At 10:33 AM +0200 6/24/03, Leopold Toetsch wrote:
I've added a check for too deeply nested stacks now.
I probably ought to get started on the stack-chunk-as-PMC patch for garbage collection of stack frames. :)
The first question is: Do we need such a limit check on the register backing stacks too?
Second is efficiency on copying COWed stacks. Currently only Control, Pad, and User stacks are lazy copied, whole stack at once and not chunkwise. So it would make sense to have stack chunks derived from buffer-like objects, which can be unCOWed per chunk.
But this would need some rework in the COW GC code, which currently assumes that only strings might be COWed.
WRT coroutines: They have new_stack()s for Pad, User, and Control. The latter is probably wrong due to the planned exception mechanism. I presume that a coroutine works like a subroutine WRT exceptions, so that some parent can catch an exception that has originated in a coroutine.
leo
