BUG: TestNG crashing if test method depends on nonexistent group

494 views
Skip to first unread message

Gili T.

unread,
Nov 29, 2012, 2:41:07 PM11/29/12
to testn...@googlegroups.com
Hi,
Can someone please take a look at the issue I reported here? http://netbeans.org/bugzilla/show_bug.cgi?id=219097
I've attached a testcase. Simply unzip and run and you should see: "The forked VM terminated without saying properly goodbye. VM crash or System.exit called ?"

Thank you,
Gili

Cédric Beust ♔

unread,
Nov 29, 2012, 3:55:51 PM11/29/12
to testn...@googlegroups.com
This looks like a JVM bug, not a TestNG one.


-- 
Cédric




--
You received this message because you are subscribed to the Google Groups "testng-dev" group.
To view this discussion on the web visit https://groups.google.com/d/msg/testng-dev/-/paj7rxvY1TAJ.
To post to this group, send email to testn...@googlegroups.com.
To unsubscribe from this group, send email to testng-dev+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/testng-dev?hl=en.

Gili T.

unread,
Dec 7, 2012, 2:30:20 PM12/7/12
to testn...@googlegroups.com, ced...@beust.com
Cédric,

I need your help to establish exactly which component is at fault. I simply don't have the technical background to debug the glue code between the JVM, TestNG and Maven in order to establish who is at fault.

If the JVM is really at fault I'm going to need more information in order to file a bug report. Can you please help?

Thanks,
Gili

Cédric Beust ♔

unread,
Dec 7, 2012, 2:53:34 PM12/7/12
to testn...@googlegroups.com
Not sure I can help, the bottom line is that a JVM crash is a JVM bug, period. Legal Java bytecode should never cause the JVM to crash...

-- 
Cédric




To view this discussion on the web visit https://groups.google.com/d/msg/testng-dev/-/FubE8YVZfp8J.

Gili T.

unread,
Dec 7, 2012, 3:13:27 PM12/7/12
to testn...@googlegroups.com, ced...@beust.com
Cédric,

All I was asking you to do was run the testcase and report what you saw from inside TestNG. We both know that error messages can be misleading and even completely wrong. This seems to be one of these cases. The JVM was not crashing. It turns out to be a Surefire bug: http://jira.codehaus.org/browse/SUREFIRE-931

Thanks anyway,
Gili
Reply all
Reply to author
Forward
0 new messages