RFR: 8349146: [REDO] Implement a better allocator for downcalls [v6]

[Per Minborg]

On Thu, 1 May 2025 09:09:16 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

>> Ugh - it's a record, I see. Which brings up the question -- should it be a record? Should it advertize its `size and `tl` components (they seem rather implementation specific?).
>
> Was this done so that the final field could be trusted? (I believe all final fields are implicitly trusted in jdk/internal/foreign):
> 
> 
>       holder->is_in_package("jdk/internal/foreign/layout") || holder->is_in_package("jdk/internal/foreign") ||

Yes. We should revert to a normal class.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/24829#discussion_r2071249389