* gcc_release (build_sources): Use two new variables EXPORTTAG and
[official-gcc.git] / gcc / c-lex.c
blob47515e7b7032020e8963ce4962b339bf126cf708
1 /* Mainly the interface between cpplib and the C front ends.
2 Copyright (C) 1987, 1988, 1989, 1992, 1994, 1995, 1996, 1997
3 1998, 1999, 2000, 2001, 2002, 2003 Free Software Foundation, Inc.
5 This file is part of GCC.
7 GCC is free software; you can redistribute it and/or modify it under
8 the terms of the GNU General Public License as published by the Free
9 Software Foundation; either version 2, or (at your option) any later
10 version.
12 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
13 WARRANTY; without even the implied warranty of MERCHANTABILITY or
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
15 for more details.
17 You should have received a copy of the GNU General Public License
18 along with GCC; see the file COPYING. If not, write to the Free
19 Software Foundation, 59 Temple Place - Suite 330, Boston, MA
20 02111-1307, USA. */
22 #include "config.h"
23 #include "system.h"
24 #include "coretypes.h"
25 #include "tm.h"
27 #include "real.h"
28 #include "rtl.h"
29 #include "tree.h"
30 #include "expr.h"
31 #include "input.h"
32 #include "output.h"
33 #include "c-tree.h"
34 #include "c-common.h"
35 #include "flags.h"
36 #include "timevar.h"
37 #include "cpplib.h"
38 #include "c-pragma.h"
39 #include "toplev.h"
40 #include "intl.h"
41 #include "tm_p.h"
42 #include "splay-tree.h"
43 #include "debug.h"
45 /* The current line map. */
46 static const struct line_map *map;
48 /* The line used to refresh the lineno global variable after each token. */
49 static unsigned int src_lineno;
51 /* We may keep statistics about how long which files took to compile. */
52 static int header_time, body_time;
53 static splay_tree file_info_tree;
55 #undef WCHAR_TYPE_SIZE
56 #define WCHAR_TYPE_SIZE TYPE_PRECISION (wchar_type_node)
58 /* Number of bytes in a wide character. */
59 #define WCHAR_BYTES (WCHAR_TYPE_SIZE / BITS_PER_UNIT)
61 int pending_lang_change; /* If we need to switch languages - C++ only */
62 int c_header_level; /* depth in C headers - C++ only */
64 static tree interpret_integer (const cpp_token *, unsigned int);
65 static tree interpret_float (const cpp_token *, unsigned int);
66 static enum integer_type_kind
67 narrowest_unsigned_type (tree, unsigned int);
68 static enum integer_type_kind
69 narrowest_signed_type (tree, unsigned int);
70 static enum cpp_ttype lex_string (const cpp_token *, tree *, bool);
71 static tree lex_charconst (const cpp_token *);
72 static void update_header_times (const char *);
73 static int dump_one_header (splay_tree_node, void *);
74 static void cb_line_change (cpp_reader *, const cpp_token *, int);
75 static void cb_dir_change (cpp_reader *, const char *);
76 static void cb_ident (cpp_reader *, unsigned int, const cpp_string *);
77 static void cb_def_pragma (cpp_reader *, unsigned int);
78 static void cb_define (cpp_reader *, unsigned int, cpp_hashnode *);
79 static void cb_undef (cpp_reader *, unsigned int, cpp_hashnode *);
81 void
82 init_c_lex (void)
84 struct cpp_callbacks *cb;
85 struct c_fileinfo *toplevel;
87 /* Set up filename timing. Must happen before cpp_read_main_file. */
88 file_info_tree = splay_tree_new ((splay_tree_compare_fn)strcmp,
90 (splay_tree_delete_value_fn)free);
91 toplevel = get_fileinfo ("<top level>");
92 if (flag_detailed_statistics)
94 header_time = 0;
95 body_time = get_run_time ();
96 toplevel->time = body_time;
99 cb = cpp_get_callbacks (parse_in);
101 cb->line_change = cb_line_change;
102 cb->dir_change = cb_dir_change;
103 cb->ident = cb_ident;
104 cb->def_pragma = cb_def_pragma;
105 cb->valid_pch = c_common_valid_pch;
106 cb->read_pch = c_common_read_pch;
108 /* Set the debug callbacks if we can use them. */
109 if (debug_info_level == DINFO_LEVEL_VERBOSE
110 && (write_symbols == DWARF_DEBUG || write_symbols == DWARF2_DEBUG
111 || write_symbols == VMS_AND_DWARF2_DEBUG))
113 cb->define = cb_define;
114 cb->undef = cb_undef;
118 struct c_fileinfo *
119 get_fileinfo (const char *name)
121 splay_tree_node n;
122 struct c_fileinfo *fi;
124 n = splay_tree_lookup (file_info_tree, (splay_tree_key) name);
125 if (n)
126 return (struct c_fileinfo *) n->value;
128 fi = xmalloc (sizeof (struct c_fileinfo));
129 fi->time = 0;
130 fi->interface_only = 0;
131 fi->interface_unknown = 1;
132 splay_tree_insert (file_info_tree, (splay_tree_key) name,
133 (splay_tree_value) fi);
134 return fi;
137 static void
138 update_header_times (const char *name)
140 /* Changing files again. This means currently collected time
141 is charged against header time, and body time starts back at 0. */
142 if (flag_detailed_statistics)
144 int this_time = get_run_time ();
145 struct c_fileinfo *file = get_fileinfo (name);
146 header_time += this_time - body_time;
147 file->time += this_time - body_time;
148 body_time = this_time;
152 static int
153 dump_one_header (splay_tree_node n, void *dummy ATTRIBUTE_UNUSED)
155 print_time ((const char *) n->key,
156 ((struct c_fileinfo *) n->value)->time);
157 return 0;
160 void
161 dump_time_statistics (void)
163 struct c_fileinfo *file = get_fileinfo (input_filename);
164 int this_time = get_run_time ();
165 file->time += this_time - body_time;
167 fprintf (stderr, "\n******\n");
168 print_time ("header files (total)", header_time);
169 print_time ("main file (total)", this_time - body_time);
170 fprintf (stderr, "ratio = %g : 1\n",
171 (double)header_time / (double)(this_time - body_time));
172 fprintf (stderr, "\n******\n");
174 splay_tree_foreach (file_info_tree, dump_one_header, 0);
177 static void
178 cb_ident (cpp_reader *pfile ATTRIBUTE_UNUSED,
179 unsigned int line ATTRIBUTE_UNUSED,
180 const cpp_string *str ATTRIBUTE_UNUSED)
182 #ifdef ASM_OUTPUT_IDENT
183 if (! flag_no_ident)
185 /* Convert escapes in the string. */
186 cpp_string cstr = { 0, 0 };
187 if (cpp_interpret_string (pfile, str, 1, &cstr, false))
189 ASM_OUTPUT_IDENT (asm_out_file, (const char *) cstr.text);
190 free ((void *)cstr.text);
193 #endif
196 /* Called at the start of every non-empty line. TOKEN is the first
197 lexed token on the line. Used for diagnostic line numbers. */
198 static void
199 cb_line_change (cpp_reader *pfile ATTRIBUTE_UNUSED, const cpp_token *token,
200 int parsing_args ATTRIBUTE_UNUSED)
202 src_lineno = SOURCE_LINE (map, token->line);
205 static void
206 cb_dir_change (cpp_reader *pfile ATTRIBUTE_UNUSED, const char *dir)
208 if (! set_src_pwd (dir))
209 warning ("too late for # directive to set debug directory");
212 void
213 fe_file_change (const struct line_map *new_map)
215 unsigned int to_line = SOURCE_LINE (new_map, new_map->to_line);
217 if (new_map->reason == LC_ENTER)
219 /* Don't stack the main buffer on the input stack;
220 we already did in compile_file. */
221 if (map == NULL)
222 main_input_filename = new_map->to_file;
223 else
225 int included_at = SOURCE_LINE (new_map - 1, new_map->from_line - 1);
227 input_line = included_at;
228 push_srcloc (new_map->to_file, 1);
229 (*debug_hooks->start_source_file) (included_at, new_map->to_file);
230 #ifndef NO_IMPLICIT_EXTERN_C
231 if (c_header_level)
232 ++c_header_level;
233 else if (new_map->sysp == 2)
235 c_header_level = 1;
236 ++pending_lang_change;
238 #endif
241 else if (new_map->reason == LC_LEAVE)
243 #ifndef NO_IMPLICIT_EXTERN_C
244 if (c_header_level && --c_header_level == 0)
246 if (new_map->sysp == 2)
247 warning ("badly nested C headers from preprocessor");
248 --pending_lang_change;
250 #endif
251 pop_srcloc ();
253 (*debug_hooks->end_source_file) (to_line);
256 update_header_times (new_map->to_file);
257 in_system_header = new_map->sysp != 0;
258 input_filename = new_map->to_file;
259 input_line = to_line;
260 map = new_map;
262 /* Hook for C++. */
263 extract_interface_info ();
266 static void
267 cb_def_pragma (cpp_reader *pfile, unsigned int line)
269 /* Issue a warning message if we have been asked to do so. Ignore
270 unknown pragmas in system headers unless an explicit
271 -Wunknown-pragmas has been given. */
272 if (warn_unknown_pragmas > in_system_header)
274 const unsigned char *space, *name;
275 const cpp_token *s;
277 space = name = (const unsigned char *) "";
278 s = cpp_get_token (pfile);
279 if (s->type != CPP_EOF)
281 space = cpp_token_as_text (pfile, s);
282 s = cpp_get_token (pfile);
283 if (s->type == CPP_NAME)
284 name = cpp_token_as_text (pfile, s);
287 input_line = SOURCE_LINE (map, line);
288 warning ("ignoring #pragma %s %s", space, name);
292 /* #define callback for DWARF and DWARF2 debug info. */
293 static void
294 cb_define (cpp_reader *pfile, unsigned int line, cpp_hashnode *node)
296 (*debug_hooks->define) (SOURCE_LINE (map, line),
297 (const char *) cpp_macro_definition (pfile, node));
300 /* #undef callback for DWARF and DWARF2 debug info. */
301 static void
302 cb_undef (cpp_reader *pfile ATTRIBUTE_UNUSED, unsigned int line,
303 cpp_hashnode *node)
305 (*debug_hooks->undef) (SOURCE_LINE (map, line),
306 (const char *) NODE_NAME (node));
309 static inline const cpp_token *
310 get_nonpadding_token (void)
312 const cpp_token *tok;
313 timevar_push (TV_CPP);
315 tok = cpp_get_token (parse_in);
316 while (tok->type == CPP_PADDING);
317 timevar_pop (TV_CPP);
319 /* The C++ front end does horrible things with the current line
320 number. To ensure an accurate line number, we must reset it
321 every time we advance a token. */
322 input_line = src_lineno;
324 return tok;
328 c_lex (tree *value)
330 const cpp_token *tok;
331 location_t atloc;
332 static bool no_more_pch;
334 retry:
335 tok = get_nonpadding_token ();
337 retry_after_at:
338 switch (tok->type)
340 case CPP_NAME:
341 *value = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node));
342 break;
344 case CPP_NUMBER:
346 unsigned int flags = cpp_classify_number (parse_in, tok);
348 switch (flags & CPP_N_CATEGORY)
350 case CPP_N_INVALID:
351 /* cpplib has issued an error. */
352 *value = error_mark_node;
353 break;
355 case CPP_N_INTEGER:
356 *value = interpret_integer (tok, flags);
357 break;
359 case CPP_N_FLOATING:
360 *value = interpret_float (tok, flags);
361 break;
363 default:
364 abort ();
367 break;
369 case CPP_ATSIGN:
370 /* An @ may give the next token special significance in Objective-C. */
371 atloc = input_location;
372 tok = get_nonpadding_token ();
373 if (c_dialect_objc ())
375 tree val;
376 switch (tok->type)
378 case CPP_NAME:
379 val = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node));
380 if (C_IS_RESERVED_WORD (val)
381 && OBJC_IS_AT_KEYWORD (C_RID_CODE (val)))
383 *value = val;
384 return CPP_AT_NAME;
386 break;
388 case CPP_STRING:
389 case CPP_WSTRING:
390 return lex_string (tok, value, true);
392 default: break;
396 /* ... or not. */
397 error ("%Hstray '@' in program", &atloc);
398 goto retry_after_at;
400 case CPP_OTHER:
402 cppchar_t c = tok->val.str.text[0];
404 if (c == '"' || c == '\'')
405 error ("missing terminating %c character", (int) c);
406 else if (ISGRAPH (c))
407 error ("stray '%c' in program", (int) c);
408 else
409 error ("stray '\\%o' in program", (int) c);
411 goto retry;
413 case CPP_CHAR:
414 case CPP_WCHAR:
415 *value = lex_charconst (tok);
416 break;
418 case CPP_STRING:
419 case CPP_WSTRING:
420 return lex_string (tok, value, false);
421 break;
423 /* These tokens should not be visible outside cpplib. */
424 case CPP_HEADER_NAME:
425 case CPP_COMMENT:
426 case CPP_MACRO_ARG:
427 abort ();
429 default:
430 *value = NULL_TREE;
431 break;
434 if (! no_more_pch)
436 no_more_pch = true;
437 c_common_no_more_pch ();
440 return tok->type;
443 /* Returns the narrowest C-visible unsigned type, starting with the
444 minimum specified by FLAGS, that can fit VALUE, or itk_none if
445 there isn't one. */
446 static enum integer_type_kind
447 narrowest_unsigned_type (tree value, unsigned int flags)
449 enum integer_type_kind itk;
451 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
452 itk = itk_unsigned_int;
453 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
454 itk = itk_unsigned_long;
455 else
456 itk = itk_unsigned_long_long;
458 /* int_fits_type_p must think the type of its first argument is
459 wider than its second argument, or it won't do the proper check. */
460 TREE_TYPE (value) = widest_unsigned_literal_type_node;
462 for (; itk < itk_none; itk += 2 /* skip unsigned types */)
463 if (int_fits_type_p (value, integer_types[itk]))
464 return itk;
466 return itk_none;
469 /* Ditto, but narrowest signed type. */
470 static enum integer_type_kind
471 narrowest_signed_type (tree value, unsigned int flags)
473 enum integer_type_kind itk;
475 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
476 itk = itk_int;
477 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
478 itk = itk_long;
479 else
480 itk = itk_long_long;
482 /* int_fits_type_p must think the type of its first argument is
483 wider than its second argument, or it won't do the proper check. */
484 TREE_TYPE (value) = widest_unsigned_literal_type_node;
486 for (; itk < itk_none; itk += 2 /* skip signed types */)
487 if (int_fits_type_p (value, integer_types[itk]))
488 return itk;
490 return itk_none;
493 /* Interpret TOKEN, an integer with FLAGS as classified by cpplib. */
494 static tree
495 interpret_integer (const cpp_token *token, unsigned int flags)
497 tree value, type;
498 enum integer_type_kind itk;
499 cpp_num integer;
500 cpp_options *options = cpp_get_options (parse_in);
502 integer = cpp_interpret_integer (parse_in, token, flags);
503 integer = cpp_num_sign_extend (integer, options->precision);
504 value = build_int_2_wide (integer.low, integer.high);
506 /* The type of a constant with a U suffix is straightforward. */
507 if (flags & CPP_N_UNSIGNED)
508 itk = narrowest_unsigned_type (value, flags);
509 else
511 /* The type of a potentially-signed integer constant varies
512 depending on the base it's in, the standard in use, and the
513 length suffixes. */
514 enum integer_type_kind itk_u = narrowest_unsigned_type (value, flags);
515 enum integer_type_kind itk_s = narrowest_signed_type (value, flags);
517 /* In both C89 and C99, octal and hex constants may be signed or
518 unsigned, whichever fits tighter. We do not warn about this
519 choice differing from the traditional choice, as the constant
520 is probably a bit pattern and either way will work. */
521 if ((flags & CPP_N_RADIX) != CPP_N_DECIMAL)
522 itk = MIN (itk_u, itk_s);
523 else
525 /* In C99, decimal constants are always signed.
526 In C89, decimal constants that don't fit in long have
527 undefined behavior; we try to make them unsigned long.
528 In GCC's extended C89, that last is true of decimal
529 constants that don't fit in long long, too. */
531 itk = itk_s;
532 if (itk_s > itk_u && itk_s > itk_long)
534 if (!flag_isoc99)
536 if (itk_u < itk_unsigned_long)
537 itk_u = itk_unsigned_long;
538 itk = itk_u;
539 warning ("this decimal constant is unsigned only in ISO C90");
541 else if (warn_traditional)
542 warning ("this decimal constant would be unsigned in ISO C90");
547 if (itk == itk_none)
548 /* cpplib has already issued a warning for overflow. */
549 type = ((flags & CPP_N_UNSIGNED)
550 ? widest_unsigned_literal_type_node
551 : widest_integer_literal_type_node);
552 else
553 type = integer_types[itk];
555 if (itk > itk_unsigned_long
556 && (flags & CPP_N_WIDTH) != CPP_N_LARGE
557 && ! in_system_header && ! flag_isoc99)
558 pedwarn ("integer constant is too large for \"%s\" type",
559 (flags & CPP_N_UNSIGNED) ? "unsigned long" : "long");
561 TREE_TYPE (value) = type;
563 /* Convert imaginary to a complex type. */
564 if (flags & CPP_N_IMAGINARY)
565 value = build_complex (NULL_TREE, convert (type, integer_zero_node), value);
567 return value;
570 /* Interpret TOKEN, a floating point number with FLAGS as classified
571 by cpplib. */
572 static tree
573 interpret_float (const cpp_token *token, unsigned int flags)
575 tree type;
576 tree value;
577 REAL_VALUE_TYPE real;
578 char *copy;
579 size_t copylen;
580 const char *typename;
582 /* FIXME: make %T work in error/warning, then we don't need typename. */
583 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
585 type = long_double_type_node;
586 typename = "long double";
588 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL
589 || flag_single_precision_constant)
591 type = float_type_node;
592 typename = "float";
594 else
596 type = double_type_node;
597 typename = "double";
600 /* Copy the constant to a nul-terminated buffer. If the constant
601 has any suffixes, cut them off; REAL_VALUE_ATOF/ REAL_VALUE_HTOF
602 can't handle them. */
603 copylen = token->val.str.len;
604 if ((flags & CPP_N_WIDTH) != CPP_N_MEDIUM)
605 /* Must be an F or L suffix. */
606 copylen--;
607 if (flags & CPP_N_IMAGINARY)
608 /* I or J suffix. */
609 copylen--;
611 copy = alloca (copylen + 1);
612 memcpy (copy, token->val.str.text, copylen);
613 copy[copylen] = '\0';
615 real_from_string (&real, copy);
616 real_convert (&real, TYPE_MODE (type), &real);
618 /* A diagnostic is required for "soft" overflow by some ISO C
619 testsuites. This is not pedwarn, because some people don't want
620 an error for this.
621 ??? That's a dubious reason... is this a mandatory diagnostic or
622 isn't it? -- zw, 2001-08-21. */
623 if (REAL_VALUE_ISINF (real) && pedantic)
624 warning ("floating constant exceeds range of \"%s\"", typename);
626 /* Create a node with determined type and value. */
627 value = build_real (type, real);
628 if (flags & CPP_N_IMAGINARY)
629 value = build_complex (NULL_TREE, convert (type, integer_zero_node), value);
631 return value;
634 /* Convert a series of STRING and/or WSTRING tokens into a tree,
635 performing string constant concatenation. TOK is the first of
636 these. VALP is the location to write the string into. OBJC_STRING
637 indicates whether an '@' token preceded the incoming token.
638 Returns the CPP token type of the result (CPP_STRING, CPP_WSTRING,
639 or CPP_OBJC_STRING).
641 This is unfortunately more work than it should be. If any of the
642 strings in the series has an L prefix, the result is a wide string
643 (6.4.5p4). Whether or not the result is a wide string affects the
644 meaning of octal and hexadecimal escapes (6.4.4.4p6,9). But escape
645 sequences do not continue across the boundary between two strings in
646 a series (6.4.5p7), so we must not lose the boundaries. Therefore
647 cpp_interpret_string takes a vector of cpp_string structures, which
648 we must arrange to provide. */
650 static enum cpp_ttype
651 lex_string (const cpp_token *tok, tree *valp, bool objc_string)
653 tree value;
654 bool wide = false;
655 size_t count = 1;
656 struct obstack str_ob;
657 cpp_string istr;
659 /* Try to avoid the overhead of creating and destroying an obstack
660 for the common case of just one string. */
661 cpp_string str = tok->val.str;
662 cpp_string *strs = &str;
664 if (tok->type == CPP_WSTRING)
665 wide = true;
667 tok = get_nonpadding_token ();
668 if (c_dialect_objc () && tok->type == CPP_ATSIGN)
670 objc_string = true;
671 tok = get_nonpadding_token ();
673 if (tok->type == CPP_STRING || tok->type == CPP_WSTRING)
675 gcc_obstack_init (&str_ob);
676 obstack_grow (&str_ob, &str, sizeof (cpp_string));
680 count++;
681 if (tok->type == CPP_WSTRING)
682 wide = true;
683 obstack_grow (&str_ob, &tok->val.str, sizeof (cpp_string));
685 tok = get_nonpadding_token ();
686 if (c_dialect_objc () && tok->type == CPP_ATSIGN)
688 objc_string = true;
689 tok = get_nonpadding_token ();
692 while (tok->type == CPP_STRING || tok->type == CPP_WSTRING);
693 strs = obstack_finish (&str_ob);
696 /* We have read one more token than we want. */
697 _cpp_backup_tokens (parse_in, 1);
699 if (count > 1 && !objc_string && warn_traditional && !in_system_header)
700 warning ("traditional C rejects string constant concatenation");
702 if (cpp_interpret_string (parse_in, strs, count, &istr, wide))
704 value = build_string (istr.len, (char *)istr.text);
705 free ((void *)istr.text);
707 else
709 /* Callers cannot generally handle error_mark_node in this context,
710 so return the empty string instead. cpp_interpret_string has
711 issued an error. */
712 if (wide)
713 value = build_string (TYPE_PRECISION (wchar_type_node)
714 / TYPE_PRECISION (char_type_node),
715 "\0\0\0"); /* widest supported wchar_t
716 is 32 bits */
717 else
718 value = build_string (1, "");
721 TREE_TYPE (value) = wide ? wchar_array_type_node : char_array_type_node;
722 *valp = fix_string_type (value);
724 if (strs != &str)
725 obstack_free (&str_ob, 0);
727 return objc_string ? CPP_OBJC_STRING : wide ? CPP_WSTRING : CPP_STRING;
730 /* Converts a (possibly wide) character constant token into a tree. */
731 static tree
732 lex_charconst (const cpp_token *token)
734 cppchar_t result;
735 tree type, value;
736 unsigned int chars_seen;
737 int unsignedp;
739 result = cpp_interpret_charconst (parse_in, token,
740 &chars_seen, &unsignedp);
742 /* Cast to cppchar_signed_t to get correct sign-extension of RESULT
743 before possibly widening to HOST_WIDE_INT for build_int_2. */
744 if (unsignedp || (cppchar_signed_t) result >= 0)
745 value = build_int_2 (result, 0);
746 else
747 value = build_int_2 ((cppchar_signed_t) result, -1);
749 if (token->type == CPP_WCHAR)
750 type = wchar_type_node;
751 /* In C, a character constant has type 'int'.
752 In C++ 'char', but multi-char charconsts have type 'int'. */
753 else if (!c_dialect_cxx () || chars_seen > 1)
754 type = integer_type_node;
755 else
756 type = char_type_node;
758 TREE_TYPE (value) = type;
759 return value;