*
* Copyright (C) 1996 Andrew Tridgell
* Copyright (C) 1996 Paul Mackerras
- * Copyright (C) 2003-2008 Wayne Davison
+ * Copyright (C) 2003-2020 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
*/
#include "rsync.h"
+#include "inums.h"
extern int checksum_seed;
extern int append_mode;
-extern int checksum_len;
+extern int xfersum_type;
int updating_basis_file;
char sender_file_sum[MAX_DIGEST_LEN];
if (hash_table)
free(hash_table);
hash_table = new_array(int32, tablesize);
- if (!hash_table)
- out_of_memory("build_hash_table");
alloc_size = tablesize;
}
static OFF_T last_match;
-/**
- * Transmit a literal and/or match token.
+/* Transmit a literal and/or match token.
*
* This delightfully-named function is called either when we find a
* match and need to transmit all the unmatched data leading up to it,
* transmit it. As a result of this second case, it is called even if
* we have not matched at all!
*
- * @param i If >0, the number of a matched token. If 0, indicates we
- * have only literal data.
- **/
-static void matched(int f, struct sum_struct *s, struct map_struct *buf,
- OFF_T offset, int32 i)
+ * If i >= 0, the number of a matched token. If < 0, indicates we have
+ * only literal data. A -1 will send a 0-token-int too, and a -2 sends
+ * only literal data, w/o any token-int. */
+static void matched(int f, struct sum_struct *s, struct map_struct *buf, OFF_T offset, int32 i)
{
int32 n = (int32)(offset - last_match); /* max value: block_size (int32) */
int32 j;
if (DEBUG_GTE(DELTASUM, 2) && i >= 0) {
rprintf(FINFO,
"match at %s last_match=%s j=%d len=%ld n=%ld\n",
- big_num(offset, 0), big_num(last_match, 0), i,
+ big_num(offset), big_num(last_match), i,
(long)s->sums[i].len, (long)n);
}
static void hash_search(int f,struct sum_struct *s,
struct map_struct *buf, OFF_T len)
{
- OFF_T offset, end;
- int32 k, want_i, backup;
+ OFF_T offset, aligned_offset, end;
+ int32 k, want_i, aligned_i, backup;
char sum2[SUM_LENGTH];
uint32 s1, s2, sum;
int more;
if (DEBUG_GTE(DELTASUM, 2)) {
rprintf(FINFO, "hash search b=%ld len=%s\n",
- (long)s->blength, big_num(len, 0));
+ (long)s->blength, big_num(len));
}
k = (int32)MIN(len, (OFF_T)s->blength);
if (DEBUG_GTE(DELTASUM, 3))
rprintf(FINFO, "sum=%.8x k=%ld\n", sum, (long)k);
- offset = 0;
+ offset = aligned_offset = aligned_i = 0;
end = len + 1 - s->sums[s->count-1].len;
if (DEBUG_GTE(DELTASUM, 3)) {
rprintf(FINFO, "hash search s->blength=%ld len=%s count=%s\n",
- (long)s->blength, big_num(len, 0), big_num(s->count, 0));
+ (long)s->blength, big_num(len), big_num(s->count));
}
do {
int done_csum2 = 0;
- int32 i;
+ uint32 hash_entry;
+ int32 i, *prev;
if (DEBUG_GTE(DELTASUM, 4)) {
rprintf(FINFO, "offset=%s sum=%04x%04x\n",
- big_num(offset, 0), s2 & 0xFFFF, s1 & 0xFFFF);
+ big_num(offset), s2 & 0xFFFF, s1 & 0xFFFF);
}
if (tablesize == TRADITIONAL_TABLESIZE) {
- if ((i = hash_table[SUM2HASH2(s1,s2)]) < 0)
+ hash_entry = SUM2HASH2(s1,s2);
+ if ((i = hash_table[hash_entry]) < 0)
goto null_hash;
sum = (s1 & 0xffff) | (s2 << 16);
} else {
sum = (s1 & 0xffff) | (s2 << 16);
- if ((i = hash_table[BIG_SUM2HASH(sum)]) < 0)
+ hash_entry = BIG_SUM2HASH(sum);
+ if ((i = hash_table[hash_entry]) < 0)
goto null_hash;
}
+ prev = &hash_table[hash_entry];
hash_hits++;
do {
int32 l;
+ /* When updating in-place, the chunk's offset must be
+ * either >= our offset or identical data at that offset.
+ * Remove any bypassed entries that we can never use. */
+ if (updating_basis_file && s->sums[i].offset < offset
+ && !(s->sums[i].flags & SUMFLG_SAME_OFFSET)) {
+ *prev = s->sums[i].chain;
+ continue;
+ }
+ prev = &s->sums[i].chain;
+
if (sum != s->sums[i].sum1)
continue;
if (l != s->sums[i].len)
continue;
- /* in-place: ensure chunk's offset is either >= our
- * offset or that the data didn't move. */
- if (updating_basis_file && s->sums[i].offset < offset
- && !(s->sums[i].flags & SUMFLG_SAME_OFFSET))
- continue;
-
if (DEBUG_GTE(DELTASUM, 3)) {
rprintf(FINFO,
"potential match at %s i=%ld sum=%08x\n",
- big_num(offset, 0), (long)i, sum);
+ big_num(offset), (long)i, sum);
}
if (!done_csum2) {
/* When updating in-place, the best possible match is
* one with an identical offset, so we prefer that over
- * the following want_i optimization. */
+ * the adjacent want_i optimization. */
if (updating_basis_file) {
- int32 i2;
- for (i2 = i; i2 >= 0; i2 = s->sums[i2].chain) {
- if (s->sums[i2].offset != offset)
- continue;
- if (i2 != i) {
- if (sum != s->sums[i2].sum1)
- break;
- if (memcmp(sum2, s->sums[i2].sum2,
- s->s2length) != 0)
- break;
- i = i2;
+ /* All the generator's chunks start at blength boundaries. */
+ while (aligned_offset < offset) {
+ aligned_offset += s->blength;
+ aligned_i++;
+ }
+ if ((offset == aligned_offset
+ || (sum == 0 && l == s->blength && aligned_offset + l <= len))
+ && aligned_i < s->count) {
+ if (i != aligned_i) {
+ if (sum != s->sums[aligned_i].sum1
+ || l != s->sums[aligned_i].len
+ || memcmp(sum2, s->sums[aligned_i].sum2, s->s2length) != 0)
+ goto check_want_i;
+ i = aligned_i;
+ }
+ if (offset != aligned_offset) {
+ /* We've matched some zeros in a spot that is also zeros
+ * further along in the basis file, if we find zeros ahead
+ * in the sender's file, we'll output enough literal data
+ * to re-align with the basis file, and get back to seeking
+ * instead of writing. */
+ backup = (int32)(aligned_offset - last_match);
+ if (backup < 0)
+ backup = 0;
+ map = (schar *)map_ptr(buf, aligned_offset - backup, l + backup)
+ + backup;
+ sum = get_checksum1((char *)map, l);
+ if (sum != s->sums[i].sum1)
+ goto check_want_i;
+ get_checksum2((char *)map, l, sum2);
+ if (memcmp(sum2, s->sums[i].sum2, s->s2length) != 0)
+ goto check_want_i;
+ /* OK, we have a re-alignment match. Bump the offset
+ * forward to the new match point. */
+ offset = aligned_offset;
}
- /* This chunk was at the same offset on
- * both the sender and the receiver. */
+ /* This identical chunk is in the same spot in the old and new file. */
s->sums[i].flags |= SUMFLG_SAME_OFFSET;
- goto set_want_i;
+ want_i = i;
}
}
+ check_want_i:
/* we've found a match, but now check to see
* if want_i can hint at a better match. */
if (i != want_i && want_i < s->count
- && (!updating_basis_file || s->sums[want_i].offset >= offset
- || s->sums[want_i].flags & SUMFLG_SAME_OFFSET)
- && sum == s->sums[want_i].sum1
- && memcmp(sum2, s->sums[want_i].sum2, s->s2length) == 0) {
+ && (!updating_basis_file || s->sums[want_i].offset >= offset
+ || s->sums[want_i].flags & SUMFLG_SAME_OFFSET)
+ && sum == s->sums[want_i].sum1
+ && memcmp(sum2, s->sums[want_i].sum2, s->s2length) == 0) {
/* we've found an adjacent match - the RLL coder
* will be happy */
i = want_i;
}
- set_want_i:
want_i = i + 1;
matched(f,s,buf,offset,i);
/* Trim off the first byte from the checksum */
more = offset + k < len;
- map = (schar *)map_ptr(buf, offset - backup, k + more + backup)
- + backup;
+ map = (schar *)map_ptr(buf, offset - backup, k + more + backup) + backup;
s1 -= map[0] + CHAR_OFFSET;
s2 -= k * (map[0]+CHAR_OFFSET);
**/
void match_sums(int f, struct sum_struct *s, struct map_struct *buf, OFF_T len)
{
+ int sum_len;
+
last_match = 0;
false_alarms = 0;
hash_hits = 0;
matches = 0;
data_transfer = 0;
- sum_init(checksum_seed);
+ sum_init(xfersum_type, checksum_seed);
if (append_mode > 0) {
if (append_mode == 2) {
matched(f, s, buf, len, -1);
}
- if (sum_end(sender_file_sum) != checksum_len)
- overflow_exit("checksum_len"); /* Impossible... */
+ sum_len = sum_end(sender_file_sum);
/* If we had a read error, send a bad checksum. We use all bits
* off as long as the checksum doesn't happen to be that, in
* which case we turn the last 0 bit into a 1. */
if (buf && buf->status != 0) {
int i;
- for (i = 0; i < checksum_len && sender_file_sum[i] == 0; i++) {}
- memset(sender_file_sum, 0, checksum_len);
- if (i == checksum_len)
+ for (i = 0; i < sum_len && sender_file_sum[i] == 0; i++) {}
+ memset(sender_file_sum, 0, sum_len);
+ if (i == sum_len)
sender_file_sum[i-1]++;
}
if (DEBUG_GTE(DELTASUM, 2))
rprintf(FINFO,"sending file_sum\n");
- write_buf(f, sender_file_sum, checksum_len);
+ write_buf(f, sender_file_sum, sum_len);
if (DEBUG_GTE(DELTASUM, 2)) {
rprintf(FINFO, "false_alarms=%d hash_hits=%d matches=%d\n",
rprintf(FINFO,
"total: matches=%d hash_hits=%d false_alarms=%d data=%s\n",
total_matches, total_hash_hits, total_false_alarms,
- big_num(stats.literal_data, 0));
+ big_num(stats.literal_data));
}
git.samba.org / rsync.git / blobdiff