[8u20] Request for approval: 6545422: TEST BUG: NativeErrors.java uses wrong path name in exec (original) (raw)

Ivan Gerasimov ivan.gerasimov at oracle.com
Tue Jul 1 21:29:50 UTC 2014

Hello!

I'm rerequesting an approval to backport this test fix into jdk8u. To address the issue with NoClassDefFoundError I added the implicit @build line to the test (as it was done in JDK-8043520).

BUGURL: https://bugs.openjdk.java.net/browse/JDK-6545422 WEBREV: http://cr.openjdk.java.net/~igerasim/6545422/0/webrev/ Jdk9 change: http://hg.openjdk.java.net/jdk9/dev/jdk/rev/efeab0eae50f Review: http://mail.openjdk.java.net/pipermail/serviceability-dev/2014-February/014152.html

Would you please approve the backport?

Sincerely yours, Ivan

On 24.02.2014 13:15, Ivan Gerasimov wrote:

On 24.02.2014 13:03, Alan Bateman wrote: On 24/02/2014 08:56, Ivan Gerasimov wrote:

Hello!

Would you please approve porting back this test bug fix? The fix applies cleanly to jdk8u. Master Bug: https://bugs.openjdk.java.net/browse/JDK-6545422 Jdk9 change: http://hg.openjdk.java.net/jdk9/dev/jdk/rev/efeab0eae50f Review: http://mail.openjdk.java.net/pipermail/serviceability-dev/2014-February/014152.html Ivan - would it be possible to hold back on this one for a bit. The updated test is dependent on implicit compilation of the classes in the @library and this is causing problems for concurrent test runs. The same pattern has slipped into 15-20 other tests with the result that they all fail intermittently due to NoClassDefFoundError issues. I think they should be fixed in jdk9-dev but any of these "improved" tests are backported. Yes, sure! Actually, I was going to port this fix into jdk6 (the test fails there), but modified in such a way that it does not depend on testlibrary. I can still work on jdk6 port without first integrating it into jdk8u and jdk7u. Sincerely yours, Ivan -Alan.

More information about the core-libs-dev mailing list