#12313: Fix yet another memory leak caused by caching of coercion data
-------------------------+--------------------------------------------------
Reporter: SimonKing | Owner: rlm
Type: defect | Status: new
Priority: major | Milestone: sage-5.0
Component: memleak | Keywords: coercion weak dictionary
Work_issues: | Upstream: N/A
Reviewer: | Author:
Merged: | Dependencies: #715
-------------------------+--------------------------------------------------
Comment(by SimonKing):
Memory leaks are quite tricky!!
The post here is mainly a memo for myself, documenting what went wrong in
the last couple of days.
I tried to directly modify my code from #715. The approach was:
`TripleDict` or `MonoDict` should store the data in lists, with some list
items being the address of the keys; and in a separate dictionary, a
`weakref.KeyedRef` with a callback function is preserved.
And what happened when I replaced `Parent._coerce_from_hash` by a
`MonoDict`? A memory leak was ''created'' by using weak references!
Namely,
* There is a strong reference to `weakref.KeyedRef`.
* The `weakref.KeyedRef` has a strong reference to the callback function
* The callback function is an instance of something that I'll call
`MonoDictEraser` (similar to the `TripleDictEraser` from #715)
* The Eraser has a strong reference to the `MonoDict`
* The `MonoDict` has a strong reference to the coerce map.
* The coerce map has a strong reference to domain and codomain
* Hence, domain and codomain can't be collected.
I think the best place to break the reference chain is from the Eraser to
the `MonoDict`. Perhaps one could do the same thing at #715, rather than
having a weak reference to the underlying set of an action?
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/12313#comment:1>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sage-trac?hl=en.
