Add mimalloc to the malloc() providers.

[lzlrd]

Is your feature request related to a problem? Please describe.

Microsoft's mimalloc is a fast memory allocator and, according to their benchmarks, is faster than the competition in nearly every case:

Describe the solution you'd like

I would like to be able to use mimalloc as an option when building as specified by make MALLOC=mimalloc.

Describe alternatives you've considered

n/a

Additional context

https://github.com/microsoft/mimalloc

[lzlrd]

I'd like to specifically take a look at the Redis test here, it shows both mimalloc and TCMalloc (#177) outshine the competition including jemalloc.

[victorstewart]

+1

[lzlrd]

mimalloc is great, but TCMalloc outperforms it for Redis (which is our case here), so I'm going to close this issue and direct y'all to #122 where @JohnSully can hopefully update us if TCMalloc has been fixed upstream.

[lzlrd]

@JohnSully, Mimalloc's benchmarks now show it outperforms TCMalloc on Redis so I think it's a good time to re-open this issue. I've been using it for quite a few projects myself, and it's catching traction amongst others (mold, Android, etc.). I think it would be helpful if we could at least benchmark Mimalloc to see if it is helpful.

[Dan-Do]

@JohnSully, Mimalloc's benchmarks now show it outperforms TCMalloc on Redis so I think it's a good time to re-open this issue. I've been using it for quite a few projects myself, and it's catching traction amongst others (mold, Android, etc.). I think it would be helpful if we could at least benchmark Mimalloc to see if it is helpful.

I also have some personal projects tested on mimalloc and it's great. Do you have successfully modify keydb source code so that I can test it?

[osevan]

And now where are developer Benchmarks for interested in peoples?