[issue34751] Hash collisions for tuples

[Jeroen Demeyer]

Jeroen Demeyer <J.Demeyer at UGent.be> added the comment:

> So, jdemeyer, if it's possible to show (or describe) to us an example of a problem you had, such that we could repeat it, that would be helpful (and necessary) to evaluate any proposed changes.  What were the inputs to hash() that caused a problem, and how did that problem manifest itself?

In all honesty, I don't remember. This was a while ago and at that time I didn't care enough to put up a CPython bug report. Still, this is a collision for tuples of short length (3) containing small integers (0 and -2). Why do you find that contrived?

What prompted me to report this bug now anyway is that I discovered bad hashing practices for other classes too. For example, meth_hash in Objects/methodobject.c simply XORs two hashes and XOR tends to suffer from catastrophic cancellation. So I was hoping to fix tuple hashing as an example for other hash functions to follow.

----------

_______________________________________
Python tracker <report at bugs.python.org>
<https://bugs.python.org/issue34751>
_______________________________________