yep as you figured specs eventually end up here as they get merged, just check history :
fb1001f + a9b4d28
... now CI is unfortunately, once again, in a very 🔴 shape so maybe its best to tag new failures at this point (so that its less confusing for the person doing the cleanup to worry about new failures as well) ?

Sorry about some confusion on master about failing tests. I made a change for 9.1 that caused a number of suites to start terminating early with a successful return code, so it looked like things were suddenly green.

We will follow normal process for updating specs, so we get the appropriate logs. @perlun If you like you could rebase off current master and force push, so we'd get an accounting of new failures.

@headius Thank you for that. Rebased & force-pushed, we should be start seeing the failures in a short while.

Interestingly enough, I can't see the expected test failures now (but it fails on a random bunch of other tests. 😉) I'll try to update my jruby-head in the near future to re-test this locally with the latest master code.

To summarize, for specs fixing bugs in JRuby, the recommended process is adding the spec directly in JRuby's spec/ruby, which is synchronized ~once a month with ruby/spec and other copies of the specs.
This has the advantage that the fix in JRuby can be immediately tested with the new added spec :)

If there is any question regarding new specs, feel free to ping me.

@enebo @headius Maybe I should add a spec/README similar to what's in MRI?
https://github.com/ruby/ruby/tree/trunk/spec

To summarize, for specs fixing bugs in JRuby

Alright. In this case, some of it was fixed in JRuby but some of it is still unfixed. I was under the impression that the CI runs of this PR should highlight the latter, but I couldn't find it in the output (there are unfortunately a few other CI errors at the moment.)

I tried rebuilding my local JRuby and can verify that the specs still fail with the latest master version (well, 1cc0674 which is probably not latest anymore but still)

$ ../mspec/bin/mspec spec/ruby/core/process/status/
$ /Volumes/git/3rd-party/jruby/bin/jruby /Volumes/git/3rd-party/mspec/bin/mspec-run spec/ruby/core/process/status/
jruby 9.2.0.0-SNAPSHOT (2.5.0) 2018-04-19 6bdfd05 Java HotSpot(TM) 64-Bit Server VM 10+46 on 10+46 +jit [darwin-x86_64]

1)
Process::Status#exitstatus for a child that raised SignalException returns nil FAILED
Expected 143
 to equal nil

/Volumes/git/3rd-party/jruby/spec/ruby/core/process/status/exitstatus_spec.rb:21:in `block in (root)'
org/jruby/RubyBasicObject.java:2596:in `instance_eval'
org/jruby/RubyEnumerable.java:1741:in `all?'
org/jruby/RubyArray.java:1791:in `each'
org/jruby/RubyArray.java:1791:in `each'
/Volumes/git/3rd-party/jruby/spec/ruby/core/process/status/exitstatus_spec.rb:3:in `<main>'
org/jruby/RubyKernel.java:990:in `load'
org/jruby/RubyBasicObject.java:2596:in `instance_eval'
org/jruby/RubyArray.java:1791:in `each'

2)
Process::Status#termsig for a child that raised SignalException returns the signal FAILED
Expected nil
 to equal 15

/Volumes/git/3rd-party/jruby/spec/ruby/core/process/status/termsig_spec.rb:23:in `block in (root)'
org/jruby/RubyBasicObject.java:2596:in `instance_eval'
org/jruby/RubyEnumerable.java:1741:in `all?'
org/jruby/RubyArray.java:1791:in `each'
org/jruby/RubyArray.java:1791:in `each'
/Volumes/git/3rd-party/jruby/spec/ruby/core/process/status/termsig_spec.rb:3:in `<main>'
org/jruby/RubyKernel.java:990:in `load'
org/jruby/RubyBasicObject.java:2596:in `instance_eval'
org/jruby/RubyArray.java:1791:in `each'
[- | ==================100%================== | 00:00:00]      2F      0E

Finished in 29.725798 seconds

16 files, 13 examples, 13 expectations, 2 failures, 0 errors, 0 tagged

@kares @headius Are you fine with merging this for now? The change is already in ruby/spec as mentioned, so it will come our way sooner or later anyway.

Ping - are there any major problems with merging this?

Looks fine for me, since this is the exact same diff as in ruby/spec it should be easy to ignore it when doing the ruby/spec update (in a few days, I do it at the end of the month).

I'll let JRuby devs merge this if they want though.

I updated all specs in #5154.

I recommend to add specs directly in JRuby when working on a bug, so then the specs can be immediately run in JRuby's CI. The specs are anyway integrated in ruby/spec ~once a month, so other implementations can also benefit from them.