[foreign-memaccess+abi] RFR: 8278151: Heap segments should handle alignment constraints in a deterministic fashion [v5]

[Maurizio Cimadamore]

On Fri, 3 Dec 2021 19:41:16 GMT, John R Rose <jrose at openjdk.org> wrote:

> I would prefer to reserve the term "hyper aligned" for the RFE in https://bugs.openjdk.java.net/browse/JDK-8167334 which talks about support for 128-bit alignment (and more) in the JVM. The odd condition being called "hyper aligned" in this PR is, IIUC, an overflow condition where the alignment of a layout exceeds its size.

I don't think the term leaks anywhere in the API - but only in an helper function and some tests. That said, I thought the term was fitting to describe a datum whose alignment constraints were bigger than its natural one (read, the size). But I'm happy to change the names there.

-------------

PR: https://git.openjdk.java.net/panama-foreign/pull/622