From d58ab810a6e325cc351684d174c48cabce01bcc1 Mon Sep 17 00:00:00 2001 From: Wilco Dijkstra Date: Wed, 14 Dec 2016 15:12:18 +0000 Subject: Improve strtok and strtok_r performance. Instead of calling strpbrk which calls strcspn, call strcspn directly so we get the end of the token without an extra call to rawmemchr. Also avoid an unnecessary call to strcspn after the last token by adding an early exit for an empty string. Change strtok to tailcall strtok_r to avoid unnecessary code duplication. Remove the special header optimization for strtok_r of a 1-character constant string - both strspn and strcspn contain optimizations for this case. Benchmarking this showed similar performance in the worst case, but up to 5.5x better performance in the "found" case for large inputs. * benchtests/bench-strtok.c (oldstrtok): Add old implementation. * string/strtok.c (strtok): Change to tailcall __strtok_r. * string/strtok_r.c (__strtok_r): Optimize for performance. * string/string-inlines.c (__old_strtok_r_1c): New function. * string/bits/string2.h (__strtok_r): Move to string-inlines.c. --- string/bits/string2.h | 39 --------------------------------------- string/string-inlines.c | 30 ++++++++++++++++++++++++++++++ string/strtok.c | 38 +++----------------------------------- string/strtok_r.c | 31 ++++++++++++++++--------------- 4 files changed, 49 insertions(+), 89 deletions(-) (limited to 'string') diff --git a/string/bits/string2.h b/string/bits/string2.h index ca1eda9..e39d4f1 100644 --- a/string/bits/string2.h +++ b/string/bits/string2.h @@ -180,45 +180,6 @@ extern void *__rawmemchr (const void *__s, int __c); #endif -#if !defined _HAVE_STRING_ARCH_strtok_r || defined _FORCE_INLINES -# ifndef _HAVE_STRING_ARCH_strtok_r -# define __strtok_r(s, sep, nextp) \ - (__extension__ (__builtin_constant_p (sep) && __string2_1bptr_p (sep) \ - && ((const char *) (sep))[0] != '\0' \ - && ((const char *) (sep))[1] == '\0' \ - ? __strtok_r_1c (s, ((const char *) (sep))[0], nextp) \ - : __strtok_r (s, sep, nextp))) -# endif - -__STRING_INLINE char *__strtok_r_1c (char *__s, char __sep, char **__nextp); -__STRING_INLINE char * -__strtok_r_1c (char *__s, char __sep, char **__nextp) -{ - char *__result; - if (__s == NULL) - __s = *__nextp; - while (*__s == __sep) - ++__s; - __result = NULL; - if (*__s != '\0') - { - __result = __s++; - while (*__s != '\0') - if (*__s++ == __sep) - { - __s[-1] = '\0'; - break; - } - } - *__nextp = __s; - return __result; -} -# ifdef __USE_POSIX -# define strtok_r(s, sep, nextp) __strtok_r (s, sep, nextp) -# endif -#endif - - #if !defined _HAVE_STRING_ARCH_strsep || defined _FORCE_INLINES # ifndef _HAVE_STRING_ARCH_strsep diff --git a/string/string-inlines.c b/string/string-inlines.c index 1091468..d43e589 100644 --- a/string/string-inlines.c +++ b/string/string-inlines.c @@ -35,6 +35,36 @@ #include "shlib-compat.h" +#if SHLIB_COMPAT (libc, GLIBC_2_1_1, GLIBC_2_25) +/* The inline functions are not used from GLIBC 2.25 and forward, however + they are required to provide the symbols through string-inlines.c + (if inlining is not possible for compatibility reasons). */ + +char * +__old_strtok_r_1c (char *__s, char __sep, char **__nextp) +{ + char *__result; + if (__s == NULL) + __s = *__nextp; + while (*__s == __sep) + ++__s; + __result = NULL; + if (*__s != '\0') + { + __result = __s++; + while (*__s != '\0') + if (*__s++ == __sep) + { + __s[-1] = '\0'; + break; + } + } + *__nextp = __s; + return __result; +} +compat_symbol (libc, __old_strtok_r_1c, __strtok_r_1c, GLIBC_2_1_1); +#endif + #if SHLIB_COMPAT (libc, GLIBC_2_1_1, GLIBC_2_24) /* The inline functions are not used from GLIBC 2.24 and forward, however they are required to provide the symbols through string-inlines.c diff --git a/string/strtok.c b/string/strtok.c index 7a4574d..482cdc1 100644 --- a/string/strtok.c +++ b/string/strtok.c @@ -18,14 +18,6 @@ #include -static char *olds; - -#undef strtok - -#ifndef STRTOK -# define STRTOK strtok -#endif - /* Parse S into tokens separated by characters in DELIM. If S is NULL, the last string strtok() was called with is used. For example: @@ -36,32 +28,8 @@ static char *olds; // s = "abc\0=-def\0" */ char * -STRTOK (char *s, const char *delim) +strtok (char *s, const char *delim) { - char *token; - - if (s == NULL) - s = olds; - - /* Scan leading delimiters. */ - s += strspn (s, delim); - if (*s == '\0') - { - olds = s; - return NULL; - } - - /* Find the end of the token. */ - token = s; - s = strpbrk (token, delim); - if (s == NULL) - /* This token finishes the string. */ - olds = __rawmemchr (token, '\0'); - else - { - /* Terminate the token and make OLDS point past it. */ - *s = '\0'; - olds = s + 1; - } - return token; + static char *olds; + return __strtok_r (s, delim, &olds); } diff --git a/string/strtok_r.c b/string/strtok_r.c index f351304..2d251f9 100644 --- a/string/strtok_r.c +++ b/string/strtok_r.c @@ -22,14 +22,10 @@ #include -#undef strtok_r -#undef __strtok_r - #ifndef _LIBC /* Get specification. */ # include "strtok_r.h" # define __strtok_r strtok_r -# define __rawmemchr strchr #endif /* Parse S into tokens separated by characters in DELIM. @@ -45,11 +41,17 @@ char * __strtok_r (char *s, const char *delim, char **save_ptr) { - char *token; + char *end; if (s == NULL) s = *save_ptr; + if (*s == '\0') + { + *save_ptr = s; + return NULL; + } + /* Scan leading delimiters. */ s += strspn (s, delim); if (*s == '\0') @@ -59,18 +61,17 @@ __strtok_r (char *s, const char *delim, char **save_ptr) } /* Find the end of the token. */ - token = s; - s = strpbrk (token, delim); - if (s == NULL) - /* This token finishes the string. */ - *save_ptr = __rawmemchr (token, '\0'); - else + end = s + strcspn (s, delim); + if (*end == '\0') { - /* Terminate the token and make *SAVE_PTR point past it. */ - *s = '\0'; - *save_ptr = s + 1; + *save_ptr = end; + return s; } - return token; + + /* Terminate the token and make *SAVE_PTR point past it. */ + *end = '\0'; + *save_ptr = end + 1; + return s; } #ifdef weak_alias libc_hidden_def (__strtok_r) -- cgit v1.1