iProvide better comment for chunking decisions - dedup - deduplicating backup program	Err	bitreich.org	70
hgit clone git://bitreich.org/dedup/ git://enlrupgkhuxnvlhsf6lc3fziv5h2hhfrinws65d7roiv6bfj7d652fid.onion/dedup/	URL:git://bitreich.org/dedup/ git://enlrupgkhuxnvlhsf6lc3fziv5h2hhfrinws65d7roiv6bfj7d652fid.onion/dedup/	bitreich.org	70
1Log	/scm/dedup/log.gph	bitreich.org	70
1Files	/scm/dedup/files.gph	bitreich.org	70
1Refs	/scm/dedup/refs.gph	bitreich.org	70
1Tags	/scm/dedup/tag	bitreich.org	70
1README	/scm/dedup/file/README.gph	bitreich.org	70
1LICENSE	/scm/dedup/file/LICENSE.gph	bitreich.org	70
i---	Err	bitreich.org	70
1commit 39f907c919176a8fed990e10e95a7f7371aedf40	/scm/dedup/commit/39f907c919176a8fed990e10e95a7f7371aedf40.gph	bitreich.org	70
1parent 90f2ff9fcf896060da4be75c70cd148f18be28e9	/scm/dedup/commit/90f2ff9fcf896060da4be75c70cd148f18be28e9.gph	bitreich.org	70
hAuthor: z3bra <contactatz3bradotorg>	URL:mailto:contactatz3bradotorg	bitreich.org	70
iDate:   Sun, 17 Feb 2019 13:04:03 +0100	Err	bitreich.org	70
i	Err	bitreich.org	70
iProvide better comment for chunking decisions	Err	bitreich.org	70
i	Err	bitreich.org	70
iDiffstat:	Err	bitreich.org	70
i  M dedup.c                             |      21 ++++++++++++++-------	Err	bitreich.org	70
i	Err	bitreich.org	70
i1 file changed, 14 insertions(+), 7 deletions(-)	Err	bitreich.org	70
i---	Err	bitreich.org	70
1diff --git a/dedup.c b/dedup.c	/scm/dedup/file/dedup.c.gph	bitreich.org	70
i@@ -74,6 +74,11 @@ char *argv0;	Err	bitreich.org	70
i /*	Err	bitreich.org	70
i  * Static table for use in buzhash algorithm.	Err	bitreich.org	70
i  * 256 * 32 bits randomly generated unique integers	Err	bitreich.org	70
i+ *	Err	bitreich.org	70
i+ * To get better pseudo-random results, there is exactly the same number	Err	bitreich.org	70
i+ * of 0 and 1 spread amongst these integers. It means that there is	Err	bitreich.org	70
i+ * exactly 50% chance that a XOR operation would flip all the bits in	Err	bitreich.org	70
i+ * the hash.	Err	bitreich.org	70
i  */	Err	bitreich.org	70
i uint32_t buz[] = {	Err	bitreich.org	70
i         0xbc9fa594,0x30a8f827,0xced627a7,0xdb46a745,0xcfa4a9e8,0x77cccb59,0xddb66276,0x3adc532f,	Err	bitreich.org	70
i@@ -115,9 +120,9 @@ uint32_t	Err	bitreich.org	70
i buzh_init(uint8_t *buf, size_t size)	Err	bitreich.org	70
i {	Err	bitreich.org	70
i         size_t i;	Err	bitreich.org	70
i-        uint32_t fp = 0;	Err	bitreich.org	70
i+        uint32_t fp;	Err	bitreich.org	70
i 	Err	bitreich.org	70
i-        for (i = size - 1; i > 0; i--, buf++)	Err	bitreich.org	70
i+        for (i = size - 1, fp = 0; i > 0; i--, buf++)	Err	bitreich.org	70
i                 fp ^= ROTL(buz[*buf], i % 32);	Err	bitreich.org	70
i 	Err	bitreich.org	70
i         return fp ^ buz[*buf];	Err	bitreich.org	70
i@@ -136,11 +141,13 @@ chunk_blk(uint8_t *buf, size_t size)	Err	bitreich.org	70
i         uint32_t fp;	Err	bitreich.org	70
i 	Err	bitreich.org	70
i         /*	Err	bitreich.org	70
i-         * Chunking blocks is decided using a rolling hash + binary pattern.	Err	bitreich.org	70
i-         * The buzhash algorithm is used to "fingerprint" a fixed size window.	Err	bitreich.org	70
i-         * Once the lower bits of this fingerprint are all zeros,	Err	bitreich.org	70
i-         * the block is chunked.	Err	bitreich.org	70
i-         * If the pattern can't be matched, then we return the buffer size.	Err	bitreich.org	70
i+         * To achieve better deduplication, we chunk blocks based on a	Err	bitreich.org	70
i+         * recurring pattern occuring on the data stream. A fixed window	Err	bitreich.org	70
i+         * of WINSIZ bytes is slid over the data, and a rolling hash is	Err	bitreich.org	70
i+         * computed for this window.	Err	bitreich.org	70
i+         * When the rolling hash matches a given pattern (see HASHMSK),	Err	bitreich.org	70
i+         * the block is chunked at the end of that window, thus making	Err	bitreich.org	70
i+         * WINSIZ the smallest possible block size.	Err	bitreich.org	70
i          */	Err	bitreich.org	70
i         fp = buzh_init(buf, WINSIZ);	Err	bitreich.org	70
i         for (i = 1; i < size - WINSIZ; i++) {	Err	bitreich.org	70
.