greg/pcompress
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Optimize index lookup for 8-byte keys.

Browse source 
More cleanups.
This commit is contained in:
Moinak Ghosh 2013-04-24 19:49:43 +05:30
parent 5d6ffd969d
commit 6c5d8d9e18
2 changed files with 14 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
rabin/global/index.c
View file

@ -168,7 +168,7 @@ set_cfg:
} else { } else {
cfg->intervals = 100 / *pct_interval; cfg->intervals = 100 / *pct_interval;
cfg->sub_intervals = cfg->intervals; cfg->sub_intervals = cfg->intervals;
*hash_slots = file_sz / cfg->segment_sz_bytes + 1; *hash_slots = file_sz / cfg->segment_sz_bytes;
*hash_slots *= cfg->sub_intervals; *hash_slots *= cfg->sub_intervals;
} }
@ -320,7 +320,7 @@ int
db_segcache_map(archive_config_t *cfg, int tid, uint32_t *blknum, uint64_t *offset, uchar_t **blocks) db_segcache_map(archive_config_t *cfg, int tid, uint32_t *blknum, uint64_t *offset, uchar_t **blocks)
{ {
uchar_t *mapbuf, *hdr; uchar_t *mapbuf, *hdr;
int fd; int fd, dummy;
uint32_t len, adj; uint32_t len, adj;
uint64_t pos; uint64_t pos;
@ -364,6 +364,7 @@ db_segcache_map(archive_config_t *cfg, int tid, uint32_t *blknum, uint64_t *offs
*blknum = *((uint32_t *)(hdr)); *blknum = *((uint32_t *)(hdr));
*offset = *((uint64_t *)(hdr + 4)); *offset = *((uint64_t *)(hdr + 4));
*blocks = hdr + SEGCACHE_HDR_SZ; *blocks = hdr + SEGCACHE_HDR_SZ;
dummy = *(hdr + SEGCACHE_HDR_SZ);
cfg->seg_fd_r[tid].mapping = mapbuf; cfg->seg_fd_r[tid].mapping = mapbuf;
cfg->seg_fd_r[tid].len = len + adj; cfg->seg_fd_r[tid].len = len + adj;
@ -436,6 +437,15 @@ db_lookup_insert_s(archive_config_t *cfg, uchar_t *sim_cksum, int interval,
pent = &(ent->next); pent = &(ent->next);
ent = ent->next; ent = ent->next;
} }
} else if (cfg->similarity_cksum_sz == 8) {
while (ent) {
if (*((uint64_t *)sim_cksum) == *((uint64_t *)ent->cksum) &&
ent->item_offset != item_offset) {
return (ent);
}
pent = &(ent->next);
ent = ent->next;
}
} else { } else {
while (ent) { while (ent) {
if (mycmp(sim_cksum, ent->cksum, cfg->similarity_cksum_sz) == 0 && if (mycmp(sim_cksum, ent->cksum, cfg->similarity_cksum_sz) == 0 &&

6
rabin/rabin_dedup.c
View file

@ -866,16 +866,14 @@ process_blocks:
*/ */
sim_ck = ctx->similarity_cksums; sim_ck = ctx->similarity_cksums;
sub_i = cfg->sub_intervals; sub_i = cfg->sub_intervals;
len = length;
tgt = seg_heap; tgt = seg_heap;
increment = cfg->chunk_cksum_sz; increment = cfg->chunk_cksum_sz;
if (increment * sub_i > len) if (increment * sub_i > length)
sub_i = len / increment; sub_i = length / increment;
for (j = 0; j<sub_i; j++) { for (j = 0; j<sub_i; j++) {
crc = lzma_crc64(tgt, increment/4, 0); crc = lzma_crc64(tgt, increment/4, 0);
*((uint64_t *)sim_ck) = crc; *((uint64_t *)sim_ck) = crc;
tgt += increment; tgt += increment;
len -= increment;
sim_ck += cfg->similarity_cksum_sz; sim_ck += cfg->similarity_cksum_sz;
} }