Bug #5619: Segfault on OS X

• Status: Closed
• Assignee: mrkn (Kenta Murata)
• Target version: 2.0.0
• ruby -v: ruby 2.0.0dev (2012-03-10 trunk 34964) [x86_64-darwin11.3.0]

[mutru (Otto Hilska)]

When I run 'roadie' gem specs on OS X, Ruby 1.9.3p0 segfaults.

Here's how to reproduce:

$ rvm use 1.9.3
$ git clone git://github.com/Mange/roadie.git
$ cd roadie
$ bundle install
$ bundle exec rake

I've attached the command output and the crash report.

It seems it might be a nokogiri issue, but since nokogiri's own tests run fine, I'm not sure.

This is my environment:

$ uname -a
Darwin becherovkax.local 11.2.0 Darwin Kernel Version 11.2.0: Tue Aug 9 20:54:00 PDT 2011; root:xnu-1699.24.8~1/RELEASE_X86_64 x86_64
$ ruby -v
ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-darwin11.2.0]

It also happens with current 1.9.3-head (installed from RVM):
ruby 1.9.3p0 (2011-11-08 revision 33661) [x86_64-darwin11.2.0]

However, everything works fine on Linux.

[kosaki (Motohiro KOSAKI)]

Seems memory curruption issue.
Can you try valgrind or other memory error detection tools?

[nobu (Nobuyoshi Nakada)]

Seems libxml related finalizer issue.

[mutru (Otto Hilska)]

• File valgrind-log.txt valgrind-log.txt added

Here's what I get running valgrind. Let me know if I should enable some extra checks.

[kosaki (Motohiro KOSAKI)]

• ruby -v changed from ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-darwin11.2.0] to -

File valgrind-log.txt added

Here's what I get running valgrind. Let me know if I should enable some extra checks.

Hmmm.. This seems to don't have a useful hint.

[valo (Valentin Mihov)]

• File ruby_2011-11-14-154022_valo-mac.crash ruby_2011-11-14-154022_valo-mac.crash added
• File crash_output.txt crash_output.txt added

I verified the segmentation fault on the same version of OS X with ruby 1.9.3. I am attaching the crash report from OS X and the output on the command line. I will try later to run it through valgrind and see if I can find anything there...

[mrkn (Kenta Murata)]

• Assignee set to mrkn (Kenta Murata)

[mrkn (Kenta Murata)]

• Target version changed from 1.9.3 to 2.0.0
• ruby -v changed from - to ruby 2.0.0dev (2012-03-10 trunk 34964) [x86_64-darwin11.3.0]

[shyouhei (Shyouhei Urabe)]

• Status changed from Open to Assigned

[mrkn (Kenta Murata)]

• Status changed from Assigned to Closed

I couldn't reproduce this issue on the latest trunk, r38731.
It may be fixed on the current trunk, I think.