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, 2004, 2005
4 Free Software Foundation, Inc.
6 This file is part of GCC.
8 GCC is free software; you can redistribute it and/or modify it under
9 the terms of the GNU General Public License as published by the Free
10 Software Foundation; either version 2, or (at your option) any later
13 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
14 WARRANTY; without even the implied warranty of MERCHANTABILITY or
15 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
18 You should have received a copy of the GNU General Public License
19 along with GCC; see the file COPYING. If not, write to the Free
20 Software Foundation, 51 Franklin Street, Fifth Floor, Boston, MA
25 #include "coretypes.h"
42 #include "splay-tree.h"
45 /* We may keep statistics about how long which files took to compile. */
46 static int header_time
, body_time
;
47 static splay_tree file_info_tree
;
49 int pending_lang_change
; /* If we need to switch languages - C++ only */
50 int c_header_level
; /* depth in C headers - C++ only */
52 /* If we need to translate characters received. This is tri-state:
53 0 means use only the untranslated string; 1 means use only
54 the translated string; -1 means chain the translated string
55 to the untranslated one. */
56 int c_lex_string_translate
= 1;
58 /* True if strings should be passed to the caller of c_lex completely
59 unmolested (no concatenation, no translation). */
60 bool c_lex_return_raw_strings
= false;
62 static tree
interpret_integer (const cpp_token
*, unsigned int);
63 static tree
interpret_float (const cpp_token
*, unsigned int);
64 static enum integer_type_kind narrowest_unsigned_type
65 (unsigned HOST_WIDE_INT
, unsigned HOST_WIDE_INT
, unsigned int);
66 static enum integer_type_kind narrowest_signed_type
67 (unsigned HOST_WIDE_INT
, unsigned HOST_WIDE_INT
, unsigned int);
68 static enum cpp_ttype
lex_string (const cpp_token
*, tree
*, bool);
69 static tree
lex_charconst (const cpp_token
*);
70 static void update_header_times (const char *);
71 static int dump_one_header (splay_tree_node
, void *);
72 static void cb_line_change (cpp_reader
*, const cpp_token
*, int);
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
*);
81 struct cpp_callbacks
*cb
;
82 struct c_fileinfo
*toplevel
;
84 /* The get_fileinfo data structure must be initialized before
85 cpp_read_main_file is called. */
86 toplevel
= get_fileinfo ("<top level>");
87 if (flag_detailed_statistics
)
90 body_time
= get_run_time ();
91 toplevel
->time
= body_time
;
94 cb
= cpp_get_callbacks (parse_in
);
96 cb
->line_change
= cb_line_change
;
98 cb
->def_pragma
= cb_def_pragma
;
99 cb
->valid_pch
= c_common_valid_pch
;
100 cb
->read_pch
= c_common_read_pch
;
102 /* Set the debug callbacks if we can use them. */
103 if (debug_info_level
== DINFO_LEVEL_VERBOSE
104 && (write_symbols
== DWARF2_DEBUG
105 || write_symbols
== VMS_AND_DWARF2_DEBUG
))
107 cb
->define
= cb_define
;
108 cb
->undef
= cb_undef
;
113 get_fileinfo (const char *name
)
116 struct c_fileinfo
*fi
;
119 file_info_tree
= splay_tree_new ((splay_tree_compare_fn
) strcmp
,
121 (splay_tree_delete_value_fn
) free
);
123 n
= splay_tree_lookup (file_info_tree
, (splay_tree_key
) name
);
125 return (struct c_fileinfo
*) n
->value
;
127 fi
= XNEW (struct c_fileinfo
);
129 fi
->interface_only
= 0;
130 fi
->interface_unknown
= 1;
131 splay_tree_insert (file_info_tree
, (splay_tree_key
) name
,
132 (splay_tree_value
) fi
);
137 update_header_times (const char *name
)
139 /* Changing files again. This means currently collected time
140 is charged against header time, and body time starts back at 0. */
141 if (flag_detailed_statistics
)
143 int this_time
= get_run_time ();
144 struct c_fileinfo
*file
= get_fileinfo (name
);
145 header_time
+= this_time
- body_time
;
146 file
->time
+= this_time
- body_time
;
147 body_time
= this_time
;
152 dump_one_header (splay_tree_node n
, void * ARG_UNUSED (dummy
))
154 print_time ((const char *) n
->key
,
155 ((struct c_fileinfo
*) n
->value
)->time
);
160 dump_time_statistics (void)
162 struct c_fileinfo
*file
= get_fileinfo (input_filename
);
163 int this_time
= get_run_time ();
164 file
->time
+= this_time
- body_time
;
166 fprintf (stderr
, "\n******\n");
167 print_time ("header files (total)", header_time
);
168 print_time ("main file (total)", this_time
- body_time
);
169 fprintf (stderr
, "ratio = %g : 1\n",
170 (double) header_time
/ (double) (this_time
- body_time
));
171 fprintf (stderr
, "\n******\n");
173 splay_tree_foreach (file_info_tree
, dump_one_header
, 0);
177 cb_ident (cpp_reader
* ARG_UNUSED (pfile
),
178 unsigned int ARG_UNUSED (line
),
179 const cpp_string
* ARG_UNUSED (str
))
181 #ifdef ASM_OUTPUT_IDENT
184 /* Convert escapes in the string. */
185 cpp_string cstr
= { 0, 0 };
186 if (cpp_interpret_string (pfile
, str
, 1, &cstr
, false))
188 ASM_OUTPUT_IDENT (asm_out_file
, (const char *) cstr
.text
);
189 free ((void *) cstr
.text
);
195 /* Called at the start of every non-empty line. TOKEN is the first
196 lexed token on the line. Used for diagnostic line numbers. */
198 cb_line_change (cpp_reader
* ARG_UNUSED (pfile
), const cpp_token
*token
,
201 if (token
->type
!= CPP_EOF
&& !parsing_args
)
202 #ifdef USE_MAPPED_LOCATION
203 input_location
= token
->src_loc
;
206 source_location loc
= token
->src_loc
;
207 const struct line_map
*map
= linemap_lookup (&line_table
, loc
);
208 input_line
= SOURCE_LINE (map
, loc
);
214 fe_file_change (const struct line_map
*new_map
)
219 if (new_map
->reason
== LC_ENTER
)
221 /* Don't stack the main buffer on the input stack;
222 we already did in compile_file. */
223 if (!MAIN_FILE_P (new_map
))
225 #ifdef USE_MAPPED_LOCATION
226 int included_at
= LAST_SOURCE_LINE_LOCATION (new_map
- 1);
228 input_location
= included_at
;
229 push_srcloc (new_map
->start_location
);
231 int included_at
= LAST_SOURCE_LINE (new_map
- 1);
233 input_line
= included_at
;
234 push_srcloc (new_map
->to_file
, 1);
236 (*debug_hooks
->start_source_file
) (included_at
, new_map
->to_file
);
237 #ifndef NO_IMPLICIT_EXTERN_C
240 else if (new_map
->sysp
== 2)
243 ++pending_lang_change
;
248 else if (new_map
->reason
== LC_LEAVE
)
250 #ifndef NO_IMPLICIT_EXTERN_C
251 if (c_header_level
&& --c_header_level
== 0)
253 if (new_map
->sysp
== 2)
254 warning (0, "badly nested C headers from preprocessor");
255 --pending_lang_change
;
260 (*debug_hooks
->end_source_file
) (new_map
->to_line
);
263 update_header_times (new_map
->to_file
);
264 in_system_header
= new_map
->sysp
!= 0;
265 #ifdef USE_MAPPED_LOCATION
266 input_location
= new_map
->start_location
;
268 input_filename
= new_map
->to_file
;
269 input_line
= new_map
->to_line
;
274 cb_def_pragma (cpp_reader
*pfile
, source_location loc
)
276 /* Issue a warning message if we have been asked to do so. Ignore
277 unknown pragmas in system headers unless an explicit
278 -Wunknown-pragmas has been given. */
279 if (warn_unknown_pragmas
> in_system_header
)
281 const unsigned char *space
, *name
;
283 #ifndef USE_MAPPED_LOCATION
285 const struct line_map
*map
= linemap_lookup (&line_table
, loc
);
286 fe_loc
.file
= map
->to_file
;
287 fe_loc
.line
= SOURCE_LINE (map
, loc
);
289 location_t fe_loc
= loc
;
292 space
= name
= (const unsigned char *) "";
293 s
= cpp_get_token (pfile
);
294 if (s
->type
!= CPP_EOF
)
296 space
= cpp_token_as_text (pfile
, s
);
297 s
= cpp_get_token (pfile
);
298 if (s
->type
== CPP_NAME
)
299 name
= cpp_token_as_text (pfile
, s
);
302 warning (OPT_Wunknown_pragmas
, "%Hignoring #pragma %s %s",
303 &fe_loc
, space
, name
);
307 /* #define callback for DWARF and DWARF2 debug info. */
309 cb_define (cpp_reader
*pfile
, source_location loc
, cpp_hashnode
*node
)
311 const struct line_map
*map
= linemap_lookup (&line_table
, loc
);
312 (*debug_hooks
->define
) (SOURCE_LINE (map
, loc
),
313 (const char *) cpp_macro_definition (pfile
, node
));
316 /* #undef callback for DWARF and DWARF2 debug info. */
318 cb_undef (cpp_reader
* ARG_UNUSED (pfile
), source_location loc
,
321 const struct line_map
*map
= linemap_lookup (&line_table
, loc
);
322 (*debug_hooks
->undef
) (SOURCE_LINE (map
, loc
),
323 (const char *) NODE_NAME (node
));
326 /* Read a token and return its type. Fill *VALUE with its value, if
327 applicable. Fill *CPP_FLAGS with the token's flags, if it is
331 c_lex_with_flags (tree
*value
, location_t
*loc
, unsigned char *cpp_flags
)
333 static bool no_more_pch
;
334 const cpp_token
*tok
;
337 timevar_push (TV_CPP
);
339 tok
= cpp_get_token (parse_in
);
343 #ifdef USE_MAPPED_LOCATION
346 *loc
= input_location
;
354 *value
= HT_IDENT_TO_GCC_IDENT (HT_NODE (tok
->val
.node
));
359 unsigned int flags
= cpp_classify_number (parse_in
, tok
);
361 switch (flags
& CPP_N_CATEGORY
)
364 /* cpplib has issued an error. */
365 *value
= error_mark_node
;
370 *value
= interpret_integer (tok
, flags
);
374 *value
= interpret_float (tok
, flags
);
384 /* An @ may give the next token special significance in Objective-C. */
385 if (c_dialect_objc ())
387 location_t atloc
= input_location
;
390 tok
= cpp_get_token (parse_in
);
399 type
= lex_string (tok
, value
, true);
403 *value
= HT_IDENT_TO_GCC_IDENT (HT_NODE (tok
->val
.node
));
404 if (objc_is_reserved_word (*value
))
413 error ("%Hstray %<@%> in program", &atloc
);
423 unsigned char name
[4];
425 *cpp_spell_token (parse_in
, tok
, name
, true) = 0;
427 error ("stray %qs in program", name
);
434 cppchar_t c
= tok
->val
.str
.text
[0];
436 if (c
== '"' || c
== '\'')
437 error ("missing terminating %c character", (int) c
);
438 else if (ISGRAPH (c
))
439 error ("stray %qc in program", (int) c
);
441 error ("stray %<\\%o%> in program", (int) c
);
447 *value
= lex_charconst (tok
);
452 if (!c_lex_return_raw_strings
)
454 type
= lex_string (tok
, value
, false);
461 *value
= build_string (tok
->val
.str
.len
, (char *) tok
->val
.str
.text
);
464 /* These tokens should not be visible outside cpplib. */
465 case CPP_HEADER_NAME
:
476 *cpp_flags
= tok
->flags
;
481 c_common_no_more_pch ();
484 timevar_pop (TV_CPP
);
493 return c_lex_with_flags (value
, &loc
, NULL
);
496 /* Returns the narrowest C-visible unsigned type, starting with the
497 minimum specified by FLAGS, that can fit HIGH:LOW, or itk_none if
500 static enum integer_type_kind
501 narrowest_unsigned_type (unsigned HOST_WIDE_INT low
,
502 unsigned HOST_WIDE_INT high
,
505 enum integer_type_kind itk
;
507 if ((flags
& CPP_N_WIDTH
) == CPP_N_SMALL
)
508 itk
= itk_unsigned_int
;
509 else if ((flags
& CPP_N_WIDTH
) == CPP_N_MEDIUM
)
510 itk
= itk_unsigned_long
;
512 itk
= itk_unsigned_long_long
;
514 for (; itk
< itk_none
; itk
+= 2 /* skip unsigned types */)
516 tree upper
= TYPE_MAX_VALUE (integer_types
[itk
]);
518 if ((unsigned HOST_WIDE_INT
) TREE_INT_CST_HIGH (upper
) > high
519 || ((unsigned HOST_WIDE_INT
) TREE_INT_CST_HIGH (upper
) == high
520 && TREE_INT_CST_LOW (upper
) >= low
))
527 /* Ditto, but narrowest signed type. */
528 static enum integer_type_kind
529 narrowest_signed_type (unsigned HOST_WIDE_INT low
,
530 unsigned HOST_WIDE_INT high
, unsigned int flags
)
532 enum integer_type_kind itk
;
534 if ((flags
& CPP_N_WIDTH
) == CPP_N_SMALL
)
536 else if ((flags
& CPP_N_WIDTH
) == CPP_N_MEDIUM
)
542 for (; itk
< itk_none
; itk
+= 2 /* skip signed types */)
544 tree upper
= TYPE_MAX_VALUE (integer_types
[itk
]);
546 if ((unsigned HOST_WIDE_INT
) TREE_INT_CST_HIGH (upper
) > high
547 || ((unsigned HOST_WIDE_INT
) TREE_INT_CST_HIGH (upper
) == high
548 && TREE_INT_CST_LOW (upper
) >= low
))
555 /* Interpret TOKEN, an integer with FLAGS as classified by cpplib. */
557 interpret_integer (const cpp_token
*token
, unsigned int flags
)
560 enum integer_type_kind itk
;
562 cpp_options
*options
= cpp_get_options (parse_in
);
564 integer
= cpp_interpret_integer (parse_in
, token
, flags
);
565 integer
= cpp_num_sign_extend (integer
, options
->precision
);
567 /* The type of a constant with a U suffix is straightforward. */
568 if (flags
& CPP_N_UNSIGNED
)
569 itk
= narrowest_unsigned_type (integer
.low
, integer
.high
, flags
);
572 /* The type of a potentially-signed integer constant varies
573 depending on the base it's in, the standard in use, and the
575 enum integer_type_kind itk_u
576 = narrowest_unsigned_type (integer
.low
, integer
.high
, flags
);
577 enum integer_type_kind itk_s
578 = narrowest_signed_type (integer
.low
, integer
.high
, flags
);
580 /* In both C89 and C99, octal and hex constants may be signed or
581 unsigned, whichever fits tighter. We do not warn about this
582 choice differing from the traditional choice, as the constant
583 is probably a bit pattern and either way will work. */
584 if ((flags
& CPP_N_RADIX
) != CPP_N_DECIMAL
)
585 itk
= MIN (itk_u
, itk_s
);
588 /* In C99, decimal constants are always signed.
589 In C89, decimal constants that don't fit in long have
590 undefined behavior; we try to make them unsigned long.
591 In GCC's extended C89, that last is true of decimal
592 constants that don't fit in long long, too. */
595 if (itk_s
> itk_u
&& itk_s
> itk_long
)
599 if (itk_u
< itk_unsigned_long
)
600 itk_u
= itk_unsigned_long
;
602 warning (0, "this decimal constant is unsigned only in ISO C90");
605 warning (OPT_Wtraditional
,
606 "this decimal constant would be unsigned in ISO C90");
612 /* cpplib has already issued a warning for overflow. */
613 type
= ((flags
& CPP_N_UNSIGNED
)
614 ? widest_unsigned_literal_type_node
615 : widest_integer_literal_type_node
);
617 type
= integer_types
[itk
];
619 if (itk
> itk_unsigned_long
620 && (flags
& CPP_N_WIDTH
) != CPP_N_LARGE
621 && !in_system_header
&& !flag_isoc99
)
622 pedwarn ("integer constant is too large for %qs type",
623 (flags
& CPP_N_UNSIGNED
) ? "unsigned long" : "long");
625 value
= build_int_cst_wide (type
, integer
.low
, integer
.high
);
627 /* Convert imaginary to a complex type. */
628 if (flags
& CPP_N_IMAGINARY
)
629 value
= build_complex (NULL_TREE
, build_int_cst (type
, 0), value
);
634 /* Interpret TOKEN, a floating point number with FLAGS as classified
637 interpret_float (const cpp_token
*token
, unsigned int flags
)
641 REAL_VALUE_TYPE real
;
644 const char *type_name
;
646 /* FIXME: make %T work in error/warning, then we don't need type_name. */
647 if ((flags
& CPP_N_WIDTH
) == CPP_N_LARGE
)
649 type
= long_double_type_node
;
650 type_name
= "long double";
652 else if ((flags
& CPP_N_WIDTH
) == CPP_N_SMALL
653 || flag_single_precision_constant
)
655 type
= float_type_node
;
660 type
= double_type_node
;
661 type_name
= "double";
664 /* Copy the constant to a nul-terminated buffer. If the constant
665 has any suffixes, cut them off; REAL_VALUE_ATOF/ REAL_VALUE_HTOF
666 can't handle them. */
667 copylen
= token
->val
.str
.len
;
668 if ((flags
& CPP_N_WIDTH
) != CPP_N_MEDIUM
)
669 /* Must be an F or L suffix. */
671 if (flags
& CPP_N_IMAGINARY
)
675 copy
= (char *) alloca (copylen
+ 1);
676 memcpy (copy
, token
->val
.str
.text
, copylen
);
677 copy
[copylen
] = '\0';
679 real_from_string (&real
, copy
);
680 real_convert (&real
, TYPE_MODE (type
), &real
);
682 /* Both C and C++ require a diagnostic for a floating constant
683 outside the range of representable values of its type. Since we
684 have __builtin_inf* to produce an infinity, it might now be
685 appropriate for this to be a mandatory pedwarn rather than
686 conditioned on -pedantic. */
687 if (REAL_VALUE_ISINF (real
) && pedantic
)
688 pedwarn ("floating constant exceeds range of %<%s%>", type_name
);
690 /* Create a node with determined type and value. */
691 value
= build_real (type
, real
);
692 if (flags
& CPP_N_IMAGINARY
)
693 value
= build_complex (NULL_TREE
, convert (type
, integer_zero_node
), value
);
698 /* Convert a series of STRING and/or WSTRING tokens into a tree,
699 performing string constant concatenation. TOK is the first of
700 these. VALP is the location to write the string into. OBJC_STRING
701 indicates whether an '@' token preceded the incoming token.
702 Returns the CPP token type of the result (CPP_STRING, CPP_WSTRING,
705 This is unfortunately more work than it should be. If any of the
706 strings in the series has an L prefix, the result is a wide string
707 (6.4.5p4). Whether or not the result is a wide string affects the
708 meaning of octal and hexadecimal escapes (6.4.4.4p6,9). But escape
709 sequences do not continue across the boundary between two strings in
710 a series (6.4.5p7), so we must not lose the boundaries. Therefore
711 cpp_interpret_string takes a vector of cpp_string structures, which
712 we must arrange to provide. */
714 static enum cpp_ttype
715 lex_string (const cpp_token
*tok
, tree
*valp
, bool objc_string
)
720 struct obstack str_ob
;
723 /* Try to avoid the overhead of creating and destroying an obstack
724 for the common case of just one string. */
725 cpp_string str
= tok
->val
.str
;
726 cpp_string
*strs
= &str
;
728 if (tok
->type
== CPP_WSTRING
)
732 tok
= cpp_get_token (parse_in
);
738 if (c_dialect_objc ())
755 gcc_obstack_init (&str_ob
);
756 obstack_grow (&str_ob
, &str
, sizeof (cpp_string
));
760 obstack_grow (&str_ob
, &tok
->val
.str
, sizeof (cpp_string
));
764 /* We have read one more token than we want. */
765 _cpp_backup_tokens (parse_in
, 1);
767 strs
= XOBFINISH (&str_ob
, cpp_string
*);
769 if (concats
&& !objc_string
&& !in_system_header
)
770 warning (OPT_Wtraditional
,
771 "traditional C rejects string constant concatenation");
773 if ((c_lex_string_translate
774 ? cpp_interpret_string
: cpp_interpret_string_notranslate
)
775 (parse_in
, strs
, concats
+ 1, &istr
, wide
))
777 value
= build_string (istr
.len
, (char *) istr
.text
);
778 free ((void *) istr
.text
);
780 if (c_lex_string_translate
== -1)
782 int xlated
= cpp_interpret_string_notranslate (parse_in
, strs
,
785 /* Assume that, if we managed to translate the string above,
786 then the untranslated parsing will always succeed. */
789 if (TREE_STRING_LENGTH (value
) != (int) istr
.len
790 || 0 != strncmp (TREE_STRING_POINTER (value
), (char *) istr
.text
,
793 /* Arrange for us to return the untranslated string in
794 *valp, but to set up the C type of the translated
796 *valp
= build_string (istr
.len
, (char *) istr
.text
);
797 valp
= &TREE_CHAIN (*valp
);
799 free ((void *) istr
.text
);
804 /* Callers cannot generally handle error_mark_node in this context,
805 so return the empty string instead. cpp_interpret_string has
808 value
= build_string (TYPE_PRECISION (wchar_type_node
)
809 / TYPE_PRECISION (char_type_node
),
810 "\0\0\0"); /* widest supported wchar_t
813 value
= build_string (1, "");
816 TREE_TYPE (value
) = wide
? wchar_array_type_node
: char_array_type_node
;
817 *valp
= fix_string_type (value
);
820 obstack_free (&str_ob
, 0);
822 return objc_string
? CPP_OBJC_STRING
: wide
? CPP_WSTRING
: CPP_STRING
;
825 /* Converts a (possibly wide) character constant token into a tree. */
827 lex_charconst (const cpp_token
*token
)
831 unsigned int chars_seen
;
834 result
= cpp_interpret_charconst (parse_in
, token
,
835 &chars_seen
, &unsignedp
);
837 if (token
->type
== CPP_WCHAR
)
838 type
= wchar_type_node
;
839 /* In C, a character constant has type 'int'.
840 In C++ 'char', but multi-char charconsts have type 'int'. */
841 else if (!c_dialect_cxx () || chars_seen
> 1)
842 type
= integer_type_node
;
844 type
= char_type_node
;
846 /* Cast to cppchar_signed_t to get correct sign-extension of RESULT
847 before possibly widening to HOST_WIDE_INT for build_int_cst. */
848 if (unsignedp
|| (cppchar_signed_t
) result
>= 0)
849 value
= build_int_cst_wide (type
, result
, 0);
851 value
= build_int_cst_wide (type
, (cppchar_signed_t
) result
, -1);