Bug #18393: TestReadline#test_interrupt_in_other_thread fails on armv7hl

• Status: Open
• Priority: Normal
• Assignee: -
• Target version: -
• ruby -v: ruby 3.1.0dev (2021-12-07 master ec878dac90) [armv7hl-linux]
• Backport: 2.6: UNKNOWN, 2.7: UNKNOWN, 3.0: UNKNOWN

[vo.x (Vit Ondruch)]

Testing Ruby 3.1 on Fedora, I observe following issue:

  1) Failure:
TestReadline#test_interrupt_in_other_thread [/builddir/build/BUILD/ruby-3.1.0-ec878dac90/test/readline/test_readline.rb:574]:
Unknown failure with exit status pid 12195 exit 1
Log: ** START **Readline::VERSION is 8.1.
#<Thread:0xb4619004 /tmp/interrupt_in_other_thread20211207-8101-5l3i8y:9 run> terminated with exception (report_on_exception is true):
/tmp/interrupt_in_other_thread20211207-8101-5l3i8y:13:in `readline': stack level too deep (SystemStackError)
	from /tmp/interrupt_in_other_thread20211207-8101-5l3i8y:13:in `block in <main>'
/tmp/interrupt_in_other_thread20211207-8101-5l3i8y:13:in `readline': stack level too deep (SystemStackError)
	from /tmp/interrupt_in_other_thread20211207-8101-5l3i8y:13:in `block in <main>'

It seems this is know issue. But I wonder:

1. Can it be disabled similarly as the i686 so one does not need to trying to figure this out.
2. If this is possibly readline issue, is there some upstream ticket somewhere? I see just reference to CI etc.