[foreign-memaccess] RFR 8234814: Eager layout size computation trips on unbound sequence layouts

[Maurizio Cimadamore]

After a discussion offline with Jorn I've decided to revert the changes 
to constants:

cr.openjdk.java.net/~mcimadamore/panama/8234814_v3

As some of the tests demonstrate, among the original goals of these 
JAVA_INT constants and friends was interop with ByteBuffers, hence the 
endianness set to BE.

I'll leave it as is for now (after all, the behavior is documented). We 
can revise later, if needs be.

Maurizio

On 26/11/2019 14:46, Maurizio Cimadamore wrote:
>
> On 26/11/2019 14:40, Maurizio Cimadamore wrote:
>> * constants such as JAVA_INT should use native order (this is, after 
>> all, the layout of an int in the VM) 
>
> To expand a bit - we eventually would like to see these constants 
> moved in the primitive wrapper classes - e.g.
>
> MemoryLayout::JAVA_INT -> Integer::LAYOUT
>
> in which case I think these constants should return whatever layout 
> the VM thinks a Java int (in this case) has.
>
> Maurizio
>