RFC: Fix for boolean arguments in interpreter

[Andrew Haley]

Mark Wielaard writes:
 > Hi,
 > 
 > This implements option 2 of the various ways that Anthony Green
 > described to fix bug #21115 "false boolean argument passed from
 > pre-compiled to interpreted method is true"
 > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21115
 > 
 > I think this is a fix with the least impact and I would like to have a
 > fix on the 4.0.x branch for this. One improvement to this could be to
 > add a flag to _Jv_InterpMethod to indicate whether the method needs to
 > fixup its arguments so we only have to scan the signature once.
 > 
 > Comments?

Looks good.  However, we should perhaps think about whether this
approach is portable to all systems.  There's also the matter of
16-bit args.

Andrew.