[nestmates] validation of ClassFile for hidden classes

[John Rose]

On Nov 13, 2019, at 8:49 PM, David Holmes <david.holmes at oracle.com> wrote:
> 
> We would want the programmatic nesthost to take precedence, then try to resolve the static NH etc.

Yes, determining the dynamic nest host of a HC must happen before the attrbutes
are examined.  In this sense the attributes will be ignored for HCs.  The JVM has
a rule for computing the host H or H'.

I think it’s harmless (in practice) that the derivation of the dynamic nest host H’ or H
for the non-HC part of a resolution request might throw because a non-hidden
class has a broken attribute.  That’s just the cost of doing business with the
bidirectional relations of the NH/NMs attributes.

> Not sure if Mandy has the reflection aspect documented outside of email at the moment.

I think she has that covered in the API spec for jl.Class.

http://cr.openjdk.java.net/~mchung/valhalla/webrevs/8171335/specdiff/java.base/java/lang/Class.html

— John