Ruby Issue Tracking System

01/27/2026

03:50 AM Ruby Bug #21836: RUBY_MN_THREADS deadlock and sleep issues

To close the loop here, yes, I was wrong: 1sec should be expected, not 10.
Without MN_THREADS, the 0.1msec sleep becomes a 1msec sleep. With MN_THREADS, the 0.1msec sleep is skipped completely. I see the second of theses issues is now... jpl-coconut (Jacob Lacouture)

01/13/2026

09:29 PM Ruby Bug #21836 (Closed): RUBY_MN_THREADS deadlock and sleep issues

I created a benchmark for the purpose of testing [a fix](https://github.com/ruby/ruby/pull/15840) for [Issue 21685](https://bugs.ruby-lang.org/issues/21685)
The benchmark is inline below and saved as 100usec.rb. It should always take ... jpl-coconut (Jacob Lacouture)

01/10/2026

08:21 AM Ruby Bug #21685: Unnecessary context-switching, especially bad on multi-core machines.

I have a much better solution here: https://github.com/ruby/ruby/pull/15840
It's simpler than my original PR. It dispenses with the additional thread and the magic number (sleep time). And in addition to bringing multi-core perf into li... jpl-coconut (Jacob Lacouture)

12/09/2025

07:52 PM Ruby Bug #21685: Unnecessary context-switching, especially bad on multi-core machines.

Interesting. I get similar results on both linux and macos, on both benchmarks. But I'm using ARM in both cases, not x86. jpl-coconut (Jacob Lacouture)

11/22/2025

05:47 AM Ruby Bug #21685: Unnecessary context-switching, especially bad on multi-core machines.

I created a PR with my original changes applied to master. It's [here](https://github.com/ruby/ruby/pull/15294).
The proposal to use a single-thread per process rather than one per ractor involves some complexity, but I think it's sim... jpl-coconut (Jacob Lacouture)

11/19/2025

11:28 PM Ruby Bug #21685: Unnecessary context-switching, especially bad on multi-core machines.

How can I help move the discussion forward? Is it better if I create a PR and we discuss there? jpl-coconut (Jacob Lacouture)

11/13/2025

11:34 PM Ruby Bug #21685 (Open): Unnecessary context-switching, especially bad on multi-core machines.

While debugging a performance issue in a large rails application, I wrote a minimal microbenchmark that reproduces the issue. [[here]](https://gist.github.com/jpl-coconut/cb3679ce885eb578e1071c4b3a525d5c) I was surprised to see that the ... jpl-coconut (Jacob Lacouture)