From: takashikkbn@... Date: 2018-12-23T00:57:33+00:00 Subject: [ruby-core:90676] [Ruby trunk Bug#15430][Assigned] test_fork_while_parent_locked is failing status on Ruby CI Issue #15430 has been updated by k0kubun (Takashi Kokubun). Status changed from Closed to Assigned As of r66497 or r66498, some Linux environments are getting: ~~~ 1) Failure: TestThread#test_fork_while_parent_locked [/home/hsbt/chkbuild/tmp/build/20181222T213003Z/ruby/test/ruby/test_thread.rb:1278]: lower `nr' if SIGKILL because of RLIMIT_NPROC limit. Expected "#" to be empty. ~~~ https://rubyci.org/logs/rubyci.s3.amazonaws.com/debian8/ruby-trunk/log/20181222T213003Z.fail.html.gz https://rubyci.org/logs/rubyci.s3.amazonaws.com/centos7/ruby-trunk/log/20181222T230003Z.fail.html.gz https://rubyci.org/logs/rubyci.s3.amazonaws.com/arch/ruby-trunk/log/20181222T210001Z.fail.html.gz So reopening this. If it's just a bad check, please consider eliminating the check or skipping the test itself to avoid false-positive alarms. > Btw, can we get dmesg or kernel logs from those CI machines to confirm it was OOM killer? I don't have the access to most of RubyCI servers. Please contact @hsbt, @naruse, or possibly @akr about those servers. ---------------------------------------- Bug #15430: test_fork_while_parent_locked is failing status on Ruby CI https://bugs.ruby-lang.org/issues/15430#change-75847 * Author: hsbt (Hiroshi SHIBATA) * Status: Assigned * Priority: Normal * Assignee: normalperson (Eric Wong) * Target version: * ruby -v: * Backport: 2.4: UNKNOWN, 2.5: UNKNOWN ---------------------------------------- After r66230, TestThread#test_fork_while_parent_locked is failing status on Ruby CI. See * CentOS 7 https://rubyci.org/logs/rubyci.s3.amazonaws.com/centos7/ruby-trunk/log/20181217T230002Z.fail.html.gz * Amazon Lunux 2 https://rubyci.org/logs/rubyci.s3.amazonaws.com/amazon2/ruby-trunk/log/20181218T003003Z.fail.html.gz * Arch Linux https://rubyci.org/logs/rubyci.s3.amazonaws.com/arch/ruby-trunk/log/20181218T000002Z.fail.html.gz -- https://bugs.ruby-lang.org/ Unsubscribe: