Takashi Kokubun

Three worthwhile reads for anyone interested in the new (M)JIT: https://medium.com/@k0kubun/the-method-jit-compiler-for-ruby-2-6-388ee… https://medium.com/square-corner-blog/rubys-new-jit-91a5c864dd10 https://www.johnhawthorn.com/2018/02/playing-with-ruby-jit-mjit/

Is that something different to what vmakarov is doing [1], the work of current MJIT from k0kubun [2] seems to be based on this earlier work. And I believe vmakarov is working on RTL from YARV instructions, and a light weight JIT [3]. If I remember correctly there was a Tracing JIT for CRuby in 2016, but discontinued due to memory consumption [1] https://bugs.ruby-lang.org/issues/12589 [2] https://medium.com/@k0kubun/the-method-jit-compiler-for-ruby-2-6-388ee… [3] https://www.youtube.com/watch?v=emhYoI_RiOA

Not sure why this comment is downvoted, this seems like a legitimate question as the JIT has some serious performance benefits (https://medium.com/@k0kubun/ruby-2-6-jit-progress-and-future-84e0a830e…

> Are there any benchmarks for the new JIT? From the horse's mouth: "...apparently JIT on Ruby 2.6 failed to be ready for production." [1] [1] https://medium.com/@k0kubun/ruby-2-6-jit-progress-and-future-84e0a830e…

Benchmarks here: https://medium.com/@k0kubun/ruby-2-6-jit-progress-and-future-84e0a830e… tl;dr: your Rails app is not faster with this feature in this version, but might be in the future.

>The initial results of MJIT for simply removing the instruction dispatch overhead and doing some basic optimizations are a 30-230% performance increase on a small but real-world benchmark. No type specialization and specular optimization required So, this amounts to a small improvement for some types of code. Indeed, it is "easy" to get that by "just" using some basic JIT techniques. The trick is to get consistently better performance across the board. Relevant tweet at https://medium.com/@k0kubun/the-method-jit-compiler-for-ruby-2-6-388ee… >I've just committed the initial JIT compiler for Ruby. It's not still so fast yet (especially it's performing badly with Rails for now), but we have much time to improve it until Ruby 2.6 (or 3.0) release.

Don't get too excited. This is a micro-benchmark and not representative of what most real world apps will see. A benchmark of something the scope of a typical rails app is likely to be much much less. Earlier descriptions of MJIT mention benchmark tests that were only ~11% improvement until the benchmark was "tuned" for the JIT. [1] If you're modifying the the benchmark specifically for the JIT, it's not really a valid benchmark IMHO. [1] https://medium.com/@k0kubun/the-method-jit-compiler-for-ruby-2-6-388ee…