BST: remove some of the remaining pointers, earlier boxed code destruction #1384
Conversation
we don't need it anymore now that we eliminated nearly all pointers to nodes.
now that the FlattenVisitor is gone nobody uses this anymore
we will soon need this
…ointer Replace it with a index into a table of definitions and code objects This also now frees much more code objects early because they get proper refcounted now
…able In the future we should just support embedding tuples as constants (of course only when all elements are constant) and then use this mechanism instead of the 'std::vector<BoxedString*>'
|
Not critical for this PR, but re the "remove pointer to FunctionDef and ClassDef" commit, I think we can clean that up -- we should be able to use the same constants strategy for the BoxedCode objects that we do for the other objects. And I think at this point we shouldn't need the parent BST node to keep a reference to the child BST node? I haven't verified that but it seems like all of the relevant information should already be included in the code object. I'd also like to see us kill InternedString rather than keep cludging it. I think your Anyway, these are just cleanups, which we can do at any time. |
This is a prerequisite for #1381 and removes the
InternedStringfields and pointers toBoxedCode. The code objects are now all currently refcounted which means we will free them when the are unused.