Querstion about ForkJoinPool / SecurityManager interoperability

Doug Lea dl at cs.oswego.edu
Thu Dec 13 14:15:12 UTC 2018

On 12/13/18 8:44 AM, Patrick Reinhart wrote:
> This special case could have been handled also by the
> InnocuousForkJoinWorkerThread
> could in my opinion be relaxed to accept null or the system classloader
> to be set
> using setContextClassLoader() 

Thanks. We should/will do this. The unconditional throw was clearly too
strong; innocuous calls can be allowed. This doesn't address the general
issues of dynamic security manager installation, but at least removes an
obstacle for people trying to cope.

I created CR: https://bugs.openjdk.java.net/browse/JDK-8215359


More information about the core-libs-dev mailing list