Issue #13338 has been updated by MSP-Greg (Greg L).


I recently changed my build system from a dual core notebook to a quad core desktop.  On the notebook, I never recall a SEGV that occurred in test-all that did not occur in the test pulled from the framework.  On my desktop, I changed the j parameter in the make test-all call to `j2` from the notebook's `j1`.  I'm now seeing tests that fail or SEGV in test-all, but not when run separately.

Hence, I have more testing to do.  Sorry.

I should have a better understanding in a few hours; I'll post results.

----------------------------------------
Bug #13338: MinGW SEGV in test/ruby/test_keyword.rb svn 58034, ok in svn 58021.
https://bugs.ruby-lang.org/issues/13338#change-63696

* Author: MSP-Greg (Greg L)
* Status: Open
* Priority: Normal
* Assignee: 
* Target version: 
* ruby -v: ruby 2.5.0dev (2017-03-20 trunk 58034) [x64-mingw32]
* Backport: 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: UNKNOWN
----------------------------------------
In a MinGW / MSYS2 64 bit build, the following line generates a SEGV in svn 58034 [test/ruby/test_keyword.rb:L373](https://github.com/ruby/ruby/blob/57d6dd5a461202f9824c3d1b8d3e5304aca2d21c/test/ruby/test_keyword.rb#L373).  With 373 commented out, all tests in the file pass.

This does not occur in svn 58021 for either 32 or 64.  Haven't built 32 bit in 58034 yet.

No SEGV info is generated, either in test-all or running the test after build...



-- 
https://bugs.ruby-lang.org/

Unsubscribe: <mailto:ruby-core-request / ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>