Re: binary compatibility ABI (was: Re: [boehm-gc] Import 6.3 alpha1)

[Jeff Sturm <jsturm at one-point dot com>]

On Thu, 14 Aug 2003, Andrew Haley wrote:
>  > static int a, b;
>  > ...
>  > return a + b;
>  >
>  > Does this require 2, 3 or 4 memory loads?
>  >
>  > (I believe 2 is the answer for non-PIC, 4 for PIC.  If we take advantage
>  > of the fact that static fields are at a fixed offset from the class
>  > record, it can be done in three, while preserving position independence.)
>
> Unless a and b are in the exact same class as the 'return a + b'
> statment, the positions of a and b relative to the class record may
> change.

Right.  Well, provided they are in the same compilation unit anyway, not
necessarily the same class.  And if the class/method symbols do become
non-public in the new ABI that'll be enforced.

> I guess this isn't true if the whole application is compiled
> and we give up on binary compatibility.

Suppose gcj defaults to binary compability between compilation units, but
relaxes all the rules within a single compilation unit.

The nice thing about that is a "compilation unit" can be a whole package
or even an entire application if you like.

Jeff