On Wed, Dec 12, 2007 at 03:45:09PM +0800, zou lunkai wrote: > see added tests at the end of Function.as. > > I have a feeling that function is not special, but a group of opcodes > that shares the same VM stack as its caller. As observed from the > tests, the caller's runtime stack could be completely violated when > calling a function.
I guess we both know this, as_environment only beeing still needed for the 'target' member. I belive Chad was working on unifying the stack. > Next step would be checking if the vm stack should be protected at > frame boundary. Indeed. Gnash GC relies on the assumption that all stacks are empty after an advance. If that needs to be changed we'll need to take care of properly marking stack elements as reachable. --strk; _______________________________________________ Gnash-commit mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnash-commit
