From 2c05d33f90861d074dc12808dafbde30f487b1a0 Mon Sep 17 00:00:00 2001 From: Ulrich Drepper Date: Wed, 28 Sep 2005 17:33:18 +0000 Subject: [BZ #1302] 2005-09-06 Paul Eggert Ulrich Drepper [BZ #1302] Change bitset word type from unsigned int to unsigned long int, as this has better performance on typical 64-bit hosts. Change bitset type name to bitset_t. * posix/regcomp.c (build_equiv_class, build_charclass): (build_range_exp, build_collating_symbol): Prefer bitset_t to re_bitset_ptr_t in prototypes, when the actual argument is a bitset. This is merely a style issue, but it makes it clearer that an entire array is expected. (re_compile_fastmap_iter, init_dfa, init_word_char, optimize_subexps, lower_subexp): Adjust for new bitset_t definition. (lower_subexp, parse_bracket_exp, built_charclass_op): Likewise. * posix/regex_internal.h (bitset_set, bitset_clear, bitset_contain, bitset_not, bitset_merge, bitset_set_all, bitset_mask): Likewise. * posix/regexec.c (check_dst_limits_calc_pos_1, check_subexp_matching_top, build_trtable, group_nodes_into_DFAstates): Likewise. * posix/regcomp.c (utf8_sb_map): Don't assume initializer == 0xffffffff. * posix/regex_internal.h (BITSET_WORD_BITS): Renamed from UINT_BITS. All uses changed. (BITSET_WORDS): Renamed from BITSET_UINTS. All uses changed. (bitset_word_t): New type, replacing 'unsigned int' for bitset uses. All uses changed. (BITSET_WORD_MAX): New macro. (bitset_set, bitset_clear, bitset_contain, bitset_empty, (bitset_set_all, bitset_copy): Adjust for bitset_t change. (bitset_empty, bitset_copy): Prefer sizeof (bitset_t) to multiplying it out ourselves. (bitset_not_merge): Remove; unused. (bitset_contain): Return bool, not unsigned int with one bit on. All callers changed. * posix/regexec.c (build_trtable): Don't assume bitset_t has no stricter alignment than re_node_set; do this by defining a new internal type struct dests_alloc and using it to allocate memory. --- posix/regex_internal.h | 74 +++++++++++++++++++++++++------------------------- 1 file changed, 37 insertions(+), 37 deletions(-) (limited to 'posix/regex_internal.h') diff --git a/posix/regex_internal.h b/posix/regex_internal.h index 0096bf7c91..681be1f52b 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -39,6 +39,9 @@ #if defined HAVE_WCTYPE_H || defined _LIBC # include #endif /* HAVE_WCTYPE_H || _LIBC */ +#if defined HAVE_STDBOOL_H || defined _LIBC +# include +#endif /* HAVE_STDBOOL_H || _LIBC */ #if defined _LIBC # include #else @@ -120,26 +123,31 @@ extern const char __re_error_msgid[] attribute_hidden; extern const size_t __re_error_msgid_idx[] attribute_hidden; -/* Number of bits in an unsinged int. */ -#define UINT_BITS (sizeof (unsigned int) * CHAR_BIT) -/* Number of unsigned int in an bit_set. */ -#define BITSET_UINTS ((SBC_MAX + UINT_BITS - 1) / UINT_BITS) -typedef unsigned int bitset[BITSET_UINTS]; -typedef unsigned int *re_bitset_ptr_t; -typedef const unsigned int *re_const_bitset_ptr_t; - -#define bitset_set(set,i) (set[i / UINT_BITS] |= 1u << i % UINT_BITS) -#define bitset_clear(set,i) (set[i / UINT_BITS] &= ~(1u << i % UINT_BITS)) -#define bitset_contain(set,i) (set[i / UINT_BITS] & (1u << i % UINT_BITS)) -#define bitset_empty(set) memset (set, 0, sizeof (unsigned int) * BITSET_UINTS) -#define bitset_set_all(set) \ - memset (set, 255, sizeof (unsigned int) * BITSET_UINTS) -#define bitset_copy(dest,src) \ - memcpy (dest, src, sizeof (unsigned int) * BITSET_UINTS) -static inline void bitset_not (bitset set); -static inline void bitset_merge (bitset dest, const bitset src); -static inline void bitset_not_merge (bitset dest, const bitset src); -static inline void bitset_mask (bitset dest, const bitset src); +/* An integer used to represent a set of bits. It must be unsigned, + and must be at least as wide as unsigned int. */ +typedef unsigned long int bitset_word_t; +/* All bits set in a bitset_word_t. */ +#define BITSET_WORD_MAX ULONG_MAX +/* Number of bits in a bitset_word_t. */ +#define BITSET_WORD_BITS (sizeof (bitset_word_t) * CHAR_BIT) +/* Number of bitset_word_t in a bit_set. */ +#define BITSET_WORDS (SBC_MAX / BITSET_WORD_BITS) +typedef bitset_word_t bitset_t[BITSET_WORDS]; +typedef bitset_word_t *re_bitset_ptr_t; +typedef const bitset_word_t *re_const_bitset_ptr_t; + +#define bitset_set(set,i) \ + (set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS) +#define bitset_clear(set,i) \ + (set[i / BITSET_WORD_BITS] &= ~((bitset_word_t) 1 << i % BITSET_WORD_BITS)) +#define bitset_contain(set,i) \ + (set[i / BITSET_WORD_BITS] & ((bitset_word_t) 1 << i % BITSET_WORD_BITS)) +#define bitset_empty(set) memset (set, '\0', sizeof (bitset_t)) +#define bitset_set_all(set) memset (set, '\xff', sizeof (bitset_t)) +#define bitset_copy(dest,src) memcpy (dest, src, sizeof (bitset_t)) +static inline void bitset_not (bitset_t set); +static inline void bitset_merge (bitset_t dest, const bitset_t src); +static inline void bitset_mask (bitset_t dest, const bitset_t src); #define PREV_WORD_CONSTRAINT 0x0001 #define PREV_NOTWORD_CONSTRAINT 0x0002 @@ -648,8 +656,8 @@ struct re_dfa_t int nbackref; /* The number of backreference in this dfa. */ /* Bitmap expressing which backreference is used. */ - unsigned int used_bkref_map; - unsigned int completed_bkref_map; + bitset_word_t used_bkref_map; + bitset_word_t completed_bkref_map; unsigned int has_plural_match : 1; /* If this dfa has "multibyte node", which is a backreference or @@ -660,7 +668,7 @@ struct re_dfa_t unsigned int map_notascii : 1; unsigned int word_ops_used : 1; int mb_cur_max; - bitset word_char; + bitset_t word_char; reg_syntax_t syntax; int *subexp_map; #ifdef DEBUG @@ -734,34 +742,26 @@ typedef struct /* Inline functions for bitset operation. */ static inline void -bitset_not (bitset set) +bitset_not (bitset_t set) { int bitset_i; - for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i) + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) set[bitset_i] = ~set[bitset_i]; } static inline void -bitset_merge (bitset dest, const bitset src) +bitset_merge (bitset_t dest, const bitset_t src) { int bitset_i; - for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i) + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) dest[bitset_i] |= src[bitset_i]; } static inline void -bitset_not_merge (bitset dest, const bitset src) -{ - int i; - for (i = 0; i < BITSET_UINTS; ++i) - dest[i] |= ~src[i]; -} - -static inline void -bitset_mask (bitset dest, const bitset src) +bitset_mask (bitset_t dest, const bitset_t src) { int bitset_i; - for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i) + for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) dest[bitset_i] &= src[bitset_i]; } -- cgit v1.2.3