products. In the case of @code{VEC_WIDEN_MULT_LO_EXPR} the
1724low @code{N/2} elements of the two vector are multiplied to produce the
1725vector of @code{N/2} products.
1726
1727@item VEC_UNPACK_HI_EXPR
1728@itemx VEC_UNPACK_LO_EXPR
1729These nodes represent unpacking of the high and low parts of the input vector,
1730respectively.  The single operand is a vector that contains @code{N} elements
1731of the same integral or floating point type.  The result is a vector
1732that contains half as many elements, of an integral or floating point type
1733whose size is twice as wide.  In the case of @code{VEC_UNPACK_HI_EXPR} the
1734high @code{N/2} elements of the vector are extracted and widened (promoted).
1735In the case of @code{VEC_UNPACK_LO_EXPR} the low @code{N/2} elements of the
1736vector are extracted and widened (promoted).
1737
1738@item VEC_UNPACK_FLOAT_HI_EXPR
1739@itemx VEC_UNPACK_FLOAT_LO_EXPR
1740These nodes represent unpacking of the high and low parts of the input vector,
1741where the values are converted from fixed point to floating point.  The
1742single operand is a vector that contains @code{N} elements of the same
1743integral type.  The result is a vector that contains half as many elements
1744of a floating point type whose size is twice as wide.  In the case of
1745@code{VEC_UNPACK_HI_EXPR} the high @code{N/2} elements of the vector are
1746extracted, converted and widened.  In the case of @code{VEC_UNPACK_LO_EXPR}
1747the low @code{N/2} elements of the vector are extracted, converted and widened.
1748
1749@item VEC_PACK_TRUNC_EXPR
1750This node represents packing of truncated elements of the two input vectors
1751into the output vector.  Input operands are vectors that contain the same
1752number of elements of the same integral or floating point type.  The result
1753is a vector that contains twice as many elements of an integral or floating
1754point type whose size is half as wide. The elements of the two vectors are
1755demoted and merged (concatenated) to form the output vector.
1756
1757@item VEC_PACK_SAT_EXPR
1758This node represents packing of elements of the two input vectors into the
1759output vector using saturation.  Input operands are vectors that contain
1760the same number of elements of the same integral type.  The result is a
1761vector that contains twice as many elements of an integral type whose size
1762is half as wide.  The elements of the two vectors are demoted and merged
1763(concatenated) to form the output vector.
1764
1765@item VEC_PACK_FIX_TRUNC_EXPR
1766This node represents packing of elements of the two input vectors into the
1767output vector, where the values are converted from floating point
1768to fixed point.  Input operands are vectors that contain the same number
1769of elements of a floating point type.  The result is a vector that contains
1770twice as many elements of an integral type whose size is half as wide.  The
1771elements of the two vectors are merged (concatenated) to form the output
1772vector.
1773
1774@item VEC_COND_EXPR
1775These nodes represent @code{?:} expressions.  The three operands must be
1776vectors of the same size and number of elements.  The second and third
1777operands must have the same type as the entire expression.  The first
1778operand is of signed integral vector type.  If an element of the first
1779operand evaluates to a zero value, the corresponding element of the
1780result is taken from the third operand. If it evaluates to a minus one
1781value, it is taken from the second operand. It should never evaluate to
1782any other value currently, but optimizations should not rely on that
1783property. In contrast with a @code{COND_EXPR}, all operands are always
1784evaluated.
1785@end table
1786
1787
1788@c ---------------------------------------------------------------------
1789@c Statements
1790@c ---------------------------------------------------------------------
1791
1792@node Statements
1793@section Statements
1794@cindex Statements
1795
1796Most statements in GIMPLE are assignment statements, represented by
1797@code{GIMPLE_ASSIGN}.  No other C expressions can appear at statement level;
1798a reference to a volatile object is converted into a
1799@code{GIMPLE_ASSIGN}.
1800
1801There are also several varieties of complex statements.
1802
1803@menu
1804* Basic Statements::
1805* Blocks::
1806* Statement Sequences::
1807* Empty Statements::
1808* Jumps::
1809* Cleanups::
1810* OpenMP::
1811@end menu
1812
1813@node Basic Statements
1814@subsection Basic Statements
1815@cindex Basic Statements
1816
1817@table @code
1818@item ASM_EXPR
1819
1820Used to represent an inline assembly statement.  For an inline assembly
1821statement like:
1822@smallexample
1823asm ("mov x, y");
1824@end smallexample
1825The @code{ASM_STRING} macro will return a @code{STRING_CST} node for
1826@code{"mov x, y"}.  If the original statement made use of the
1827extended-assembly syntax, then @code{ASM_OUTPUTS},
1828@code{ASM_INPUTS}, and @code{ASM_CLOBBERS} will be the outputs, inputs,
1829and clobbers for the statement, represented as @code{STRING_CST} nodes.
1830The extended-assembly syntax looks like:
1831@smallexample
1832asm ("fsinx %1,%0" : "=f" (result) : "f" (angle));
1833@end smallexample
1834The first string is the @code{ASM_STRING}, containing the instruction
1835template.  The next two strings are the output and inputs, respectively;
1836this statement has no clobbers.  As this example indicates, ``plain''
1837assembly statements are merely a special case of extended assembly
1838statements; they have no cv-qualifiers, outputs, inputs, or clobbers.
1839All of the strings will be @code{NUL}-terminated, and will contain no
1840embedded @code{NUL}-characters.
1841
1842If the assembly statement is declared @code{volatile}, or if the
1843statement was not an extended assembly statement, and is therefore
1844implicitly volatile, then the predicate @code{ASM_VOLATILE_P} will hold
1845of the @code{ASM_EXPR}.
1846
1847@item DECL_EXPR
1848
1849Used to represent a local declaration.  The @code{DECL_EXPR_DECL} macro
1850can be used to obtain the entity declared.  This declaration may be a
1851@code{LABEL_DECL}, indicating that the label declared is a local label.
1852(As an extension, GCC allows the declaration of labels with scope.)  In
1853C, this declaration may be a @code{FUNCTION_DECL}, indicating the
1854use of the GCC nested function extension.  For more information,
1855@pxref{Functions}.
1856
1857@item LABEL_EXPR
1858
1859Used to represent a label.  The @code{LABEL_DECL} declared by this
1860statement can be obtained with the @code{LABEL_EXPR_LABEL} macro.  The
1861@code{IDENTIFIER_NODE} giving the name of the label can be obtained from
1862the @code{LABEL_DECL} with @code{DECL_NAME}.
1863
1864@item GOTO_EXPR
1865
1866Used to represent a @code{goto} statement.  The @code{GOTO_DESTINATION} will
1867usually be a @code{LABEL_DECL}.  However, if the ``computed goto'' extension
1868has been used, the @code{GOTO_DESTINATION} will be an arbitrary expression
1869indicating the destination.  This expression will always have pointer type.
1870
1871@item RETURN_EXPR
1872
1873Used to represent a @code{return} statement.  Operand 0 represents the
1874value to return.  It should either be the @code{RESULT_DECL} for the
1875containing function, or a @code{MODIFY_EXPR} or @code{INIT_EXPR}
1876setting the function's @code{RESULT_DECL}.  It will be
1877@code{NULL_TREE} if the statement was just
1878@smallexample
1879return;
1880@end smallexample
1881
1882@item LOOP_EXPR
1883These nodes represent ``infinite'' loops.  The @code{LOOP_EXPR_BODY}
1884represents the body of the loop.  It should be executed forever, unless
1885an @code{EXIT_EXPR} is encountered.
1886
1887@item EXIT_EXPR
1888These nodes represent conditional exits from the nearest enclosing
1889@code{LOOP_EXPR}.  The single operand is the condition; if it is
1890nonzero, then the loop should be exited.  An @code{EXIT_EXPR} will only
1891appear within a @code{LOOP_EXPR}.
1892
1893@item SWITCH_STMT
1894
1895Used to represent a @code{switch} statement.  The @code{SWITCH_STMT_COND}
1896is the expression on which the switch is occurring.  See the documentation
1897for an @code{IF_STMT} for more information on the representation used
1898for the condition.  The @code{SWITCH_STMT_BODY} is the body of the switch
1899statement.   The @code{SWITCH_STMT_TYPE} is the original type of switch
1900expression as given in the source, before any compiler conversions.
1901
1902@item CASE_LABEL_EXPR
1903
1904Use to represent a @code{case} label, range of @code{case} labels, or a
1905@code{default} label.  If @code{CASE_LOW} is @code{NULL_TREE}, then this is a
1906@code{default} label.  Otherwise, if @code{CASE_HIGH} is @code{NULL_TREE}, then
1907this is an ordinary @code{case} label.  In this case, @code{CASE_LOW} is
1908an expression giving the value of the label.  Both @code{CASE_LOW} and
1909@code{CASE_HIGH} are @code{INTEGER_CST} nodes.  These values will have
1910the same type as the condition expression in the switch statement.
1911
1912Otherwise, if both @code{CASE_LOW} and @code{CASE_HIGH} are defined, the
1913statement is a range of case labels.  Such statements originate with the
1914extension that allows users to write things of the form:
1915@smallexample
1916case 2 ... 5:
1917@end smallexample
1918The first value will be @code{CASE_LOW}, while the second will be
1919@code{CASE_HIGH}.
1920
1921@end table
1922
1923
1924@node Blocks
1925@subsection Blocks
1926@cindex Blocks
1927
1928Block scopes and the variables they declare in GENERIC are
1929expressed using the @code{BIND_EXPR} code, which in previous
1930versions of GCC was primarily used for the C statement-expression
1931extension.
1932
1933Variables in a block are collected into @code{BIND_EXPR_VARS} in
1934declaration order through their @code{TREE_CHAIN} field.  Any runtime
1935initialization is moved out of @code{DECL_INITIAL} and into a
1936statement in the controlled block.  When gimplifying from C or C++,
1937this initialization replaces the @code{DECL_STMT}.  These variables
1938will never require cleanups.  The scope of these variables is just the
1939body
1940
1941Variable-length arrays (VLAs) complicate this process, as their
1942size often refers to variables initialized earlier in the block.
1943To handle this, we currently split the block at that point, and
1944move the VLA into a new, inner @code{BIND_EXPR}.  This strategy
1945may change in the future.
1946
1947A C++ program will usually contain more @code{BIND_EXPR}s than
1948there are syntactic blocks in the source code, since several C++
1949constructs have implicit scopes associated with them.  On the
1950other hand, although the C++ front end uses pseudo-scopes to
1951handle cleanups for objects with destructors, these don't
1952translate into the GIMPLE form; multiple declarations at the same
1953level use the same @code{BIND_EXPR}.
1954
1955@node Statement Sequences
1956@subsection Statement Sequences
1957@cindex Statement Sequences
1958
1959Multiple statements at the same nesting level are collected into
1960a @code{STATEMENT_LIST}.  Statement lists are modified and
1961traversed using the interface in @samp{tree-iterator.h}.
1962
1963@node Empty Statements
1964@subsection Empty Statements
1965@cindex Empty Statements
1966
1967Whenever possible, statements with no effect are discarded.  But
1968if they are nested within another construct which cannot be
1969discarded for some reason, they are instead replaced with an
1970empty statement, generated by @code{build_empty_stmt}.
1971Initially, all empty statements were shared, after the pattern of
1972the Java front end, but this caused a lot of trouble in practice.
1973
1974An empty statement is represented as @code{(void)0}.
1975
1976@node Jumps
1977@subsection Jumps
1978@cindex Jumps
1979
1980Other jumps are expressed by either @code{GOTO_EXPR} or
1981@code{RETURN_EXPR}.
1982
1983The operand of a @code{GOTO_EXPR} must be either a label or a
1984variable containing the address to jump to.
1985
1986The operand of a @code{RETURN_EXPR} is either @code{NULL_TREE},
1987@code{RESULT_DECL}, or a @code{MODIFY_EXPR} which sets the return
1988value.  It would be nice to move the @code{MODIFY_EXPR} into a
1989separate statement, but the special return semantics in
1990@code{expand_return} make that difficult.  It may still happen in
1991the future, perhaps by moving most of that logic into
1992@code{expand_assignment}.
1993
1994@node Cleanups
1995@subsection Cleanups
1996@cindex Cleanups
1997
1998Destructors for local C++ objects and similar dynamic cleanups are
1999represented in GIMPLE by a @code{TRY_FINALLY_EXPR}.
2000@code{TRY_FINALLY_EXPR} has two operands, both of which are a sequence
2001of statements to execute.  The first sequence is executed.  When it
2002completes the second sequence is executed.
2003
2004The first sequence may complete in the following ways:
2005
2006@enumerate
2007
2008@item Execute the last statement in the sequence and fall off the
2009end.
2010
2011@item Execute a goto statement (@code{GOTO_EXPR}) to an ordinary
2012label outside the sequence.
2013
2014@item Execute a return statement (@code{RETURN_EXPR}).
2015
2016@item Throw an exception.  This is currently not explicitly represented in
2017GIMPLE.
2018
2019@end enumerate
2020
2021The second sequence is not executed if the first sequence completes by
2022calling @code{setjmp} or @code{exit} or any other function that does
2023not return.  The second sequence is also not executed if the first
2024sequence completes via a non-local goto or a computed goto (in general
2025the compiler does not know whether such a goto statement exits the
2026first sequence or not, so we assume that it doesn't).
2027
2028After the second sequence is executed, if it completes normally by
2029falling off the end, execution continues wherever the first sequence
2030would have continued, by falling off the end, or doing a goto, etc.
2031
2032@code{TRY_FINALLY_EXPR} complicates the flow graph, since the cleanup
2033needs to appear on every edge out of the controlled block; this
2034reduces the freedom to move code across these edges.  Therefore, the
2035EH lowering pass which runs before most of the optimization passes
2036eliminates these expressions by explicitly adding the cleanup to each
2037edge.  Rethrowing the exception is represented using @code{RESX_EXPR}.
2038
2039@node OpenMP
2040@subsection OpenMP
2041@tindex OMP_PARALLEL
2042@tindex OMP_FOR
2043@tindex OMP_SECTIONS
2044@tindex OMP_SINGLE
2045@tindex OMP_SECTION
2046@tindex OMP_MASTER
2047@tindex OMP_ORDERED
2048@tindex OMP_CRITICAL
2049@tindex OMP_RETURN
2050@tindex OMP_CONTINUE
2051@tindex OMP_ATOMIC
2052@tindex OMP_CLAUSE
2053
2054All the statements starting with @code{OMP_} represent directives and
2055clauses used by the OpenMP API @w{@uref{http://www.openmp.org/}}.
2056
2057@table @code
2058@item OMP_PARALLEL
2059
2060Represents @code{#pragma omp parallel [clause1 @dots{} clauseN]}. It
2061has four operands:
2062
2063Operand @code{OMP_PARALLEL_BODY} is valid while in GENERIC and
2064High GIMPLE forms.  It contains the body of code to be executed
2065by all the threads.  During GIMPLE lowering, this operand becomes
2066@code{NULL} and the body is emitted linearly after
2067@code{OMP_PARALLEL}.
2068
2069Operand @code{OMP_PARALLEL_CLAUSES} is the list of clauses
2070associated with the directive.
2071
2072Operand @code{OMP_PARALLEL_FN} is created by
2073@code{pass_lower_omp}, it contains the @code{FUNCTION_DECL}
2074for the function that will contain the body of the parallel
2075region.
2076
2077Operand @code{OMP_PARALLEL_DATA_ARG} is also created by
2078@code{pass_lower_omp}. If there are shared variables to be
2079communicated to the children threads, this operand will contain
2080the @code{VAR_DECL} that contains all the shared values and
2081variables.
2082
2083@item OMP_FOR
2084
2085Represents @code{#pragma omp for [clause1 @dots{} clauseN]}.  It
2086has 5 operands:
2087
2088Operand @code{OMP_FOR_BODY} contains the loop body.
2089
2090Operand @code{OMP_FOR_CLAUSES} is the list of clauses
2091associated with the directive.
2092
2093Operand @code{OMP_FOR_INIT} is the loop initialization code of
2094the form @code{VAR = N1}.
2095
2096Operand @code{OMP_FOR_COND} is the loop conditional expression
2097of the form @code{VAR @{<,>,<=,>=@} N2}.
2098
2099Operand @code{OMP_FOR_INCR} is the loop index increment of the
2100form @code{VAR @{+=,-=@} INCR}.
2101
2102Operand @code{OMP_FOR_PRE_BODY} contains side-effect code from
2103operands @code{OMP_FOR_INIT}, @code{OMP_FOR_COND} and
2104@code{OMP_FOR_INC}.  These side-effects are part of the
2105@code{OMP_FOR} block but must be evaluated before the start of
2106loop body.
2107
2108The loop index variable @code{VAR} must be a signed integer variable,
2109which is implicitly private to each thread.  Bounds
2110@code{N1} and @code{N2} and the increment expression
2111@code{INCR} are required to be loop invariant integer
2112expressions that are evaluated without any synchronization. The
2113evaluation order, frequency of evaluation and side-effects are
2114unspecified by the standard.
2115
2116@item OMP_SECTIONS
2117
2118Represents @code{#pragma omp sections [clause1 @dots{} clauseN]}.
2119
2120Operand @code{OMP_SECTIONS_BODY} contains the sections body,
2121which in turn contains a set of @code{OMP_SECTION} nodes for
2122each of the concurrent sections delimited by @code{#pragma omp
2123section}.
2124
2125Operand @code{OMP_SECTIONS_CLAUSES} is the list of clauses
2126associated with the directive.
2127
2128@item OMP_SECTION
2129
2130Section delimiter for @code{OMP_SECTIONS}.
2131
2132@item OMP_SINGLE
2133
2134Represents @code{#pragma omp single}.
2135
2136Operand @code{OMP_SINGLE_BODY} contains the body of code to be
2137executed by a single thread.
2138
2139Operand @code{OMP_SINGLE_CLAUSES} is the list of clauses
2140associated with the directive.
2141
2142@item OMP_MASTER
2143
2144Represents @code{#pragma omp master}.
2145
2146Operand @code{OMP_MASTER_BODY} contains the body of code to be
2147executed by the master thread.
2148
2149@item OMP_ORDERED
2150
2151Represents @code{#pragma omp ordered}.
2152
2153Operand @code{OMP_ORDERED_BODY} contains the body of code to be
2154executed in the sequential order dictated by the loop index
2155variable.
2156
2157@item OMP_CRITICAL
2158
2159Represents @code{#pragma omp critical [name]}.
2160
2161Operand @code{OMP_CRITICAL_BODY} is the critical section.
2162
2163Operand @code{OMP_CRITICAL_NAME} is an optional identifier to
2164label the critical section.
2165
2166@item OMP_RETURN
2167
2168This does not represent any OpenMP directive, it is an artificial
2169marker to indicate the end of the body of an OpenMP@. It is used
2170by the flow graph (@code{tree-cfg.c}) and OpenMP region
2171building code (@code{omp-low.c}).
2172
2173@item OMP_CONTINUE
2174
2175Similarly, this instruction does not represent an OpenMP
2176directive, it is used by @code{OMP_FOR} and
2177@code{OMP_SECTIONS} to mark the place where the code needs to
2178loop to the next iteration (in the case of @code{OMP_FOR}) or
2179the next section (in the case of @code{OMP_SECTIONS}).
2180
2181In some cases, @code{OMP_CONTINUE} is placed right before
2182@code{OMP_RETURN}.  But if there are cleanups that need to
2183occur right after the looping body, it will be emitted between
2184@code{OMP_CONTINUE} and @code{OMP_RETURN}.
2185
2186@item OMP_ATOMIC
2187
2188Represents @code{#pragma omp atomic}.
2189
2190Operand 0 is the address at which the atomic operation is to be
2191performed.
2192
2193Operand 1 is the expression to evaluate.  The gimplifier tries
2194three alternative code generation strategies.  Whenever possible,
2195an atomic update built-in is used.  If that fails, a
2196compare-and-swap loop is attempted.  If that also fails, a
2197regular critical section around the expression is used.
2198
2199@item OMP_CLAUSE
2200
2201Represents clauses associated with one of the @code{OMP_} directives.
2202Clauses are represented by separate sub-codes defined in
2203@file{tree.h}.  Clauses codes can be one of:
2204@code{OMP_CLAUSE_PRIVATE}, @code{OMP_CLAUSE_SHARED},
2205@code{OMP_CLAUSE_FIRSTPRIVATE},
2206@code{OMP_CLAUSE_LASTPRIVATE}, @code{OMP_CLAUSE_COPYIN},
2207@code{OMP_CLAUSE_COPYPRIVATE}, @code{OMP_CLAUSE_IF},
2208@code{OMP_CLAUSE_NUM_THREADS}, @code{OMP_CLAUSE_SCHEDULE},
2209@code{OMP_CLAUSE_NOWAIT}, @code{OMP_CLAUSE_ORDERED},
2210@code{OMP_CLAUSE_DEFAULT}, @code{OMP_CLAUSE_REDUCTION},
2211@code{OMP_CLAUSE_COLLAPSE}, @code{OMP_CLAUSE_UNTIED},
2212@code{OMP_CLAUSE_FINAL}, and @code{OMP_CLAUSE_MERGEABLE}.  Each code
2213represents the corresponding OpenMP clause.
2214
2215Clauses associated with the same directive are chained together
2216via @code{OMP_CLAUSE_CHAIN}. Those clauses that accept a list
2217of variables are restricted to exactly one, accessed with
2218@code{OMP_CLAUSE_VAR}.  Therefore, multiple variables under the
2219same clause @code{C} need to be represented as multiple @code{C} clauses
2220chained together.  This facilitates adding new clauses during
2221compilation.
2222
2223@end table
2224
2225@c ---------------------------------------------------------------------
2226@c Functions
2227@c ---------------------------------------------------------------------
2228
2229@node Functions
2230@section Functions
2231@cindex function
2232@tindex FUNCTION_DECL
2233
2234A function is represented by a @code{FUNCTION_DECL} node.  It stores
2235the basic pieces of the function such as body, parameters, and return
2236type as well as information on the surrounding context, visibility,
2237and linkage.
2238
2239@menu
2240* Function Basics::     Function names, body, and parameters.
2241* Function Properties:: Context, linkage, etc.
2242@end menu
2243
2244@c ---------------------------------------------------------------------
2245@c Function Basics
2246@c ---------------------------------------------------------------------
2247
2248@node Function Basics
2249@subsection Function Basics
2250@findex DECL_NAME
2251@findex DECL_ASSEMBLER_NAME
2252@findex TREE_PUBLIC
2253@findex DECL_ARTIFICIAL
2254@findex DECL_FUNCTION_SPECIFIC_TARGET
2255@findex DECL_FUNCTION_SPECIFIC_OPTIMIZATION
2256
2257A function has four core parts: the name, the parameters, the result,
2258and the body.  The following macros and functions access these parts
2259of a @code{FUNCTION_DECL} as well as other basic features:
2260@ftable @code
2261@item DECL_NAME
2262This macro returns the unqualified name of the function, as an
2263@code{IDENTIFIER_NODE}.  For an instantiation of a function template,
2264the @code{DECL_NAME} is the unqualified name of the template, not
2265something like @code{f<int>}.  The value of @code{DECL_NAME} is
2266undefined when used on a constructor, destructor, overloaded operator,
2267or type-conversion operator, or any function that is implicitly
2268generated by the compiler.  See below for macros that can be used to
2269distinguish these cases.
2270
2271@item DECL_ASSEMBLER_NAME
2272This macro returns the mangled name of the function, also an
2273@code{IDENTIFIER_NODE}.  This name does not contain leading underscores
2274on systems that prefix all identifiers with underscores.  The mangled
2275name is computed in the same way on all platforms; if special processing
2276is required to deal with the object file format used on a particular
2277platform, it is the responsibility of the back end to perform those
2278modifications.  (Of course, the back end should not modify
2279@code{DECL_ASSEMBLER_NAME} itself.)
2280
2281Using @code{DECL_ASSEMBLER_NAME} will cause additional memory to be
2282allocated (for the mangled name of the entity) so it should be used
2283only when emitting assembly code.  It should not be used within the
2284optimizers to determine whether or not two declarations are the same,
2285even though some of the existing optimizers do use it in that way.
2286These uses will be removed over time.
2287
2288@item DECL_ARGUMENTS
2289This macro returns the @code{PARM_DECL} for the first argument to the
2290function.  Subsequent @code{PARM_DECL} nodes can be obtained by
2291following the @code{TREE_CHAIN} links.
2292
2293@item DECL_RESULT
2294This macro returns the @code{RESULT_DECL} for the function.
2295
2296@item DECL_SAVED_TREE
2297This macro returns the complete body of the function.
2298
2299@item TREE_TYPE
2300This macro returns the @code{FUNCTION_TYPE} or @code{METHOD_TYPE} for
2301the function.
2302
2303@item DECL_INITIAL
2304A function that has a definition in the current translation unit will
2305have a non-@code{NULL} @code{DECL_INITIAL}.  However, back ends should not make
2306use of the particular value given by @code{DECL_INITIAL}.
2307
2308It should contain a tree of @code{BLOCK} nodes that mirrors the scopes
2309that variables are bound in the function.  Each block contains a list
2310of decls declared in a basic block, a pointer to a chain of blocks at
2311the next lower scope level, then a pointer to the next block at the
2312same level and a backpointer to the parent @code{BLOCK} or
2313@code{FUNCTION_DECL}.  So given a function as follows:
2314
2315@smallexample
2316void foo()
2317@{
2318  int a;
2319  @{
2320    int b;
2321  @}
2322  int c;
2323@}
2324@end smallexample
2325
2326you would get the following:
2327
2328@smallexample
2329tree foo = FUNCTION_DECL;
2330tree decl_a = VAR_DECL;
2331tree decl_b = VAR_DECL;
2332tree decl_c = VAR_DECL;
2333tree block_a = BLOCK;
2334tree block_b = BLOCK;
2335tree block_c = BLOCK;
2336BLOCK_VARS(block_a) = decl_a;
2337BLOCK_SUBBLOCKS(block_a) = block_b;
2338BLOCK_CHAIN(block_a) = block_c;
2339BLOCK_SUPERCONTEXT(block_a) = foo;
2340BLOCK_VARS(block_b) = decl_b;
2341BLOCK_SUPERCONTEXT(block_b) = block_a;
2342BLOCK_VARS(block_c) = decl_c;
2343BLOCK_SUPERCONTEXT(block_c) = foo;
2344DECL_INITIAL(foo) = block_a;
2345@end smallexample
2346
2347@end ftable
2348
2349@c ---------------------------------------------------------------------
2350@c Function Properties
2351@c ---------------------------------------------------------------------
2352
2353@node Function Properties
2354@subsection Function Properties
2355@cindex function properties
2356@cindex statements
2357
2358To determine the scope of a function, you can use the
2359@code{DECL_CONTEXT} macro.  This macro will return the class
2360(either a @code{RECORD_TYPE} or a @code{UNION_TYPE}) or namespace (a
2361@code{NAMESPACE_DECL}) of which the function is a member.  For a virtual
2362function, this macro returns the class in which the function was
2363actually defined, not the base class in which the virtual declaration
2364occurred.
2365
2366In C, the @code{DECL_CONTEXT} for a function maybe another function.
2367This representation indicates that the GNU nested function extension
2368is in use.  For details on the semantics of nested functions, see the
2369GCC Manual.  The nested function can refer to local variables in its
2370containing function.  Such references are not explicitly marked in the
2371tree structure; back ends must look at the @code{DECL_CONTEXT} for the
2372referenced @code{VAR_DECL}.  If the @code{DECL_CONTEXT} for the
2373referenced @code{VAR_DECL} is not the same as the function currently
2374being processed, and neither @code{DECL_EXTERNAL} nor
2375@code{TREE_STATIC} hold, then the reference is to a local variable in
2376a containing function, and the back end must take appropriate action.
2377
2378@ftable @code
2379@item DECL_EXTERNAL
2380This predicate holds if the function is undefined.
2381
2382@item TREE_PUBLIC
2383This predicate holds if the function has external linkage.
2384
2385@item TREE_STATIC
2386This predicate holds if the function has been defined.
2387
2388@item TREE_THIS_VOLATILE
2389This predicate holds if the function does not return normally.
2390
2391@item TREE_READONLY
2392This predicate holds if the function can only read its arguments.
2393
2394@item DECL_PURE_P
2395This predicate holds if the function can only read its arguments, but
2396may also read global memory.
2397
2398@item DECL_VIRTUAL_P
2399This predicate holds if the function is virtual.
2400
2401@item DECL_ARTIFICIAL
2402This macro holds if the function was implicitly generated by the
2403compiler, rather than explicitly declared.  In addition to implicitly
2404generated class member functions, this macro holds for the special
2405functions created to implement static initialization and destruction, to
2406compute run-time type information, and so forth.
2407
2408@item DECL_FUNCTION_SPECIFIC_TARGET
2409This macro returns a tree node that holds the target options that are
2410to be used to compile this particular function or @code{NULL_TREE} if
2411the function is to be compiled with the target options specified on
2412the command line.
2413
2414@item DECL_FUNCTION_SPECIFIC_OPTIMIZATION
2415This macro returns a tree node that holds the optimization options
2416that are to be used to compile this particular function or
2417@code{NULL_TREE} if the function is to be compiled with the
2418optimization options specified on the command line.
2419
2420@end ftable
2421
2422@c ---------------------------------------------------------------------
2423@c Language-dependent trees
2424@c ---------------------------------------------------------------------
2425
2426@node Language-dependent trees
2427@section Language-dependent trees
2428@cindex language-dependent trees
2429
2430Front ends may wish to keep some state associated with various GENERIC
2431trees while parsing.  To support this, trees provide a set of flags
2432that may be used by the front end.  They are accessed using
2433@code{TREE_LANG_FLAG_n} where @samp{n} is currently 0 through 6.
2434
2435If necessary, a front end can use some language-dependent tree
2436codes in its GENERIC representation, so long as it provides a
2437hook for converting them to GIMPLE and doesn't expect them to
2438work with any (hypothetical) optimizers that run before the
2439conversion to GIMPLE@. The intermediate representation used while
2440parsing C and C++ looks very little like GENERIC, but the C and
2441C++ gimplifier hooks are perfectly happy to take it as input and
2442spit out GIMPLE@.
2443
2444
2445
2446@node C and C++ Trees
2447@section C and C++ Trees
2448
2449This section documents the internal representation used by GCC to
2450represent C and C++ source programs.  When presented with a C or C++
2451source program, GCC parses the program, performs semantic analysis
2452(including the generation of error messages), and then produces the
2453internal representation described here.  This representation contains a
2454complete representation for the entire translation unit provided as
2455input to the front end.  This representation is then typically processed
2456by a code-generator in order to produce machine code, but could also be
2457used in the creation of source browsers, intelligent editors, automatic
2458documentation generators, interpreters, and any other programs needing
2459the ability to process C or C++ code.
2460
2461This section explains the internal representation.  In particular, it
2462documents the internal representation for C and C++ source
2463constructs, and the macros, functions, and variables that can be used to
2464access these constructs.  The C++ representation is largely a superset
2465of the representation used in the C front end.  There is only one
2466construct used in C that does not appear in the C++ front end and that
2467is the GNU ``nested function'' extension.  Many of the macros documented
2468here do not apply in C because the corresponding language constructs do
2469not appear in C@.
2470
2471The C and C++ front ends generate a mix of GENERIC trees and ones
2472specific to C and C++.  These language-specific trees are higher-level
2473constructs than the ones in GENERIC to make the parser's job easier.
2474This section describes those trees that aren't part of GENERIC as well
2475as aspects of GENERIC trees that are treated in a language-specific
2476manner.
2477
2478If you are developing a ``back end'', be it is a code-generator or some
2479other tool, that uses this representation, you may occasionally find
2480that you need to ask questions not easily answered by the functions and
2481macros available here.  If that situation occurs, it is quite likely
2482that GCC already supports the functionality you desire, but that the
2483interface is simply not documented here.  In that case, you should ask
2484the GCC maintainers (via mail to @email{gcc@@gcc.gnu.org}) about
2485documenting the functionality you require.  Similarly, if you find
2486yourself writing functions that do not deal directly with your back end,
2487but instead might be useful to other people using the GCC front end, you
2488should submit your patches for inclusion in GCC@.
2489
2490@menu
2491* Types for C++::               Fundamental and aggregate types.
2492* Namespaces::                  Namespaces.
2493* Classes::                     Classes.
2494* Functions for C++::           Overloading and accessors for C++.
2495* Statements for C++::          Statements specific to C and C++.
2496* C++ Expressions::    From @code{typeid} to @code{throw}.
2497@end menu
2498
2499@node Types for C++
2500@subsection Types for C++
2501@tindex UNKNOWN_TYPE
2502@tindex TYPENAME_TYPE
2503@tindex TYPEOF_TYPE
2504@findex cp_type_quals
2505@findex TYPE_UNQUALIFIED
2506@findex TYPE_QUAL_CONST
2507@findex TYPE_QUAL_VOLATILE
2508@findex TYPE_QUAL_RESTRICT
2509@findex TYPE_MAIN_VARIANT
2510@cindex qualified type
2511@findex TYPE_SIZE
2512@findex TYPE_ALIGN
2513@findex TYPE_PRECISION
2514@findex TYPE_ARG_TYPES
2515@findex TYPE_METHOD_BASETYPE
2516@findex TYPE_PTRDATAMEM_P
2517@findex TYPE_OFFSET_BASETYPE
2518@findex TREE_TYPE
2519@findex TYPE_CONTEXT
2520@findex TYPE_NAME
2521@findex TYPENAME_TYPE_FULLNAME
2522@findex TYPE_FIELDS
2523@findex TYPE_PTROBV_P
2524
2525In C++, an array type is not qualified; rather the type of the array
2526elements is qualified.  This situation is reflected in the intermediate
2527representation.  The macros described here will always examine the
2528qualification of the underlying element type when applied to an array
2529type.  (If the element type is itself an array, then the recursion
2530continues until a non-array type is found, and the qualification of this
2531type is examined.)  So, for example, @code{CP_TYPE_CONST_P} will hold of
2532the type @code{const int ()[7]}, denoting an array of seven @code{int}s.
2533
2534The following functions and macros deal with cv-qualification of types:
2535@ftable @code
2536@item cp_type_quals
2537This function returns the set of type qualifiers applied to this type.
2538This value is @code{TYPE_UNQUALIFIED} if no qualifiers have been
2539applied.  The @code{TYPE_QUAL_CONST} bit is set if the type is
2540@code{const}-qualified.  The @code{TYPE_QUAL_VOLATILE} bit is set if the
2541type is @code{volatile}-qualified.  The @code{TYPE_QUAL_RESTRICT} bit is
2542set if the type is @code{restrict}-qualified.
2543
2544@item CP_TYPE_CONST_P
2545This macro holds if the type is @code{const}-qualified.
2546
2547@item CP_TYPE_VOLATILE_P
2548This macro holds if the type is @code{volatile}-qualified.
2549
2550@item CP_TYPE_RESTRICT_P
2551This macro holds if the type is @code{restrict}-qualified.
2552
2553@item CP_TYPE_CONST_NON_VOLATILE_P
2554This predicate holds for a type that is @code{const}-qualified, but
2555@emph{not} @code{volatile}-qualified; other cv-qualifiers are ignored as
2556well: only the @code{const}-ness is tested.
2557
2558@end ftable
2559
2560A few other macros and functions are usable with all types:
2561@ftable @code
2562@item TYPE_SIZE
2563The number of bits required to represent the type, represented as an
2564@code{INTEGER_CST}.  For an incomplete type, @code{TYPE_SIZE} will be
2565@code{NULL_TREE}.
2566
2567@item TYPE_ALIGN
2568The alignment of the type, in bits, represented as an @code{int}.
2569
2570@item TYPE_NAME
2571This macro returns a declaration (in the form of a @code{TYPE_DECL}) for
2572the type.  (Note this macro does @emph{not} return an
2573@code{IDENTIFIER_NODE}, as you might expect, given its name!)  You can
2574look at the @code{DECL_NAME} of the @code{TYPE_DECL} to obtain the
2575actual name of the type.  The @code{TYPE_NAME} will be @code{NULL_TREE}
2576for a type that is not a built-in type, the result of a typedef, or a
2577named class type.
2578
2579@item CP_INTEGRAL_TYPE
2580This predicate holds if the type is an integral type.  Notice that in
2581C++, enumerations are @emph{not} integral types.
2582
2583@item ARITHMETIC_TYPE_P
2584This predicate holds if the type is an integral type (in the C++ sense)
2585or a floating point type.
2586
2587@item CLASS_TYPE_P
2588This predicate holds for a class-type.
2589
2590@item TYPE_BUILT_IN
2591This predicate holds for a built-in type.
2592
2593@item TYPE_PTRDATAMEM_P
2594This predicate holds if the type is a pointer to data member.
2595
2596@item TYPE_PTR_P
2597This predicate holds if the type is a pointer type, and the pointee is
2598not a data member.
2599
2600@item TYPE_PTRFN_P
2601This predicate holds for a pointer to function type.
2602
2603@item TYPE_PTROB_P
2604This predicate holds for a pointer to object type.  Note however that it
2605does not hold for the generic pointer to object type @code{void *}.  You
2606may use @code{TYPE_PTROBV_P} to test for a pointer to object type as
2607well as @code{void *}.
2608
2609@end ftable
2610
2611The table below describes types specific to C and C++ as well as
2612language-dependent info about GENERIC types.
2613
2614@table @code
2615
2616@item POINTER_TYPE
2617Used to represent pointer types, and pointer to data member types.  If
2618@code{TREE_TYPE}
2619is a pointer to data member type, then @code{TYPE_PTRDATAMEM_P} will hold.
2620For a pointer to data member type of the form @samp{T X::*},
2621@code{TYPE_PTRMEM_CLASS_TYPE} will be the type @code{X}, while
2622@code{TYPE_PTRMEM_POINTED_TO_TYPE} will be the type @code{T}.
2623
2624@item RECORD_TYPE
2625Used to represent @code{struct} and @code{class} types in C and C++.  If
2626@code{TYPE_PTRMEMFUNC_P} holds, then this type is a pointer-to-member
2627type.  In that case, the @code{TYPE_PTRMEMFUNC_FN_TYPE} is a
2628@code{POINTER_TYPE} pointing to a @code{METHOD_TYPE}.  The
2629@code{METHOD_TYPE} is the type of a function pointed to by the
2630pointer-to-member function.  If @code{TYPE_PTRMEMFUNC_P} does not hold,
2631this type is a class type.  For more information, @pxref{Classes}.
2632
2633@item UNKNOWN_TYPE
2634This node is used to represent a type the knowledge of which is
2635insufficient for a sound processing.
2636
2637@item TYPENAME_TYPE
2638Used to represent a construct of the form @code{typename T::A}.  The
2639@code{TYPE_CONTEXT} is @code{T}; the @code{TYPE_NAME} is an
2640@code{IDENTIFIER_NODE} for @code{A}.  If the type is specified via a
2641template-id, then @code{TYPENAME_TYPE_FULLNAME} yields a
2642@code{TEMPLATE_ID_EXPR}.  The @code{TREE_TYPE} is non-@code{NULL} if the
2643node is implicitly generated in support for the implicit typename
2644extension; in which case the @code{TREE_TYPE} is a type node for the
2645base-class.
2646
2647@item TYPEOF_TYPE
2648Used to represent the @code{__typeof__} extension.  The
2649@code{TYPE_FIELDS} is the expression the type of which is being
2650represented.
2651
2652@end table
2653
2654
2655@c ---------------------------------------------------------------------
2656@c Namespaces
2657@c ---------------------------------------------------------------------
2658
2659@node Namespaces
2660@subsection Namespaces
2661@cindex namespace, scope
2662@tindex NAMESPACE_DECL
2663
2664The root of the entire intermediate representation is the variable
2665@code{global_namespace}.  This is the namespace specified with @code{::}
2666in C++ source code.  All other namespaces, types, variables, functions,
2667and so forth can be found starting with this namespace.
2668
2669However, except for the fact that it is distinguished as the root of the
2670representation, the global namespace is no different from any other
2671namespace.  Thus, in what follows, we describe namespaces generally,
2672rather than the global namespace in particular.
2673
2674A namespace is represented by a @code{NAMESPACE_DECL} node.
2675
2676The following macros and functions can be used on a @code{NAMESPACE_DECL}:
2677
2678@ftable @code
2679@item DECL_NAME
2680This macro is used to obtain the @code{IDENTIFIER_NODE} corresponding to
2681the unqualified name of the name of the namespace (@pxref{Identifiers}).
2682The name of the global namespace is @samp{::}, even though in C++ the
2683global namespace is unnamed.  However, you should use comparison with
2684@code{global_namespace}, rather than @code{DECL_NAME} to determine
2685whether or not a namespace is the global one.  An unnamed namespace
2686will have a @code{DECL_NAME} equal to @code{anonymous_namespace_name}.
2687Within a single translation unit, all unnamed namespaces will have the
2688same name.
2689
2690@item DECL_CONTEXT
2691This macro returns the enclosing namespace.  The @code{DECL_CONTEXT} for
2692the @code{global_namespace} is @code{NULL_TREE}.
2693
2694@item DECL_NAMESPACE_ALIAS
2695If this declaration is for a namespace alias, then
2696@code{DECL_NAMESPACE_ALIAS} is the namespace for which this one is an
2697alias.
2698
2699Do not attempt to use @code{cp_namespace_decls} for a namespace which is
2700an alias.  Instead, follow @code{DECL_NAMESPACE_ALIAS} links until you
2701reach an ordinary, non-alias, namespace, and call
2702@code{cp_namespace_decls} there.
2703
2704@item DECL_NAMESPACE_STD_P
2705This predicate holds if the namespace is the special @code{::std}
2706namespace.
2707
2708@item cp_namespace_decls
2709This function will return the declarations contained in the namespace,
2710including types, overloaded functions, other namespaces, and so forth.
2711If there are no declarations, this function will return
2712@code{NULL_TREE}.  The declarations are connected through their
2713@code{TREE_CHAIN} fields.
2714
2715Although most entries on this list will be declarations,
2716@code{TREE_LIST} nodes may also appear.  In this case, the
2717@code{TREE_VALUE} will be an @code{OVERLOAD}.  The value of the
2718@code{TREE_PURPOSE} is unspecified; back ends should ignore this value.
2719As with the other kinds of declarations returned by
2720@code{cp_namespace_decls}, the @code{TREE_CHAIN} will point to the next
2721declaration in this list.
2722
2723For more information on the kinds of declarations that can occur on this
2724list, @xref{Declarations}.  Some declarations will not appear on this
2725list.  In particular, no @code{FIELD_DECL}, @code{LABEL_DECL}, or
2726@code{PARM_DECL} nodes will appear here.
2727
2728This function cannot be used with namespaces that have
2729@code{DECL_NAMESPACE_ALIAS} set.
2730
2731@end ftable
2732
2733@c ---------------------------------------------------------------------
2734@c Classes
2735@c ---------------------------------------------------------------------
2736
2737@node Classes
2738@subsection Classes
2739@cindex class, scope
2740@tindex RECORD_TYPE
2741@tindex UNION_TYPE
2742@findex CLASSTYPE_DECLARED_CLASS
2743@findex TYPE_BINFO
2744@findex BINFO_TYPE
2745@findex TYPE_FIELDS
2746@findex TYPE_VFIELD
2747@findex TYPE_METHODS
2748
2749Besides namespaces, the other high-level scoping construct in C++ is the
2750class.  (Throughout this manual the term @dfn{class} is used to mean the
2751types referred to in the ANSI/ISO C++ Standard as classes; these include
2752types defined with the @code{class}, @code{struct}, and @code{union}
2753keywords.)
2754
2755A class type is represented by either a @code{RECORD_TYPE} or a
2756@code{UNION_TYPE}.  A class declared with the @code{union} tag is
2757represented by a @code{UNION_TYPE}, while classes declared with either
2758the @code{struct} or the @code{class} tag are represented by
2759@code{RECORD_TYPE}s.  You can use the @code{CLASSTYPE_DECLARED_CLASS}
2760macro to discern whether or not a particular type is a @code{class} as
2761opposed to a @code{struct}.  This macro will be true only for classes
2762declared with the @code{class} tag.
2763
2764Almost all non-function members are available on the @code{TYPE_FIELDS}
2765list.  Given one member, the next can be found by following the
2766@code{TREE_CHAIN}.  You should not depend in any way on the order in
2767which fields appear on this list.  All nodes on this list will be
2768@samp{DECL} nodes.  A @code{FIELD_DECL} is used to represent a non-static
2769data member, a @code{VAR_DECL} is used to represent a static data
2770member, and a @code{TYPE_DECL} is used to represent a type.  Note that
2771the @code{CONST_DECL} for an enumeration constant will appear on this
2772list, if the enumeration type was declared in the class.  (Of course,
2773the @code{TYPE_DECL} for the enumeration type will appear here as well.)
2774There are no entries for base classes on this list.  In particular,
2775there is no @code{FIELD_DECL} for the ``base-class portion'' of an
2776object.
2777
2778The @code{TYPE_VFIELD} is a compiler-generated field used to point to
2779virtual function tables.  It may or may not appear on the
2780@code{TYPE_FIELDS} list.  However, back ends should handle the
2781@code{TYPE_VFIELD} just like all the entries on the @code{TYPE_FIELDS}
2782list.
2783
2784The function members are available on the @code{TYPE_METHODS} list.
2785Again, subsequent members are found by following the @code{TREE_CHAIN}
2786field.  If a function is overloaded, each of the overloaded functions
2787appears; no @code{OVERLOAD} nodes appear on the @code{TYPE_METHODS}
2788list.  Implicitly declared functions (including default constructors,
2789copy constructors, assignment operators, and destructors) will appear on
2790this list as well.
2791
2792Every class has an associated @dfn{binfo}, which can be obtained with
2793@code{TYPE_BINFO}.  Binfos are used to represent base-classes.  The
2794binfo given by @code{TYPE_BINFO} is the degenerate case, whereby every
2795class is considered to be its own base-class.  The base binfos for a
2796particular binfo are held in a vector, whose length is obtained with
2797@code{BINFO_N_BASE_BINFOS}.  The base binfos themselves are obtained
2798with @code{BINFO_BASE_BINFO} and @code{BINFO_BASE_ITERATE}.  To add a
2799new binfo, use @code{BINFO_BASE_APPEND}.  The vector of base binfos can
2800be obtained with @code{BINFO_BASE_BINFOS}, but normally you do not need
2801to use that.  The class type associated with a binfo is given by
2802@code{BINFO_TYPE}.  It is not always the case that @code{BINFO_TYPE
2803(TYPE_BINFO (x))}, because of typedefs and qualified types.  Neither is
2804it the case that @code{TYPE_BINFO (BINFO_TYPE (y))} is the same binfo as
2805@code{y}.  The reason is that if @code{y} is a binfo representing a
2806base-class @code{B} of a derived class @code{D}, then @code{BINFO_TYPE
2807(y)} will be @code{B}, and @code{TYPE_BINFO (BINFO_TYPE (y))} will be
2808@code{B} as its own base-class, rather than as a base-class of @code{D}.
2809
2810The access to a base type can be found with @code{BINFO_BASE_ACCESS}.
2811This will produce @code{access_public_node}, @code{access_private_node}
2812or @code{access_protected_node}.  If bases are always public,
2813@code{BINFO_BASE_ACCESSES} may be @code{NULL}.
2814
2815@code{BINFO_VIRTUAL_P} is used to specify whether the binfo is inherited
2816virtually or not.  The other flags, @code{BINFO_MARKED_P} and
2817@code{BINFO_FLAG_1} to @code{BINFO_FLAG_6} can be used for language
2818specific use.
2819
2820The following macros can be used on a tree node representing a class-type.
2821
2822@ftable @code
2823@item LOCAL_CLASS_P
2824This predicate holds if the class is local class @emph{i.e.}@: declared
2825inside a function body.
2826
2827@item TYPE_POLYMORPHIC_P
2828This predicate holds if the class has at least one virtual function
2829(declared or inherited).
2830
2831@item TYPE_HAS_DEFAULT_CONSTRUCTOR
2832This predicate holds whenever its argument represents a class-type with
2833default constructor.
2834
2835@item CLASSTYPE_HAS_MUTABLE
2836@itemx TYPE_HAS_MUTABLE_P
2837These predicates hold for a class-type having a mutable data member.
2838
2839@item CLASSTYPE_NON_POD_P
2840This predicate holds only for class-types that are not PODs.
2841
2842@item TYPE_HAS_NEW_OPERATOR
2843This predicate holds for a class-type that defines
2844@code{operator new}.
2845
2846@item TYPE_HAS_ARRAY_NEW_OPERATOR
2847This predicate holds for a class-type for which
2848@code{operator new[]} is defined.
2849
2850@item TYPE_OVERLOADS_CALL_EXPR
2851This predicate holds for class-type for which the function call
2852@code{operator()} is overloaded.
2853
2854@item TYPE_OVERLOADS_ARRAY_REF
2855This predicate holds for a class-type that overloads
2856@code{operator[]}
2857
2858@item TYPE_OVERLOADS_ARROW
2859This predicate holds for a class-type for which @code{operator->} is
2860overloaded.
2861
2862@end ftable
2863
2864@node Functions for C++
2865@subsection Functions for C++
2866@cindex function
2867@tindex FUNCTION_DECL
2868@tindex OVERLOAD
2869@findex OVL_CURRENT
2870@findex OVL_NEXT
2871
2872A function is represented by a @code{FUNCTION_DECL} node.  A set of
2873overloaded functions is sometimes represented by an @code{OVERLOAD} node.
2874
2875An @code{OVERLOAD} node is not a declaration, so none of the
2876@samp{DECL_} macros should be used on an @code{OVERLOAD}.  An
2877@code{OVERLOAD} node is similar to a @code{TREE_LIST}.  Use
2878@code{OVL_CURRENT} to get the function associated with an
2879@code{OVERLOAD} node; use @code{OVL_NEXT} to get the next
2880@code{OVERLOAD} node in the list of overloaded functions.  The macros
2881@code{OVL_CURRENT} and @code{OVL_NEXT} are actually polymorphic; you can
2882use them to work with @code{FUNCTION_DECL} nodes as well as with
2883overloads.  In the case of a @code{FUNCTION_DECL}, @code{OVL_CURRENT}
2884will always return the function itself, and @code{OVL_NEXT} will always
2885be @code{NULL_TREE}.
2886
2887To determine the scope of a function, you can use the
2888@code{DECL_CONTEXT} macro.  This macro will return the class
2889(either a @code{RECORD_TYPE} or a @code{UNION_TYPE}) or namespace (a
2890@code{NAMESPACE_DECL}) of which the function is a member.  For a virtual
2891function, this macro returns the class in which the function was
2892actually defined, not the base class in which the virtual declaration
2893occurred.
2894
2895If a friend function is defined in a class scope, the
2896@code{DECL_FRIEND_CONTEXT} macro can be used to determine the class in
2897which it was defined.  For example, in
2898@smallexample
2899class C @{ friend void f() @{@} @};
2900@end smallexample
2901@noindent
2902the @code{DECL_CONTEXT} for @code{f} will be the
2903@code{global_namespace}, but the @code{DECL_FRIEND_CONTEXT} will be the
2904@code{RECORD_TYPE} for @code{C}.
2905
2906
2907The following macros and functions can be used on a @code{FUNCTION_DECL}:
2908@ftable @code
2909@item DECL_MAIN_P
2910This predicate holds for a function that is the program entry point
2911@code{::code}.
2912
2913@item DECL_LOCAL_FUNCTION_P
2914This predicate holds if the function was declared at block scope, even
2915though it has a global scope.
2916
2917@item DECL_ANTICIPATED
2918This predicate holds if the function is a built-in function but its
2919prototype is not yet explicitly declared.
2920
2921@item DECL_EXTERN_C_FUNCTION_P
2922This predicate holds if the function is declared as an
2923`@code{extern "C"}' function.
2924
2925@item DECL_LINKONCE_P
2926This macro holds if multiple copies of this function may be emitted in
2927various translation units.  It is the responsibility of the linker to
2928merge the various copies.  Template instantiations are the most common
2929example of functions for which @code{DECL_LINKONCE_P} holds; G++
2930instantiates needed templates in all translation units which require them,
2931and then relies on the linker to remove duplicate instantiations.
2932
2933FIXME: This macro is not yet implemented.
2934
2935@item DECL_FUNCTION_MEMBER_P
2936This macro holds if the function is a member of a class, rather than a
2937member of a namespace.
2938
2939@item DECL_STATIC_FUNCTION_P
2940This predicate holds if the function a static member function.
2941
2942@item DECL_NONSTATIC_MEMBER_FUNCTION_P
2943This macro holds for a non-static member function.
2944
2945@item DECL_CONST_MEMFUNC_P
2946This predicate holds for a @code{const}-member function.
2947
2948@item DECL_VOLATILE_MEMFUNC_P
2949This predicate holds for a @code{volatile}-member function.
2950
2951@item DECL_CONSTRUCTOR_P
2952This macro holds if the function is a constructor.
2953
2954@item DECL_NONCONVERTING_P
2955This predicate holds if the constructor is a non-converting constructor.
2956
2957@item DECL_COMPLETE_CONSTRUCTOR_P
2958This predicate holds for a function which is a constructor for an object
2959of a complete type.
2960
2961@item DECL_BASE_CONSTRUCTOR_P
2962This predicate holds for a function which is a constructor for a base
2963class sub-object.
2964
2965@item DECL_COPY_CONSTRUCTOR_P
2966This predicate holds for a function which is a copy-constructor.
2967
2968@item DECL_DESTRUCTOR_P
2969This macro holds if the function is a destructor.
2970
2971@item DECL_COMPLETE_DESTRUCTOR_P
2972This predicate holds if the function is the destructor for an object a
2973complete type.
2974
2975@item DECL_OVERLOADED_OPERATOR_P
2976This macro holds if the function is an overloaded operator.
2977
2978@item DECL_CONV_FN_P
2979This macro holds if the function is a type-conversion operator.
2980
2981@item DECL_GLOBAL_CTOR_P
2982This predicate holds if the function is a file-scope initialization
2983function.
2984
2985@item DECL_GLOBAL_DTOR_P
2986This predicate holds if the function is a file-scope finalization
2987function.
2988
2989@item DECL_THUNK_P
2990This predicate holds if the function is a thunk.
2991
2992These functions represent stub code that adjusts the @code{this} pointer
2993and then jumps to another function.  When the jumped-to function
2994returns, control is transferred directly to the caller, without
2995returning to the thunk.  The first parameter to the thunk is always the
2996@code{this} pointer; the thunk should add @code{THUNK_DELTA} to this
2997value.  (The @code{THUNK_DELTA} is an @code{int}, not an
2998@code{INTEGER_CST}.)
2999
3000Then, if @code{THUNK_VCALL_OFFSET} (an @code{INTEGER_CST}) is nonzero
3001the adjusted @code{this} pointer must be adjusted again.  The complete
3002calculation is given by the following pseudo-code:
3003
3004@smallexample
3005this += THUNK_DELTA
3006if (THUNK_VCALL_OFFSET)
3007  this += (*((ptrdiff_t **) this))[THUNK_VCALL_OFFSET]
3008@end smallexample
3009
3010Finally, the thunk should jump to the location given
3011by @code{DECL_INITIAL}; this will always be an expression for the
3012address of a function.
3013
3014@item DECL_NON_THUNK_FUNCTION_P
3015This predicate holds if the function is @emph{not} a thunk function.
3016
3017@item GLOBAL_INIT_PRIORITY
3018If either @code{DECL_GLOBAL_CTOR_P} or @code{DECL_GLOBAL_DTOR_P} holds,
3019then this gives the initialization priority for the function.  The
3020linker will arrange that all functions for which
3021@code{DECL_GLOBAL_CTOR_P} holds are run in increasing order of priority
3022before @code{main} is called.  When the program exits, all functions for
3023which @code{DECL_GLOBAL_DTOR_P} holds are run in the reverse order.
3024
3025@item TYPE_RAISES_EXCEPTIONS
3026This macro returns the list of exceptions that a (member-)function can
3027raise.  The returned list, if non @code{NULL}, is comprised of nodes
3028whose @code{TREE_VALUE} represents a type.
3029
3030@item TYPE_NOTHROW_P
3031This predicate holds when the exception-specification of its arguments
3032is of the form `@code{()}'.
3033
3034@item DECL_ARRAY_DELETE_OPERATOR_P
3035This predicate holds if the function an overloaded
3036@code{operator delete[]}.
3037
3038@end ftable
3039
3040@c ---------------------------------------------------------------------
3041@c Function Bodies
3042@c ---------------------------------------------------------------------
3043
3044@node Statements for C++
3045@subsection Statements for C++
3046@cindex statements
3047@tindex BREAK_STMT
3048@tindex CLEANUP_STMT
3049@findex CLEANUP_DECL
3050@findex CLEANUP_EXPR
3051@tindex CONTINUE_STMT
3052@tindex DECL_STMT
3053@findex DECL_STMT_DECL
3054@tindex DO_STMT
3055@findex DO_BODY
3056@findex DO_COND
3057@tindex EMPTY_CLASS_EXPR
3058@tindex EXPR_STMT
3059@findex EXPR_STMT_EXPR
3060@tindex FOR_STMT
3061@findex FOR_INIT_STMT
3062@findex FOR_COND
3063@findex FOR_EXPR
3064@findex FOR_BODY
3065@tindex HANDLER
3066@tindex IF_STMT
3067@findex IF_COND
3068@findex THEN_CLAUSE
3069@findex ELSE_CLAUSE
3070@tindex RETURN_STMT
3071@findex RETURN_EXPR
3072@tindex SUBOBJECT
3073@findex SUBOBJECT_CLEANUP
3074@tindex SWITCH_STMT
3075@findex SWITCH_COND
3076@findex SWITCH_BODY
3077@tindex TRY_BLOCK
3078@findex TRY_STMTS
3079@findex TRY_HANDLERS
3080@findex HANDLER_PARMS
3081@findex HANDLER_BODY
3082@findex USING_STMT
3083@tindex WHILE_STMT
3084@findex WHILE_BODY
3085@findex WHILE_COND
3086
3087A function that has a definition in the current translation unit will
3088have a non-@code{NULL} @code{DECL_INITIAL}.  However, back ends should not make
3089use of the particular value given by @code{DECL_INITIAL}.
3090
3091The @code{DECL_SAVED_TREE} macro will give the complete body of the
3092function.
3093
3094@subsubsection Statements
3095
3096There are tree nodes corresponding to all of the source-level
3097statement constructs, used within the C and C++ frontends.  These are
3098enumerated here, together with a list of the various macros that can
3099be used to obtain information about them.  There are a few macros that
3100can be used with all statements:
3101
3102@ftable @code
3103@item STMT_IS_FULL_EXPR_P
3104In C++, statements normally constitute ``full expressions''; temporaries
3105created during a statement are destroyed when the statement is complete.
3106However, G++ sometimes represents expressions by statements; these
3107statements will not have @code{STMT_IS_FULL_EXPR_P} set.  Temporaries
3108created during such statements should be destroyed when the innermost
3109enclosing statement with @code{STMT_IS_FULL_EXPR_P} set is exited.
3110
3111@end ftable
3112
3113Here is the list of the various statement nodes, and the macros used to
3114access them.  This documentation describes the use of these nodes in
3115non-template functions (including instantiations of template functions).
3116In template functions, the same nodes are used, but sometimes in
3117slightly different ways.
3118
3119Many of the statements have substatements.  For example, a @code{while}
3120loop will have a body, which is itself a statement.  If the substatement
3121is @code{NULL_TREE}, it is considered equivalent to a statement
3122consisting of a single @code{;}, i.e., an expression statement in which
3123the expression has been omitted.  A substatement may in fact be a list
3124of statements, connected via their @code{TREE_CHAIN}s.  So, you should
3125always process the statement tree by looping over substatements, like
3126this:
3127@smallexample
3128void process_stmt (stmt)
3129     tree stmt;
3130@{
3131  while (stmt)
3132    @{
3133      switch (TREE_CODE (stmt))
3134        @{
3135        case IF_STMT:
3136          process_stmt (THEN_CLAUSE (stmt));
3137          /* @r{More processing here.}  */
3138          break;
3139
3140        @dots{}
3141        @}
3142
3143      stmt = TREE_CHAIN (stmt);
3144    @}
3145@}
3146@end smallexample
3147In other words, while the @code{then} clause of an @code{if} statement
3148in C++ can be only one statement (although that one statement may be a
3149compound statement), the intermediate representation will sometimes use
3150several statements chained together.
3151
3152@table @code
3153@item BREAK_STMT
3154
3155Used to represent a @code{break} statement.  There are no additional
3156fields.
3157
3158@item CLEANUP_STMT
3159
3160Used to represent an action that should take place upon exit from the
3161enclosing scope.  Typically, these actions are calls to destructors for
3162local objects, but back ends cannot rely on this fact.  If these nodes
3163are in fact representing such destructors, @code{CLEANUP_DECL} will be
3164the @code{VAR_DECL} destroyed.  Otherwise, @code{CLEANUP_DECL} will be
3165@code{NULL_TREE}.  In any case, the @code{CLEANUP_EXPR} is the
3166expression to execute.  The cleanups executed on exit from a scope
3167should be run in the reverse order of the order in which the associated
3168@code{CLEANUP_STMT}s were encountered.
3169
3170@item CONTINUE_STMT
3171
3172Used to represent a @code{continue} statement.  There are no additional
3173fields.
3174
3175@item CTOR_STMT
3176
3177Used to mark the beginning (if @code{CTOR_BEGIN_P} holds) or end (if
3178@code{CTOR_END_P} holds of the main body of a constructor.  See also
3179@code{SUBOBJECT} for more information on how to use these nodes.
3180
3181@item DO_STMT
3182
3183Used to represent a @code{do} loop.  The body of the loop is given by
3184@code{DO_BODY} while the termination condition for the loop is given by
3185@code{DO_COND}.  The condition for a @code{do}-statement is always an
3186expression.
3187
3188@item EMPTY_CLASS_EXPR
3189
3190Used to represent a temporary object of a class with no data whose
3191address is never taken.  (All such objects are interchangeable.)  The
3192@code{TREE_TYPE} represents the type of the object.
3193
3194@item EXPR_STMT
3195
3196Used to represent an expression statement.  Use @code{EXPR_STMT_EXPR} to
3197obtain the expression.
3198
3199@item FOR_STMT
3200
3201Used to represent a @code{for} statement.  The @code{FOR_INIT_STMT} is
3202the initialization statement for the loop.  The @code{FOR_COND} is the
3203termination condition.  The @code{FOR_EXPR} is the expression executed
3204right before the @code{FOR_COND} on each loop iteration; often, this
3205expression increments a counter.  The body of the loop is given by
3206@code{FOR_BODY}.  Note that @code{FOR_INIT_STMT} and @code{FOR_BODY}
3207return statements, while @code{FOR_COND} and @code{FOR_EXPR} return
3208expressions.
3209
3210@item HANDLER
3211
3212Used to represent a C++ @code{catch} block.  The @code{HANDLER_TYPE}
3213is the type of exception that will be caught by this handler; it is
3214equal (by pointer equality) to @code{NULL} if this handler is for all
3215types.  @code{HANDLER_PARMS} is the @code{DECL_STMT} for the catch
3216parameter, and @code{HANDLER_BODY} is the code for the block itself.
3217
3218@item IF_STMT
3219
3220Used to represent an @code{if} statement.  The @code{IF_COND} is the
3221expression.
3222
3223If the condition is a @code{TREE_LIST}, then the @code{TREE_PURPOSE} is
3224a statement (usually a @code{DECL_STMT}).  Each time the condition is
3225evaluated, the statement should be executed.  Then, the
3226@code{TREE_VALUE} should be used as the conditional expression itself.
3227This representation is used to handle C++ code like this:
3228
3229C++ distinguishes between this and @code{COND_EXPR} for handling templates.
3230
3231@smallexample
3232if (int i = 7) @dots{}
3233@end smallexample
3234
3235where there is a new local variable (or variables) declared within the
3236condition.
3237
3238The @code{THEN_CLAUSE} represents the statement given by the @code{then}
3239condition, while the @code{ELSE_CLAUSE} represents the statement given
3240by the @code{else} condition.
3241
3242@item SUBOBJECT
3243
3244In a constructor, these nodes are used to mark the point at which a
3245subobject of @code{this} is fully constructed.  If, after this point, an
3246exception is thrown before a @code{CTOR_STMT} with @code{CTOR_END_P} set
3247is encountered, the @code{SUBOBJECT_CLEANUP} must be executed.  The
3248cleanups must be executed in the reverse order in which they appear.
3249
3250@item SWITCH_STMT
3251
3252Used to represent a @code{switch} statement.  The @code{SWITCH_STMT_COND}
3253is the expression on which the switch is occurring.  See the documentation
3254for an @code{IF_STMT} for more information on the representation used
3255for the condition.  The @code{SWITCH_STMT_BODY} is the body of the switch
3256statement.   The @code{SWITCH_STMT_TYPE} is the original type of switch
3257expression as given in the source, before any compiler conversions.
3258
3259@item TRY_BLOCK
3260Used to represent a @code{try} block.  The body of the try block is
3261given by @code{TRY_STMTS}.  Each of the catch blocks is a @code{HANDLER}
3262node.  The first handler is given by @code{TRY_HANDLERS}.  Subsequent
3263handlers are obtained by following the @code{TREE_CHAIN} link from one
3264handler to the next.  The body of the handler is given by
3265@code{HANDLER_BODY}.
3266
3267If @code{CLEANUP_P} holds of the @code{TRY_BLOCK}, then the
3268@code{TRY_HANDLERS} will not be a @code{HANDLER} node.  Instead, it will
3269be an expression that should be executed if an exception is thrown in
3270the try block.  It must rethrow the exception after executing that code.
3271And, if an exception is thrown while the expression is executing,
3272@code{terminate} must be called.
3273
3274@item USING_STMT
3275Used to represent a @code{using} directive.  The namespace is given by
3276@code{USING_STMT_NAMESPACE}, which will be a NAMESPACE_DECL@.  This node
3277is needed inside template functions, to implement using directives
3278during instantiation.
3279
3280@item WHILE_STMT
3281
3282Used to represent a @code{while} loop.  The @code{WHILE_COND} is the
3283termination condition for the loop.  See the documentation for an
3284@code{IF_STMT} for more information on the representation used for the
3285condition.
3286
3287The @code{WHILE_BODY} is the body of the loop.
3288
3289@end table
3290
3291@node C++ Expressions
3292@subsection C++ Expressions
3293
3294This section describes expressions specific to the C and C++ front
3295ends.
3296
3297@table @code
3298@item TYPEID_EXPR
3299
3300Used to represent a @code{typeid} expression.
3301
3302@item NEW_EXPR
3303@itemx VEC_NEW_EXPR
3304
3305Used to represent a call to @code{new} and @code{new[]} respectively.
3306
3307@item DELETE_EXPR
3308@itemx VEC_DELETE_EXPR
3309
3310Used to represent a call to @code{delete} and @code{delete[]} respectively.
3311
3312@item MEMBER_REF
3313
3314Represents a reference to a member of a class.
3315
3316@item THROW_EXPR
3317
3318Represents an instance of @code{throw} in the program.  Operand 0,
3319which is the expression to throw, may be @code{NULL_TREE}.
3320
3321
3322@item AGGR_INIT_EXPR
3323An @code{AGGR_INIT_EXPR} represents the initialization as the return
3324value of a function call, or as the result of a constructor.  An
3325@code{AGGR_INIT_EXPR} will only appear as a full-expression, or as the
3326second operand of a @code{TARGET_EXPR}.  @code{AGGR_INIT_EXPR}s have
3327a representation similar to that of @code{CALL_EXPR}s.  You can use
3328the @code{AGGR_INIT_EXPR_FN} and @code{AGGR_INIT_EXPR_ARG} macros to access
3329the function to call and the arguments to pass.
3330
3331If @code{AGGR_INIT_VIA_CTOR_P} holds of the @code{AGGR_INIT_EXPR}, then
3332the initialization is via a constructor call.  The address of the
3333@code{AGGR_INIT_EXPR_SLOT} operand, which is always a @code{VAR_DECL},
3334is taken, and this value replaces the first argument in the argument
3335list.
3336
3337In either case, the expression is void.
3338
3339
3340@end table
3341
3342
3343@node Java Trees
3344@section Java Trees
3345