Hi, See below a message from Motohiro KOSAKI, a Ruby developer (Cced). Full discussion at http://redmine.ruby-lang.org/issues/5239
Short story: Some breaks in the Ruby test suite on kfreebsd. The code does: - create a process with fork() - create a (p)thread - issue a waitpid() from the pthread on the process so, the waitpid() is issued from the parent process, but not the parent thread. Linux accepts that, kfreebsd fails, Motohiro says that FreeBSD works. I don't understand the differences between FreeBSD and kfreebsd regarding pthreads. Is it documented somewhere? Can you fix this? It's not blocking for me: Motohiro accepted to apply a fix. But this behaviour might cause other bugs. - Lucas ----- Forwarded message from KOSAKI Motohiro <[email protected]> ----- From: KOSAKI Motohiro <[email protected]> To: [email protected] Date: Sun, 28 Aug 2011 09:42:01 +0900 Subject: [ruby-core:39153] Re: [Ruby 1.9 - Bug #5239][Assigned] bootstraptest/runner.rb: assert_normal_exit logic broken on Debian/GNU kFreeBSD Reply-To: [email protected] List-Id: ruby-core.ruby-lang.org >> > RATIONALE >> > A call to the wait() or waitpid() function only returns status on an >> > immediate child process of the *calling process* >> >> The pthread rule is, pthread_create() doen't introduce any process tree >> change. In the other words, kFreeBSD is buggy. > > One could argue that a different PID makes a different process, but it's > correct that it's not a different process. popen makes different process. and Thread.new don't make a process. isn't it? Therefore a subthread clealy can use waitpid() for waiting child process. Moreover real FreeBSD works this test completely. It's only kFreeBSD issue. I wonder why kFreeBSD don't handle threading correctly. Long time ago, linuxthreads made the same mistake and application users sufferd from it. I didn't expected we observed the same issue at 21th century. ----- End forwarded message ----- -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/[email protected]
