Virtual function call stuff, again

[Mark Mitchell]

>>>>> "Mark" == Mark Mitchell <mark at codesourcery.com> writes:

    Mark>   Furthermore, the actual algorithm used to perform the
    Mark> adjustments does not seem necessarily to be part of the ABI.

Similarly, I don't think the ABI should specify that the caller
adjusts the `this' pointer to the class containing the unique final
overrider.  *Any* class containing a virtual function table entry for
this function will do, and an implementation should be allowed to such
a class, if it wants to.  (Picking another class is probably inferior,
since that method is certain to require a `this' adjustment, while
picking the class with the unique final overrider may not, but that's
a quality-of-implementation issue.)

--
Mark Mitchell                   mark at codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com