From patchwork Wed Sep 10 03:03:04 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Konstantin Serebryany X-Patchwork-Id: 387514 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from sourceware.org (server1.sourceware.org [209.132.180.131]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 2897A140187 for ; Wed, 10 Sep 2014 13:03:39 +1000 (EST) DomainKey-Signature: a=rsa-sha1; c=nofws; d=sourceware.org; h=list-id :list-unsubscribe:list-subscribe:list-archive:list-post :list-help:sender:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; q=dns; s=default; b=KJ+x LM3PpKSiiY+iMIRSvCBbQEHpLQtJvnk2FdUnhYdPbVG5E9TXdg48/RF8zxOsS20f U0rItGNxNrHfcjgSIuWkHFlecFsziOGItcFDo7SXdFTHhRQNQx5wL7LwC1M71kOI BD3MiGeOXdUp1bCX6ErVj9i4r5OKyhYWNPIgFnM= DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=sourceware.org; h=list-id :list-unsubscribe:list-subscribe:list-archive:list-post :list-help:sender:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type; s=default; bh=CT0v/GTF/G 6zlD0PGxX8GQWPB5U=; b=uVrvH7eP1lS137ZmWEm0QJ4RzeA9NYQGFewmXFaV75 H8He7Bk+Nf0uoBBnh2y3KSzanigaOiVhWHC118wTqxEIHW/s72blJ3cqypPhjTpu GTyUvVT9nB+bqKYnBbj0AmOCCWPRgvHDt/xRe8sRcBEXG5L7JBKhSiGaWINTtUjE s= Received: (qmail 16947 invoked by alias); 10 Sep 2014 03:03:32 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 16933 invoked by uid 89); 10 Sep 2014 03:03:30 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-2.4 required=5.0 tests=AWL, BAYES_00, FREEMAIL_FROM, RCVD_IN_DNSWL_LOW, SPF_PASS autolearn=ham version=3.3.2 X-HELO: mail-vc0-f175.google.com X-Received: by 10.221.45.132 with SMTP id uk4mr4262211vcb.57.1410318204538; Tue, 09 Sep 2014 20:03:24 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <20140909215726.B0AE92C39D8@topped-with-meat.com> References: <20140909215726.B0AE92C39D8@topped-with-meat.com> From: Konstantin Serebryany Date: Tue, 9 Sep 2014 20:03:04 -0700 Message-ID: Subject: Re: [PATCH] un-nest findidx() To: Roland McGrath Cc: Will Newton , Andreas Schwab , GNU C Library Here it is, verified against fresh trunk. 2014-09-09 Kostya Serebryany * locale/weight.h: add include guard. (findidx): un-nest, make it static inline, add parameters. * locale/weightwc.h: add include guard, rename findidx to findidxwc. (findidxwc): un-nest, make it static inline, add parameters. * posix/fnmatch_loop.c: include weightwc.h or weight.h depending on WIDE_CHAR_VERSION. Define FINDIDX as findidxwc or findidx. (FCT): change type of 'extra' to wint_t; do not include weight.h, un-nest calls to findidx. * posix/regcomp.c: include weight.h. (build_equiv_class): don't include weight.h, un-nest findidx. * posix/regex_internal.h: include weight.h (re_string_elem_size_at): don't include weight.h, un-nest findidx. * posix/regexec.c: include weight.h. (check_node_accept_bytes): don't include weight.h, un-nest findidx. * string/strcoll_l.c: define FINDIDX, include WEIGHT_H. (get_next_seq): don't include WEIGHT_H, un-nest findidx. (get_next_seq_nocache): don't include WEIGHT_H, un-nest findidx. * string/strxfrm_l.c: define FINDIDX, include WEIGHT_H. (STRXFRM): don't include WEIGHT_H, un-nest findidx. * wcsmbs/wcscoll_l.c: define FINDIDX. * wcsmbs/wcsxfrm_l.c: define FINDIDX. On Tue, Sep 9, 2014 at 2:57 PM, Roland McGrath wrote: > When it's been very long at all, you should post a fresh patch that's > verified against today's trunk, and not include any quoted nonsense in the > message. Just post in the same thread and that should be enough context > for everybody. diff --git a/locale/weight.h b/locale/weight.h index 9eb8ac6..485526c 100644 --- a/locale/weight.h +++ b/locale/weight.h @@ -16,10 +16,16 @@ License along with the GNU C Library; if not, see . */ +#ifndef _WEIGHT_H_ +#define _WEIGHT_H_ + /* Find index of weight. */ -auto inline int32_t +static inline int32_t __attribute ((always_inline)) -findidx (const unsigned char **cpp, size_t len) +findidx (const int32_t *table, + const int32_t *indirect, + const unsigned char *extra, + const unsigned char **cpp, size_t len) { int_fast32_t i = table[*(*cpp)++]; const unsigned char *cp; @@ -130,3 +136,5 @@ findidx (const unsigned char **cpp, size_t len) /* NOTREACHED */ return 0x43219876; } + +#endif /* weight.h */ diff --git a/locale/weightwc.h b/locale/weightwc.h index 8f047e3..3348544 100644 --- a/locale/weightwc.h +++ b/locale/weightwc.h @@ -16,10 +16,16 @@ License along with the GNU C Library; if not, see . */ +#ifndef _WEIGHTWC_H_ +#define _WEIGHTWC_H_ + /* Find index of weight. */ -auto inline int32_t +static inline int32_t __attribute ((always_inline)) -findidx (const wint_t **cpp, size_t len) +findidxwc (const int32_t *table, + const int32_t *indirect, + const wint_t *extra, + const wint_t **cpp, size_t len) { wint_t ch = *(*cpp)++; int32_t i = __collidx_table_lookup ((const char *) table, ch); @@ -109,3 +115,5 @@ findidx (const wint_t **cpp, size_t len) /* NOTREACHED */ return 0x43219876; } + +#endif /* weightwc.h */ diff --git a/posix/fnmatch_loop.c b/posix/fnmatch_loop.c index f79d051..8cd2f31 100644 --- a/posix/fnmatch_loop.c +++ b/posix/fnmatch_loop.c @@ -17,6 +17,17 @@ #include +# if WIDE_CHAR_VERSION +# include +# undef FINDIDX +# define FINDIDX findidxwc +# else +# include +# undef FINDIDX +# define FINDIDX findidx +# endif + + struct STRUCT { const CHAR *pattern; @@ -376,7 +387,7 @@ FCT (pattern, string, string_end, no_leading_period, flags, ends, alloca_used) const int32_t *table; # if WIDE_CHAR_VERSION const int32_t *weights; - const int32_t *extra; + const wint_t *extra; # else const unsigned char *weights; const unsigned char *extra; @@ -385,19 +396,12 @@ FCT (pattern, string, string_end, no_leading_period, flags, ends, alloca_used) int32_t idx; const UCHAR *cp = (const UCHAR *) str; - /* This #include defines a local function! */ -# if WIDE_CHAR_VERSION -# include -# else -# include -# endif - # if WIDE_CHAR_VERSION table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); weights = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); - extra = (const int32_t *) + extra = (const wint_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); @@ -412,7 +416,7 @@ FCT (pattern, string, string_end, no_leading_period, flags, ends, alloca_used) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); # endif - idx = findidx (&cp, 1); + idx = FINDIDX (table, indirect, extra, &cp, 1); if (idx != 0) { /* We found a table entry. Now see whether the @@ -422,7 +426,8 @@ FCT (pattern, string, string_end, no_leading_period, flags, ends, alloca_used) int32_t idx2; const UCHAR *np = (const UCHAR *) n; - idx2 = findidx (&np, string_end - n); + idx2 = FINDIDX (table, indirect, extra, + &np, string_end - n); if (idx2 != 0 && (idx >> 24) == (idx2 >> 24) && len == weights[idx2 & 0xffffff]) diff --git a/posix/regcomp.c b/posix/regcomp.c index 921d0f4..b1cc8bc 100644 --- a/posix/regcomp.c +++ b/posix/regcomp.c @@ -3389,6 +3389,8 @@ parse_bracket_symbol (bracket_elem_t *elem, re_string_t *regexp, return REG_NOERROR; } +#include + /* Helper function for parse_bracket_exp. Build the equivalence class which is represented by NAME. The result are written to MBCSET and SBCSET. @@ -3413,8 +3415,6 @@ build_equiv_class (bitset_t sbcset, const unsigned char *name) int32_t idx1, idx2; unsigned int ch; size_t len; - /* This #include defines a local function! */ -# include /* Calculate the index for equivalence class. */ cp = name; table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); @@ -3424,7 +3424,7 @@ build_equiv_class (bitset_t sbcset, const unsigned char *name) _NL_COLLATE_EXTRAMB); indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); - idx1 = findidx (&cp, -1); + idx1 = findidx (table, indirect, extra, &cp, -1); if (BE (idx1 == 0 || *cp != '\0', 0)) /* This isn't a valid character. */ return REG_ECOLLATE; @@ -3435,7 +3435,7 @@ build_equiv_class (bitset_t sbcset, const unsigned char *name) { char_buf[0] = ch; cp = char_buf; - idx2 = findidx (&cp, 1); + idx2 = findidx (table, indirect, extra, &cp, 1); /* idx2 = table[ch]; */ diff --git a/posix/regex_internal.h b/posix/regex_internal.h index 75c390f..c38980b 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -733,6 +733,8 @@ re_string_wchar_at (const re_string_t *pstr, int idx) } # ifndef NOT_IN_libc +# include + static int internal_function __attribute__ ((pure, unused)) re_string_elem_size_at (const re_string_t *pstr, int idx) @@ -740,7 +742,6 @@ re_string_elem_size_at (const re_string_t *pstr, int idx) # ifdef _LIBC const unsigned char *p, *extra; const int32_t *table, *indirect; -# include uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); if (nrules != 0) @@ -751,7 +752,7 @@ re_string_elem_size_at (const re_string_t *pstr, int idx) indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); p = pstr->mbs + idx; - findidx (&p, pstr->len - idx); + findidx (table, indirect, extra, &p, pstr->len - idx); return p - pstr->mbs - idx; } else diff --git a/posix/regexec.c b/posix/regexec.c index 7032da7..3e52682 100644 --- a/posix/regexec.c +++ b/posix/regexec.c @@ -3749,6 +3749,8 @@ group_nodes_into_DFAstates (const re_dfa_t *dfa, const re_dfastate_t *state, one collating element like '.', '[a-z]', opposite to the other nodes can only accept one byte. */ +#include + static int internal_function check_node_accept_bytes (const re_dfa_t *dfa, int node_idx, @@ -3868,8 +3870,6 @@ check_node_accept_bytes (const re_dfa_t *dfa, int node_idx, const int32_t *table, *indirect; const unsigned char *weights, *extra; const char *collseqwc; - /* This #include defines a local function! */ -# include /* match with collating_symbol? */ if (cset->ncoll_syms) @@ -3925,7 +3925,7 @@ check_node_accept_bytes (const re_dfa_t *dfa, int node_idx, _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); - int32_t idx = findidx (&cp, elem_len); + int32_t idx = findidx (table, indirect, extra, &cp, elem_len); if (idx > 0) for (i = 0; i < cset->nequiv_classes; ++i) { diff --git a/string/strcoll_l.c b/string/strcoll_l.c index 10ce4a6..8764e67 100644 --- a/string/strcoll_l.c +++ b/string/strcoll_l.c @@ -33,6 +33,7 @@ # define STRCMP strcmp # define STRLEN strlen # define WEIGHT_H "../locale/weight.h" +# define FINDIDX findidx # define SUFFIX MB # define L(arg) arg #endif @@ -146,13 +147,14 @@ get_next_seq_cached (coll_seq *seq, int nrules, int pass, seq->idxnow = idxnow; } +#include WEIGHT_H + /* Get next sequence. Traverse the string as required. */ static void get_next_seq (coll_seq *seq, int nrules, const unsigned char *rulesets, const USTRING_TYPE *weights, const int32_t *table, const USTRING_TYPE *extra, const int32_t *indirect) { -#include WEIGHT_H size_t val = seq->val = 0; int len = seq->len; size_t backw_stop = seq->backw_stop; @@ -194,7 +196,7 @@ get_next_seq (coll_seq *seq, int nrules, const unsigned char *rulesets, while (*us != L('\0')) { - int32_t tmp = findidx (&us, -1); + int32_t tmp = FINDIDX (table, indirect, extra, &us, -1); rulearr[idxmax] = tmp >> 24; idxarr[idxmax] = tmp & 0xffffff; idxcnt = idxmax++; @@ -242,7 +244,6 @@ get_next_seq_nocache (coll_seq *seq, int nrules, const unsigned char *rulesets, const USTRING_TYPE *extra, const int32_t *indirect, int pass) { -#include WEIGHT_H size_t val = seq->val = 0; int len = seq->len; size_t backw_stop = seq->backw_stop; @@ -285,7 +286,7 @@ get_next_seq_nocache (coll_seq *seq, int nrules, const unsigned char *rulesets, us = seq->back_us; while (i < backw) { - int32_t tmp = findidx (&us, -1); + int32_t tmp = FINDIDX (table, indirect, extra, &us, -1); idx = tmp & 0xffffff; i++; } @@ -300,7 +301,7 @@ get_next_seq_nocache (coll_seq *seq, int nrules, const unsigned char *rulesets, while (*us != L('\0')) { - int32_t tmp = findidx (&us, -1); + int32_t tmp = FINDIDX (table, indirect, extra, &us, -1); unsigned char rule = tmp >> 24; prev_idx = idx; idx = tmp & 0xffffff; diff --git a/string/strxfrm_l.c b/string/strxfrm_l.c index 04b9338..7d41a28 100644 --- a/string/strxfrm_l.c +++ b/string/strxfrm_l.c @@ -33,6 +33,7 @@ # define STRLEN strlen # define STPNCPY __stpncpy # define WEIGHT_H "../locale/weight.h" +# define FINDIDX findidx # define SUFFIX MB # define L(arg) arg #endif @@ -80,6 +81,7 @@ utf8_encode (char *buf, int val) } #endif +#include WEIGHT_H size_t STRXFRM (STRING_TYPE *dest, const STRING_TYPE *src, size_t n, __locale_t l) @@ -104,8 +106,6 @@ STRXFRM (STRING_TYPE *dest, const STRING_TYPE *src, size_t n, __locale_t l) size_t idxcnt; int use_malloc; -#include WEIGHT_H - if (nrules == 0) { if (n != 0) @@ -174,7 +174,7 @@ STRXFRM (STRING_TYPE *dest, const STRING_TYPE *src, size_t n, __locale_t l) idxmax = 0; do { - int32_t tmp = findidx (&usrc, -1); + int32_t tmp = FINDIDX (table, indirect, extra, &usrc, -1); rulearr[idxmax] = tmp >> 24; idxarr[idxmax] = tmp & 0xffffff; diff --git a/wcsmbs/wcscoll_l.c b/wcsmbs/wcscoll_l.c index 74e2e39..a402806 100644 --- a/wcsmbs/wcscoll_l.c +++ b/wcsmbs/wcscoll_l.c @@ -26,6 +26,7 @@ #define STRCMP wcscmp #define STRLEN __wcslen #define WEIGHT_H "../locale/weightwc.h" +#define FINDIDX findidxwc #define SUFFIX WC #define L(arg) L##arg #define WIDE_CHAR_VERSION 1 diff --git a/wcsmbs/wcsxfrm_l.c b/wcsmbs/wcsxfrm_l.c index f3f3f50..23e1703 100644 --- a/wcsmbs/wcsxfrm_l.c +++ b/wcsmbs/wcsxfrm_l.c @@ -26,6 +26,7 @@ #define STRLEN __wcslen #define STPNCPY __wcpncpy #define WEIGHT_H "../locale/weightwc.h" +#define FINDIDX findidxwc #define SUFFIX WC #define L(arg) L##arg #define WIDE_CHAR_VERSION 1