rchouinard/phpass

Remote Timing Attack vulnerability

Closed this issue · 2 comments

I've looked into this previously, and all my research indicates that this is unnecessary when dealing with proper password hashing techniques. Even the article you link states, "Using one-way hashing should defeat this because it’s very hard to work backwards from timing leaks when the attacker derived input keeps changing (hashes to a different digest each time a byte changes in the original plaintext input)."

I'll be happy to look into this again, but these issues are related more to accidental data disclosure than brute-force password cracking when hashes are in use.

OK, no problem. Keep up the good work.