[
https://jira.codehaus.org/browse/SUREFIRE-583?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=356166#comment-356166
]
Tibor Digana commented on SUREFIRE-583:
---------------------------------------
@Justin
Would you provide logs from the command "mvn -X -e test" ?
> When forking and specifying a JVM, that JVM's security policy's JCE providers
> are not loaded, JAVA_HOME's are
> -------------------------------------------------------------------------------------------------------------
>
> Key: SUREFIRE-583
> URL: https://jira.codehaus.org/browse/SUREFIRE-583
> Project: Maven Surefire
> Issue Type: Bug
> Components: process forking
> Affects Versions: 2.4.2
> Environment: Windows, JAVA_HOME is Sun JDK 1.6.0u16, forked JVM is
> IBM JDK for WAS 6.1
> Reporter: Justin Searls
> Assignee: Tibor Digana
> Fix For: Backlog
>
>
> Premise:
> My test needs to run on the IBM JDK to work, but for other reasons I need to
> actually build on the Sun JVM. My application's tests are relying on
> libraries that use a message digest ("SHA", not "SHA1") that I can only find
> support for in the BouncyCastle JCE provider.
> Setup:
> 1. So I've identified in my plugin configuration something like
> <jvm>/path/to/ibm/jdk/jre/bin/javaw.exe</jvm>
> 2. Added BouncyCastle JCE provider jar to /path/to/ibm/jdk/jre/lib/ext
> 3. Setup BouncyCastle as the sole JCE provider in
> /path/to/ibm/jdk/jre/lib/security/java.security
> Expected Result: Designated IBM JVM would look for its java.security file and
> load its jre/lib/ext JARs when executing tests
> Actual Result: No such effect. After going through the same setup on my Sun
> JDK (which I'm running Maven with), that did have the effect of actually
> providing that provider and getting past the error I was experiencing.
> It seems to me that if you fork to a different JVM, that JVM's security
> policy should be used. Given the complexity of this API, however, I wouldn't
> be surprised to hear that there's a major technical hurdle in implementing
> this, however.
--
This message was sent by Atlassian JIRA
(v6.1.6#6162)
