RFR: JDK-8005845 (JEP 149) Fully Concurrent ClassLoading - VM Changes

[Christian Thalinger]

On Jan 8, 2013, at 2:11 AM, David Holmes <David.Holmes at oracle.com> wrote:

> As described here:
> 
> https://blogs.oracle.com/dholmes/entry/parallel_classloading_revisited_fully_concurrent
> 
> and discussed here:
> 
> http://mail.openjdk.java.net/pipermail/core-libs-dev/2012-December/012658.html
> 
> FullyConcurrent classloaders allow for parallel-defineClass to occur. These are the VM changes to support that:
> 
> http://cr.openjdk.java.net/~dholmes/8005845/webrev/

One nit:

    template(parallelCapable_name,                      "parallelLockMap")                                          \
+   template(fullyConcurrent_name,                      "isFullyConcurrent")                                          \
                                                                                                                    \

Please keep the backslashes aligned.

-- Chris

> 
> Quite simply we allow access to the fullyConcurrent field in java.lang.ClassLoader and if it is fullyConcurrent then we allow a parallel-defineClass to occur.
> 
> In a JDK without the ClassLoader change this is all a no-op.
> 
> Thanks,
> David