[issue19183] PEP 456 Secure and interchangeable hash algorithm

[Antoine Pitrou]

Antoine Pitrou added the comment:

So, amusingly, Christian's patch seems to be 4-5% faster than vanilla on many benchmarks here (Sandy Bridge Core i5, 64-bit, gcc 4.8.1). A couple of benchmarks are a couple % slower, but nothing severe.

This without the small strings optimization.

On top of that, the small strings optimization seems to have varying effects, some positive, some negative, so it doesn't seem to be desirable (on this platform anyway).

----------

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue19183>
_______________________________________