verifier problem

Sun Jan 27 16:03:00 GMT 2002

>>>>> "Per" == Per Bothner <per@bothner.com> writes:

Per> I think (without having analyzed it too carefully) that trying to
Per> test or set the subroutine field of a 'state' in merge is wrong.
I did this because of this text in JVMS 4.9.6:
 When executing the ret instruction, [ ... ], there must be only
 one possible subroutine from which the instruction can be
 returning. Two different subroutines cannot "merge" their
 execution to a single ret instruction.
But now I see I implemented this incorrectly. We should only be doing
this check when processing `ret', not when merging states.
I'll look into this more deeply a bit later.
Tom