3 ms·
It does nothing to protect against having to do bit-to-bit comparison of keys when two keys hash to bit-exact same hash value. This is what's behind "hash flood
by exDM69 10y ago
It does nothing to protect against having to do bit-to-bit comparison of keys when two keys hash to bit-exact same hash value. This is what's behind "hash flooding" algorithmic complexity DDoS attacks that cause O(n^2) CPU usage w.r.t. input length.
SipHash is the only hash function that has credible cryptanalysis against key recovery and collisions.
E.g. FNV1 is trivially simple to generate collisions against.
- rurban 10y agoIt's also trivial to produce SipHash collisions and its "security" claims are dubious in the hash table context. There's not much credibility in the siphash paper in section 7. linear chaining is not the recommended way to treat hash tables anymore, and even with that you can add a useful protection. siphash is not one. Yes, it's non-reversable, and it doesn't allow easy seed-independent multi collisions as with other hash functions. But one hash function used in a hash table can never be secure. The collision resolution adds the security part. (or/plus some other schemes which are obviously over your head).