[bloom] Benchmarks with Knuth multiplier-based hash production

From: Joaquin M López Muñoz (joaquinlopezmunoz_at_[hidden])
Date: 2025-06-06 19:02:14

Hi,

As per feedback given here:

https://lists.boost.org/Archives/boost//2025/05/259548.php

I've experimented with the  hash production scheme suggested by Kostas,
where h_{i+1} is calculated as h_i * 6364136223846793005 mod 2^64:

https://github.com/joaquintides/bloom/blob/refs/heads/feature/alternative-hash-production/include/boost/bloom/detail/core.hpp#L105-L125

These benchmarks show the performance of each filter configuration with
the original mechanism and, on the next row, the new mechanism (filter
names marked with *):

https://github.com/joaquintides/boost_bloom_benchmarks/blob/alternative-hash-production/README.md

I can't draw any definitive conclusion as to whether the new mechanism
affects performance, probably because the GHA-run benchmarks have a
good deal of noise and variations in CPU speed. Maybe a keener eye than
mine can tell, or if you're interested in trying the benchmark locally 
please
tell me so and I can instruct you on how to do it.

Feedback welcome,

Joaquin M Lopez Munoz