[8u20] Request for approval: 6545422: TEST BUG: NativeErrors.java uses wrong path name in exec
ivan.gerasimov at oracle.com
Tue Jul 1 21:29:50 UTC 2014
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).
Jdk9 change: http://hg.openjdk.java.net/jdk9/dev/jdk/rev/efeab0eae50f
Would you please approve the backport?
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:
>>> 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
>> 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
> I can still work on jdk6 port without first integrating it into jdk8u
> and jdk7u.
> Sincerely yours,
More information about the core-libs-dev