Narihiro Nakamura <authornari / gmail.com> wrote:
> I think it's lazy-sweep bug, so I'm going to fix it.

Hi, I've been running the nokogiri test case for this bug on Ruby trunk
with r34719 (and r34720) for 4 hours now and have yet to see any
segfaults.

Hopefully we can have a 1.9.3 release with a version of this fix
backported soon.  I suspect this bug may has caused segfaults in other
gems/modules/cases where users did/could not diagnose the problem, too.

Thank you.