* dbxout.c (current_file): Also wrap inside DBX_DEBUGGING_INFO ||
[official-gcc.git] / gcc / c-lex.c
blob5546fc9ddd1933c91ff073b6a861afb0cf2ae960
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 /* We may keep statistics about how long which files took to compile. */
49 static int header_time, body_time;
50 static splay_tree file_info_tree;
52 #undef WCHAR_TYPE_SIZE
53 #define WCHAR_TYPE_SIZE TYPE_PRECISION (wchar_type_node)
55 /* Number of bytes in a wide character. */
56 #define WCHAR_BYTES (WCHAR_TYPE_SIZE / BITS_PER_UNIT)
58 int pending_lang_change; /* If we need to switch languages - C++ only */
59 int c_header_level; /* depth in C headers - C++ only */
61 static tree interpret_integer (const cpp_token *, unsigned int);
62 static tree interpret_float (const cpp_token *, unsigned int);
63 static enum integer_type_kind
64 narrowest_unsigned_type (tree, unsigned int);
65 static enum integer_type_kind
66 narrowest_signed_type (tree, unsigned int);
67 static enum cpp_ttype lex_string (const cpp_token *, tree *, bool);
68 static tree lex_charconst (const cpp_token *);
69 static void update_header_times (const char *);
70 static int dump_one_header (splay_tree_node, void *);
71 static void cb_line_change (cpp_reader *, const cpp_token *, int);
72 static void cb_dir_change (cpp_reader *, const char *);
73 static void cb_ident (cpp_reader *, unsigned int, const cpp_string *);
74 static void cb_def_pragma (cpp_reader *, unsigned int);
75 static void cb_define (cpp_reader *, unsigned int, cpp_hashnode *);
76 static void cb_undef (cpp_reader *, unsigned int, cpp_hashnode *);
78 void
79 init_c_lex (void)
81 struct cpp_callbacks *cb;
82 struct c_fileinfo *toplevel;
84 /* Set up filename timing. Must happen before cpp_read_main_file. */
85 file_info_tree = splay_tree_new ((splay_tree_compare_fn)strcmp,
87 (splay_tree_delete_value_fn)free);
88 toplevel = get_fileinfo ("<top level>");
89 if (flag_detailed_statistics)
91 header_time = 0;
92 body_time = get_run_time ();
93 toplevel->time = body_time;
96 cb = cpp_get_callbacks (parse_in);
98 cb->line_change = cb_line_change;
99 cb->dir_change = cb_dir_change;
100 cb->ident = cb_ident;
101 cb->def_pragma = cb_def_pragma;
102 cb->valid_pch = c_common_valid_pch;
103 cb->read_pch = c_common_read_pch;
105 /* Set the debug callbacks if we can use them. */
106 if (debug_info_level == DINFO_LEVEL_VERBOSE
107 && (write_symbols == DWARF_DEBUG || write_symbols == DWARF2_DEBUG
108 || write_symbols == VMS_AND_DWARF2_DEBUG))
110 cb->define = cb_define;
111 cb->undef = cb_undef;
115 struct c_fileinfo *
116 get_fileinfo (const char *name)
118 splay_tree_node n;
119 struct c_fileinfo *fi;
121 n = splay_tree_lookup (file_info_tree, (splay_tree_key) name);
122 if (n)
123 return (struct c_fileinfo *) n->value;
125 fi = xmalloc (sizeof (struct c_fileinfo));
126 fi->time = 0;
127 fi->interface_only = 0;
128 fi->interface_unknown = 1;
129 splay_tree_insert (file_info_tree, (splay_tree_key) name,
130 (splay_tree_value) fi);
131 return fi;
134 static void
135 update_header_times (const char *name)
137 /* Changing files again. This means currently collected time
138 is charged against header time, and body time starts back at 0. */
139 if (flag_detailed_statistics)
141 int this_time = get_run_time ();
142 struct c_fileinfo *file = get_fileinfo (name);
143 header_time += this_time - body_time;
144 file->time += this_time - body_time;
145 body_time = this_time;
149 static int
150 dump_one_header (splay_tree_node n, void *dummy ATTRIBUTE_UNUSED)
152 print_time ((const char *) n->key,
153 ((struct c_fileinfo *) n->value)->time);
154 return 0;
157 void
158 dump_time_statistics (void)
160 struct c_fileinfo *file = get_fileinfo (input_filename);
161 int this_time = get_run_time ();
162 file->time += this_time - body_time;
164 fprintf (stderr, "\n******\n");
165 print_time ("header files (total)", header_time);
166 print_time ("main file (total)", this_time - body_time);
167 fprintf (stderr, "ratio = %g : 1\n",
168 (double)header_time / (double)(this_time - body_time));
169 fprintf (stderr, "\n******\n");
171 splay_tree_foreach (file_info_tree, dump_one_header, 0);
174 static void
175 cb_ident (cpp_reader *pfile ATTRIBUTE_UNUSED,
176 unsigned int line ATTRIBUTE_UNUSED,
177 const cpp_string *str ATTRIBUTE_UNUSED)
179 #ifdef ASM_OUTPUT_IDENT
180 if (! flag_no_ident)
182 /* Convert escapes in the string. */
183 cpp_string cstr = { 0, 0 };
184 if (cpp_interpret_string (pfile, str, 1, &cstr, false))
186 ASM_OUTPUT_IDENT (asm_out_file, (const char *) cstr.text);
187 free ((void *)cstr.text);
190 #endif
193 /* Called at the start of every non-empty line. TOKEN is the first
194 lexed token on the line. Used for diagnostic line numbers. */
195 static void
196 cb_line_change (cpp_reader *pfile ATTRIBUTE_UNUSED, const cpp_token *token,
197 int parsing_args)
199 if (token->type == CPP_EOF || parsing_args)
200 return;
202 input_line = 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 if (new_map == NULL)
217 map = NULL;
218 return;
221 if (new_map->reason == LC_ENTER)
223 /* Don't stack the main buffer on the input stack;
224 we already did in compile_file. */
225 if (map == NULL)
226 main_input_filename = new_map->to_file;
227 else
229 int included_at = SOURCE_LINE (new_map - 1, new_map->from_line - 1);
231 input_line = included_at;
232 push_srcloc (new_map->to_file, 1);
233 (*debug_hooks->start_source_file) (included_at, new_map->to_file);
234 #ifndef NO_IMPLICIT_EXTERN_C
235 if (c_header_level)
236 ++c_header_level;
237 else if (new_map->sysp == 2)
239 c_header_level = 1;
240 ++pending_lang_change;
242 #endif
245 else if (new_map->reason == LC_LEAVE)
247 #ifndef NO_IMPLICIT_EXTERN_C
248 if (c_header_level && --c_header_level == 0)
250 if (new_map->sysp == 2)
251 warning ("badly nested C headers from preprocessor");
252 --pending_lang_change;
254 #endif
255 pop_srcloc ();
257 (*debug_hooks->end_source_file) (new_map->to_line);
260 update_header_times (new_map->to_file);
261 in_system_header = new_map->sysp != 0;
262 input_filename = new_map->to_file;
263 input_line = new_map->to_line;
264 map = new_map;
266 /* Hook for C++. */
267 extract_interface_info ();
270 static void
271 cb_def_pragma (cpp_reader *pfile, unsigned int line)
273 /* Issue a warning message if we have been asked to do so. Ignore
274 unknown pragmas in system headers unless an explicit
275 -Wunknown-pragmas has been given. */
276 if (warn_unknown_pragmas > in_system_header)
278 const unsigned char *space, *name;
279 const cpp_token *s;
281 space = name = (const unsigned char *) "";
282 s = cpp_get_token (pfile);
283 if (s->type != CPP_EOF)
285 space = cpp_token_as_text (pfile, s);
286 s = cpp_get_token (pfile);
287 if (s->type == CPP_NAME)
288 name = cpp_token_as_text (pfile, s);
291 input_line = SOURCE_LINE (map, line);
292 warning ("ignoring #pragma %s %s", space, name);
296 /* #define callback for DWARF and DWARF2 debug info. */
297 static void
298 cb_define (cpp_reader *pfile, unsigned int line, cpp_hashnode *node)
300 (*debug_hooks->define) (SOURCE_LINE (map, line),
301 (const char *) cpp_macro_definition (pfile, node));
304 /* #undef callback for DWARF and DWARF2 debug info. */
305 static void
306 cb_undef (cpp_reader *pfile ATTRIBUTE_UNUSED, unsigned int line,
307 cpp_hashnode *node)
309 (*debug_hooks->undef) (SOURCE_LINE (map, line),
310 (const char *) NODE_NAME (node));
313 static inline const cpp_token *
314 get_nonpadding_token (void)
316 const cpp_token *tok;
317 timevar_push (TV_CPP);
319 tok = cpp_get_token (parse_in);
320 while (tok->type == CPP_PADDING);
321 timevar_pop (TV_CPP);
323 return tok;
327 c_lex (tree *value)
329 const cpp_token *tok;
330 location_t atloc;
331 static bool no_more_pch;
333 retry:
334 tok = get_nonpadding_token ();
336 retry_after_at:
337 switch (tok->type)
339 case CPP_NAME:
340 *value = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node));
341 break;
343 case CPP_NUMBER:
345 unsigned int flags = cpp_classify_number (parse_in, tok);
347 switch (flags & CPP_N_CATEGORY)
349 case CPP_N_INVALID:
350 /* cpplib has issued an error. */
351 *value = error_mark_node;
352 break;
354 case CPP_N_INTEGER:
355 *value = interpret_integer (tok, flags);
356 break;
358 case CPP_N_FLOATING:
359 *value = interpret_float (tok, flags);
360 break;
362 default:
363 abort ();
366 break;
368 case CPP_ATSIGN:
369 /* An @ may give the next token special significance in Objective-C. */
370 atloc = input_location;
371 tok = get_nonpadding_token ();
372 if (c_dialect_objc ())
374 tree val;
375 switch (tok->type)
377 case CPP_NAME:
378 val = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node));
379 if (C_IS_RESERVED_WORD (val)
380 && OBJC_IS_AT_KEYWORD (C_RID_CODE (val)))
382 *value = val;
383 return CPP_AT_NAME;
385 break;
387 case CPP_STRING:
388 case CPP_WSTRING:
389 return lex_string (tok, value, true);
391 default: break;
395 /* ... or not. */
396 error ("%Hstray '@' in program", &atloc);
397 goto retry_after_at;
399 case CPP_OTHER:
401 cppchar_t c = tok->val.str.text[0];
403 if (c == '"' || c == '\'')
404 error ("missing terminating %c character", (int) c);
405 else if (ISGRAPH (c))
406 error ("stray '%c' in program", (int) c);
407 else
408 error ("stray '\\%o' in program", (int) c);
410 goto retry;
412 case CPP_CHAR:
413 case CPP_WCHAR:
414 *value = lex_charconst (tok);
415 break;
417 case CPP_STRING:
418 case CPP_WSTRING:
419 return lex_string (tok, value, false);
420 break;
422 /* These tokens should not be visible outside cpplib. */
423 case CPP_HEADER_NAME:
424 case CPP_COMMENT:
425 case CPP_MACRO_ARG:
426 abort ();
428 default:
429 *value = NULL_TREE;
430 break;
433 if (! no_more_pch)
435 no_more_pch = true;
436 c_common_no_more_pch ();
439 return tok->type;
442 /* Returns the narrowest C-visible unsigned type, starting with the
443 minimum specified by FLAGS, that can fit VALUE, or itk_none if
444 there isn't one. */
445 static enum integer_type_kind
446 narrowest_unsigned_type (tree value, unsigned int flags)
448 enum integer_type_kind itk;
450 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
451 itk = itk_unsigned_int;
452 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
453 itk = itk_unsigned_long;
454 else
455 itk = itk_unsigned_long_long;
457 /* int_fits_type_p must think the type of its first argument is
458 wider than its second argument, or it won't do the proper check. */
459 TREE_TYPE (value) = widest_unsigned_literal_type_node;
461 for (; itk < itk_none; itk += 2 /* skip unsigned types */)
462 if (int_fits_type_p (value, integer_types[itk]))
463 return itk;
465 return itk_none;
468 /* Ditto, but narrowest signed type. */
469 static enum integer_type_kind
470 narrowest_signed_type (tree value, unsigned int flags)
472 enum integer_type_kind itk;
474 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
475 itk = itk_int;
476 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
477 itk = itk_long;
478 else
479 itk = itk_long_long;
481 /* int_fits_type_p must think the type of its first argument is
482 wider than its second argument, or it won't do the proper check. */
483 TREE_TYPE (value) = widest_unsigned_literal_type_node;
485 for (; itk < itk_none; itk += 2 /* skip signed types */)
486 if (int_fits_type_p (value, integer_types[itk]))
487 return itk;
489 return itk_none;
492 /* Interpret TOKEN, an integer with FLAGS as classified by cpplib. */
493 static tree
494 interpret_integer (const cpp_token *token, unsigned int flags)
496 tree value, type;
497 enum integer_type_kind itk;
498 cpp_num integer;
499 cpp_options *options = cpp_get_options (parse_in);
501 integer = cpp_interpret_integer (parse_in, token, flags);
502 integer = cpp_num_sign_extend (integer, options->precision);
503 value = build_int_2_wide (integer.low, integer.high);
505 /* The type of a constant with a U suffix is straightforward. */
506 if (flags & CPP_N_UNSIGNED)
507 itk = narrowest_unsigned_type (value, flags);
508 else
510 /* The type of a potentially-signed integer constant varies
511 depending on the base it's in, the standard in use, and the
512 length suffixes. */
513 enum integer_type_kind itk_u = narrowest_unsigned_type (value, flags);
514 enum integer_type_kind itk_s = narrowest_signed_type (value, flags);
516 /* In both C89 and C99, octal and hex constants may be signed or
517 unsigned, whichever fits tighter. We do not warn about this
518 choice differing from the traditional choice, as the constant
519 is probably a bit pattern and either way will work. */
520 if ((flags & CPP_N_RADIX) != CPP_N_DECIMAL)
521 itk = MIN (itk_u, itk_s);
522 else
524 /* In C99, decimal constants are always signed.
525 In C89, decimal constants that don't fit in long have
526 undefined behavior; we try to make them unsigned long.
527 In GCC's extended C89, that last is true of decimal
528 constants that don't fit in long long, too. */
530 itk = itk_s;
531 if (itk_s > itk_u && itk_s > itk_long)
533 if (!flag_isoc99)
535 if (itk_u < itk_unsigned_long)
536 itk_u = itk_unsigned_long;
537 itk = itk_u;
538 warning ("this decimal constant is unsigned only in ISO C90");
540 else if (warn_traditional)
541 warning ("this decimal constant would be unsigned in ISO C90");
546 if (itk == itk_none)
547 /* cpplib has already issued a warning for overflow. */
548 type = ((flags & CPP_N_UNSIGNED)
549 ? widest_unsigned_literal_type_node
550 : widest_integer_literal_type_node);
551 else
552 type = integer_types[itk];
554 if (itk > itk_unsigned_long
555 && (flags & CPP_N_WIDTH) != CPP_N_LARGE
556 && ! in_system_header && ! flag_isoc99)
557 pedwarn ("integer constant is too large for \"%s\" type",
558 (flags & CPP_N_UNSIGNED) ? "unsigned long" : "long");
560 TREE_TYPE (value) = type;
562 /* Convert imaginary to a complex type. */
563 if (flags & CPP_N_IMAGINARY)
564 value = build_complex (NULL_TREE, convert (type, integer_zero_node), value);
566 return value;
569 /* Interpret TOKEN, a floating point number with FLAGS as classified
570 by cpplib. */
571 static tree
572 interpret_float (const cpp_token *token, unsigned int flags)
574 tree type;
575 tree value;
576 REAL_VALUE_TYPE real;
577 char *copy;
578 size_t copylen;
579 const char *typename;
581 /* FIXME: make %T work in error/warning, then we don't need typename. */
582 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
584 type = long_double_type_node;
585 typename = "long double";
587 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL
588 || flag_single_precision_constant)
590 type = float_type_node;
591 typename = "float";
593 else
595 type = double_type_node;
596 typename = "double";
599 /* Copy the constant to a nul-terminated buffer. If the constant
600 has any suffixes, cut them off; REAL_VALUE_ATOF/ REAL_VALUE_HTOF
601 can't handle them. */
602 copylen = token->val.str.len;
603 if ((flags & CPP_N_WIDTH) != CPP_N_MEDIUM)
604 /* Must be an F or L suffix. */
605 copylen--;
606 if (flags & CPP_N_IMAGINARY)
607 /* I or J suffix. */
608 copylen--;
610 copy = alloca (copylen + 1);
611 memcpy (copy, token->val.str.text, copylen);
612 copy[copylen] = '\0';
614 real_from_string (&real, copy);
615 real_convert (&real, TYPE_MODE (type), &real);
617 /* A diagnostic is required for "soft" overflow by some ISO C
618 testsuites. This is not pedwarn, because some people don't want
619 an error for this.
620 ??? That's a dubious reason... is this a mandatory diagnostic or
621 isn't it? -- zw, 2001-08-21. */
622 if (REAL_VALUE_ISINF (real) && pedantic)
623 warning ("floating constant exceeds range of \"%s\"", typename);
625 /* Create a node with determined type and value. */
626 value = build_real (type, real);
627 if (flags & CPP_N_IMAGINARY)
628 value = build_complex (NULL_TREE, convert (type, integer_zero_node), value);
630 return value;
633 /* Convert a series of STRING and/or WSTRING tokens into a tree,
634 performing string constant concatenation. TOK is the first of
635 these. VALP is the location to write the string into. OBJC_STRING
636 indicates whether an '@' token preceded the incoming token.
637 Returns the CPP token type of the result (CPP_STRING, CPP_WSTRING,
638 or CPP_OBJC_STRING).
640 This is unfortunately more work than it should be. If any of the
641 strings in the series has an L prefix, the result is a wide string
642 (6.4.5p4). Whether or not the result is a wide string affects the
643 meaning of octal and hexadecimal escapes (6.4.4.4p6,9). But escape
644 sequences do not continue across the boundary between two strings in
645 a series (6.4.5p7), so we must not lose the boundaries. Therefore
646 cpp_interpret_string takes a vector of cpp_string structures, which
647 we must arrange to provide. */
649 static enum cpp_ttype
650 lex_string (const cpp_token *tok, tree *valp, bool objc_string)
652 tree value;
653 bool wide = false;
654 size_t count = 1;
655 struct obstack str_ob;
656 cpp_string istr;
658 /* Try to avoid the overhead of creating and destroying an obstack
659 for the common case of just one string. */
660 cpp_string str = tok->val.str;
661 cpp_string *strs = &str;
663 if (tok->type == CPP_WSTRING)
664 wide = true;
666 tok = get_nonpadding_token ();
667 if (c_dialect_objc () && tok->type == CPP_ATSIGN)
669 objc_string = true;
670 tok = get_nonpadding_token ();
672 if (tok->type == CPP_STRING || tok->type == CPP_WSTRING)
674 gcc_obstack_init (&str_ob);
675 obstack_grow (&str_ob, &str, sizeof (cpp_string));
679 count++;
680 if (tok->type == CPP_WSTRING)
681 wide = true;
682 obstack_grow (&str_ob, &tok->val.str, sizeof (cpp_string));
684 tok = get_nonpadding_token ();
685 if (c_dialect_objc () && tok->type == CPP_ATSIGN)
687 objc_string = true;
688 tok = get_nonpadding_token ();
691 while (tok->type == CPP_STRING || tok->type == CPP_WSTRING);
692 strs = obstack_finish (&str_ob);
695 /* We have read one more token than we want. */
696 _cpp_backup_tokens (parse_in, 1);
698 if (count > 1 && !objc_string && warn_traditional && !in_system_header)
699 warning ("traditional C rejects string constant concatenation");
701 if (cpp_interpret_string (parse_in, strs, count, &istr, wide))
703 value = build_string (istr.len, (char *)istr.text);
704 free ((void *)istr.text);
706 else
708 /* Callers cannot generally handle error_mark_node in this context,
709 so return the empty string instead. cpp_interpret_string has
710 issued an error. */
711 if (wide)
712 value = build_string (TYPE_PRECISION (wchar_type_node)
713 / TYPE_PRECISION (char_type_node),
714 "\0\0\0"); /* widest supported wchar_t
715 is 32 bits */
716 else
717 value = build_string (1, "");
720 TREE_TYPE (value) = wide ? wchar_array_type_node : char_array_type_node;
721 *valp = fix_string_type (value);
723 if (strs != &str)
724 obstack_free (&str_ob, 0);
726 return objc_string ? CPP_OBJC_STRING : wide ? CPP_WSTRING : CPP_STRING;
729 /* Converts a (possibly wide) character constant token into a tree. */
730 static tree
731 lex_charconst (const cpp_token *token)
733 cppchar_t result;
734 tree type, value;
735 unsigned int chars_seen;
736 int unsignedp;
738 result = cpp_interpret_charconst (parse_in, token,
739 &chars_seen, &unsignedp);
741 /* Cast to cppchar_signed_t to get correct sign-extension of RESULT
742 before possibly widening to HOST_WIDE_INT for build_int_2. */
743 if (unsignedp || (cppchar_signed_t) result >= 0)
744 value = build_int_2 (result, 0);
745 else
746 value = build_int_2 ((cppchar_signed_t) result, -1);
748 if (token->type == CPP_WCHAR)
749 type = wchar_type_node;
750 /* In C, a character constant has type 'int'.
751 In C++ 'char', but multi-char charconsts have type 'int'. */
752 else if (!c_dialect_cxx () || chars_seen > 1)
753 type = integer_type_node;
754 else
755 type = char_type_node;
757 TREE_TYPE (value) = type;
758 return value;