Possible performance improvement by replacing % (modulo) with multiply + shift

[thomasmueller]

This is described in http://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/

(I'm looking for a fast Bloom filter implementation, and found your implementation uses modulo which is a bit slow.)

[mavam]

Thanks for taking a look at the performance.

Do you have some example/profiling data showing that this is indeed on the critical path?

(Unfortunately I don't have the cycles to drive such experiments forward at the moment, but I'm happy to review and integrate community contributions.)

[thomasmueller]

Feel free to close the issue. I'm working on my own Bloom filter implementation, and for my case it does make a big difference. But if you don't have any benchmarks, maybe you don't care so much about performance... that's fine! It was just FYI.

[mavam]

[..] and for my case it does make a big difference.

Good to know, I appreciate that you shared this information.

But if you don't have any benchmarks, maybe you don't care so much about performance... that's fine!

I do care about performance actually, I just don't have cycles at the moment to advance this repository into a state where I would like to see it. 😕