 |
lua-l archive |
|
|
||
|
Florian Weimer wrote:
I went ahead and replaced the hash function with Jenkins' lookup3.c. The impact on microbenchmarks was rather mixed. For reading text files line by line, with somewhat regular content, there was a small speed-up (despite line lengths generally above 32 bytes). Fasta was slower, k-nucleotide was faster. Reading large, mostly random strings will be significantly slower, but for quite regular strings, lookup3.c's better mixing seems to pay off.
A nice surprise that you noticed speed up. I have to ask though, from a security point of view - is there any point in replacing one non-cryptographic hash with another? Note Jenkins' stern warning that the algorithm should not be used for cryptographic purposes, and indeed his hash is almost as vulnerable as the Lua one - a quick google search for "jenkins lookup3.c attack" gives this: http://www.team5150.com/~andrew/blog/2007/03/breaking_superfasthash.html, among other things. The author writes that you should randomize a, b and c, preventing his attack (although likely still leaving it open to others - again, it's not a cryptographic hash), which sounds sensible.
- Alex