RFR 8143628: Fork sun.misc.Unsafe and jdk.internal.misc.Unsafe native method tables
paul.sandoz at oracle.com
Mon Nov 30 09:07:22 UTC 2015
> On 29 Nov 2015, at 17:08, Alan Bateman <Alan.Bateman at oracle.com> wrote:
> On 27/11/2015 09:06, Paul Sandoz wrote:
>>> Also wondering how these tests will access jdk.internal once we're using the modular image?
>> I believe it is sufficient to include:
>> 28 * @modules java.base/jdk.internal.misc
>> 28 * @modules java.base/sun.misc
>> We will need to update all tests depending on sun.misc.Unsafe when it moves to a separate module.
> The changes look okay to me too.
> If Spp were more widely used when we might choose an option other than -nel but what you have is fine.
Yes, that option name is a marginal improvement on the name in the VHs sandbox repo :-)
> On @modules then it would be good to add to get @modules java.base/jdk.internal.misc on any tests using the rehoused Unsafe. That makes it easier the jake forest is sync'ed up. If there are tests using sun.misc.Unsafe then @modules isn't needed. If there are tests with @modules java.base/sun.misc then we'll do a bulk update once the separation is complete.
I presume Chris will do a sweep over all this at some point when ready.
More information about the core-libs-dev