From: Eric Wong <normalperson@...>
Date: 2018-11-19T00:22:49+00:00
Subject: [ruby-core:89859] Re: [Ruby trunk Bug#15316] rb_postponed_job_register not thread-safe

takashikkbn@gmail.com wrote:
> Well, I seem to have introduced at least two deadlocks by
> that. I already fixed the first one, and I'm going to fix the
> second one
> http://ci.rvm.jp/results/trunk-mjit@silicon-docker/1467162.
> 
> I'll take a deeper look when I come home today, but do you
> know who wakes up rb_sigwait_sleep and/or how the hang there
> could happen?

Any signal should wake it up...

If you can't solve it, can you wait until after the 11/22
developers meeting?

It's a small chance, but I really want to get Thread::Light
(auto-fiber) ready and into 2.6 (at least for Linux/FreeBSD),
and I barely have time for Ruby as it is.

It's not looking like I'll be able to afford to hack on Ruby
next year :<

Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>