]> wimlib.net Git - wimlib/commit
git://wimlib.net / wimlib / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 472366c) | patch
lz_hash_chains.c: Hand-inline do_search()
authorEric Biggers <ebiggers3@gmail.com>
Fri, 25 Jul 2014 04:35:44 +0000 (23:35 -0500)
committerEric Biggers <ebiggers3@gmail.com>
Sat, 26 Jul 2014 16:06:53 +0000 (11:06 -0500)
commite3a8f6a32b258ba8ac6929f6c468b1c873a96fc6
treed32992e23e1765a5b72c11e6c761e831b591c804tree | snapshot
parent472366c8db7cef3eaad9eea3c08048b726db525ecommit | diff
lz_hash_chains.c:  Hand-inline do_search()

On x86_64 this improves performance slightly (XPRESS compression ~3%
faster).  Probably a lot of the difference is due to redundant checks
against 'max_len' getting removed.
src/lz_hash_chains.c diff | blob | history
wimlib repository