X-Git-Url: https://mattmccutchen.net/rsync/rsync.git/blobdiff_plain/68207537326987e28d42ad3b2fe53bdb1ebcc781..fdd10da6f7995296f6bb5ee3dd7231d6dbc20b5e:/match.c diff --git a/match.c b/match.c index 19b17c5c..72947d0a 100644 --- a/match.c +++ b/match.c @@ -1,21 +1,24 @@ /* - Copyright (C) Andrew Tridgell 1996 - Copyright (C) Paul Mackerras 1996 - - This program is free software; you can redistribute it and/or modify - it under the terms of the GNU General Public License as published by - the Free Software Foundation; either version 2 of the License, or - (at your option) any later version. - - This program is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU General Public License for more details. - - You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. -*/ + * Block matching used by the file-transfer code. + * + * Copyright (C) 1996 Andrew Tridgell + * Copyright (C) 1996 Paul Mackerras + * Copyright (C) 2003, 2004, 2005, 2006 Wayne Davison + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA. + */ #include "rsync.h" @@ -37,30 +40,24 @@ static int total_matches; extern struct stats stats; -static uint32 tablesize; +#define TABLESIZE (1<<16) + static int32 *hash_table; -#define SUM2HASH(sum) ((sum)%tablesize) +#define SUM2HASH2(s1,s2) (((s1) + (s2)) & 0xFFFF) +#define SUM2HASH(sum) SUM2HASH2((sum)&0xFFFF,(sum)>>16) static void build_hash_table(struct sum_struct *s) { int32 i; - uint32 prior_size = tablesize; - - /* Dynamically calculate the hash table size so that the hash load - * for big files is about 80%. This number must be odd or s2 will - * not be able to span the entire set. */ - tablesize = (uint32)(s->count/8) * 10 + 11; - if (tablesize < 65537) - tablesize = 65537; /* a prime number */ - if (tablesize != prior_size) { - free(hash_table); - hash_table = new_array(int32, tablesize); + + if (!hash_table) { + hash_table = new_array(int32, TABLESIZE); if (!hash_table) out_of_memory("build_hash_table"); } - memset(hash_table, 0xFF, tablesize * sizeof hash_table[0]); + memset(hash_table, 0xFF, TABLESIZE * sizeof hash_table[0]); for (i = 0; i < s->count; i++) { uint32 t = SUM2HASH(s->sums[i].sum1); @@ -163,13 +160,16 @@ static void hash_search(int f,struct sum_struct *s, int done_csum2 = 0; int32 i; - if (verbose > 4) - rprintf(FINFO,"offset=%.0f sum=%08x\n",(double)offset,sum); + if (verbose > 4) { + rprintf(FINFO, "offset=%.0f sum=%04x%04x\n", + (double)offset, s2 & 0xFFFF, s1 & 0xFFFF); + } - i = hash_table[SUM2HASH(sum)]; + i = hash_table[SUM2HASH2(s1,s2)]; if (i < 0) goto null_hash; + sum = (s1 & 0xffff) | (s2 << 16); hash_hits++; do { int32 l; @@ -272,7 +272,6 @@ static void hash_search(int f,struct sum_struct *s, s2 += s1; } else --k; - sum = (s1 & 0xffff) | (s2 << 16); /* By matching early we avoid re-reading the data 3 times in the case where a token