X-Git-Url: https://mattmccutchen.net/rsync/rsync.git/blobdiff_plain/9486289ce452ab82ca8b3f16fae8c2ef5247a923..e1b3d5c4be46ee0b17a98aa48cb7cb152c4a008b:/match.c

diff --git a/match.c b/match.c
index bc759e24..272025aa 100644
--- a/match.c
+++ b/match.c
@@ -29,18 +29,18 @@ extern int remote_version;
 typedef unsigned short tag;
 
 #define TABLESIZE (1<<16)
-#define NULL_TAG ((tag)-1)
+#define NULL_TAG (-1)
 
 static int false_alarms;
 static int tag_hits;
 static int matches;
-static int data_transfer;
+static int64 data_transfer;
 
 static int total_false_alarms;
 static int total_tag_hits;
 static int total_matches;
-static int64 total_data_transfer;
 
+extern struct stats stats;
 
 struct target {
   tag t;
@@ -49,7 +49,7 @@ struct target {
 
 static struct target *targets;
 
-static tag *tag_table;
+static int *tag_table;
 
 #define gettag2(s1,s2) (((s1) + (s2)) & 0xFFFF)
 #define gettag(sum) gettag2((sum)&0xFFFF,(sum)>>16)
@@ -65,7 +65,7 @@ static void build_hash_table(struct sum_struct *s)
   int i;
 
   if (!tag_table)
-    tag_table = (tag *)malloc(sizeof(tag)*TABLESIZE);
+    tag_table = (int *)malloc(sizeof(tag_table[0])*TABLESIZE);
 
   targets = (struct target *)malloc(sizeof(targets[0])*s->count);
   if (!tag_table || !targets) 
@@ -100,14 +100,13 @@ static void matched(int f,struct sum_struct *s,struct map_struct *buf,
 		rprintf(FINFO,"match at %d last_match=%d j=%d len=%d n=%d\n",
 			(int)offset,(int)last_match,i,(int)s->sums[i].len,(int)n);
 
-	send_token(f,i,buf,last_match,n,i==-1?0:s->sums[i].len);
+	send_token(f,i,buf,last_match,n,i<0?0:s->sums[i].len);
 	data_transfer += n;
 
-	if (n > 0)
-		write_flush(f);
-
-	if (i >= 0)
+	if (i >= 0) {
+		stats.matched_data += s->sums[i].len;
 		n += s->sums[i].len;
+	}
   
 	for (j=0;j<n;j+=CHUNK_SIZE) {
 		int n1 = MIN(CHUNK_SIZE,n-j);
@@ -212,7 +211,17 @@ static void hash_search(int f,struct sum_struct *s,
 		} else {
 			--k;
 		}
-		
+
+		/* By matching early we avoid re-reading the
+		   data 3 times in the case where a token
+		   match comes a long way after last
+		   match. The 3 reads are caused by the
+		   running match, the checksum update and the
+		   literal send. */
+		if (offset-last_match >= CHUNK_SIZE+s->n && 
+		    (end-offset > CHUNK_SIZE)) {
+			matched(f,s,buf,offset - s->n, -2);
+		}
 	} while (++offset < end);
 	
 	matched(f,s,buf,len,-1);
@@ -266,7 +275,7 @@ void match_sums(int f,struct sum_struct *s,struct map_struct *buf,OFF_T len)
 	total_tag_hits += tag_hits;
 	total_false_alarms += false_alarms;
 	total_matches += matches;
-	total_data_transfer += data_transfer;
+	stats.literal_data += data_transfer;
 }
 
 void match_report(void)
@@ -275,7 +284,8 @@ void match_report(void)
 		return;
 
 	rprintf(FINFO,
-		"total: matches=%d  tag_hits=%d  false_alarms=%d  data=%ld\n",
+		"total: matches=%d  tag_hits=%d  false_alarms=%d data=%.0f\n",
 		total_matches,total_tag_hits,
-		total_false_alarms,(long)total_data_transfer);
+		total_false_alarms,
+		(double)stats.literal_data);
 }