RFC: fix PR 28892

Tom Tromey tromey@redhat.com
Thu Aug 31 19:04:00 GMT 2006

>>>>> "David" == David Daney <ddaney@avtrex.com> writes:

David> If you try to modify a final field directly (after it has been set)
David> the verifier must reject the code.  So the only option would be to use
David> reflection, for which it is only supposed to work in a very narrow set
David> of circumstances.

.. and the special cases of System.{in,out,err}, via the set* methods.

But, yeah, I'd forgotten about the non-static constraint.
In any case, I think this only applies to subsequent sets.  I don't
think there is any requirement that the initial set of a static final
variable be done in <clinit>.


More information about the Gcc-patches mailing list