[8u-dev] RFR 8190690: Impact on krb5 test cases in the 8u-CPU nightly

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

[8u-dev] RFR 8190690: Impact on krb5 test cases in the 8u-CPU nightly

Weijun Wang
Please review the fix at

   http://cr.openjdk.java.net/~weijun/8190690/webrev.00/

Before this fix, a static block in KDC.java set a customized name service provider. Some recent change (jdk8u161?) inside JDK might have added a name lookup before this provider is set and the provider is never loaded.

This is a <=jdk8u issue. We switch to a different mechanism in jdk9.

This fix just sets the provider right on the command line, for example:

- * @run main/othervm Basic
+ * @run main/othervm -Dsun.net.spi.nameservice.provider.1=ns,mock Basic

BasicProc.java and ReplayCacheTestProc.java launch child processes, the system property is also set on the child processes.

Thanks
Max

Reply | Threaded
Open this post in threaded view
|

Re: [8u-dev] RFR 8190690: Impact on krb5 test cases in the 8u-CPU nightly

Seán Coffey
Looks like some test clients can have old configuration files which
trigger early loading of InetAddress. Your edits look fine and hardens
the testcase against such issues.

regards,
Sean.

On 09/11/2017 09:04, Weijun Wang wrote:

> Please review the fix at
>
>     http://cr.openjdk.java.net/~weijun/8190690/webrev.00/
>
> Before this fix, a static block in KDC.java set a customized name service provider. Some recent change (jdk8u161?) inside JDK might have added a name lookup before this provider is set and the provider is never loaded.
>
> This is a <=jdk8u issue. We switch to a different mechanism in jdk9.
>
> This fix just sets the provider right on the command line, for example:
>
> - * @run main/othervm Basic
> + * @run main/othervm -Dsun.net.spi.nameservice.provider.1=ns,mock Basic
>
> BasicProc.java and ReplayCacheTestProc.java launch child processes, the system property is also set on the child processes.
>
> Thanks
> Max
>