On Tue, Jul 30, 2013 at 03:56:02PM -0400, Greg Stein wrote: > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > simply tells the hash function to watch for NUL rather than to > traverse N characters. Because N doesn't have to be tracked, you might > even argue the NUL-test is faster. > > IOW, passing APR_HASH_KEY_STRING does NOT perform a strlen(). So yes: > a lot of this work isn't saving much at all. > > Just look at apr/tables/apr_hash.c:hashfunc_default()
svn_hash__make uses a hash function which does a strlen(). FWIW, I'm happy to do the actual removal / changes of code to implement whatever the decision of this thread is.
