1) Unpacking a Sub PMC does: - create a constant Sub PMC - put that in the global hash with the (label-)name as key - finally make absolute (start,end) addresses from the offsets
2) So the question is: Shouldn't it be the other way round: - unpack (thaw) the global hash - which unpacks its members, i.e. the subroutines (and global vars) - fixup is done in thaw_finish
The same occurs with classes which are collected in the interpreters classhash.
The differences between 1) and 2) are: 1) needs more knowledge about thawed PMCs (put Subs into the globals...) 2) needs a thaw option to extend the given (hash) PMC, but is much simpler
Thougts?
leo