#19244: Broken doctest in src/sage/categories/fields.py
-------------------------------------+-------------------------------------
Reporter: jdemeyer | Owner:
Type: defect | Status: needs_review
Priority: blocker | Milestone: sage-6.9
Component: memleak | Resolution:
Keywords: random_fail | Merged in:
Authors: Simon King | Reviewers:
Report Upstream: N/A | Work issues:
Branch: | Commit:
u/SimonKing/broken_doctest_in_src_sage_categories_fields_py|
0d97c7db960aa058b7cb1c60aea839dbc3d161e9
Dependencies: | Stopgaps:
-------------------------------------+-------------------------------------
Changes (by {'newvalue': u'Simon King', 'oldvalue': ''}):
* status: new => needs_review
* commit: => 0d97c7db960aa058b7cb1c60aea839dbc3d161e9
* component: categories => memleak
* author: => Simon King
Comment:
I have seen that repeatedly. I think at some point I suggested to test
that the difference is not zero but is at most zero. Now I suggest a
different solution.
We have seen the following phenomenon while we introduced Sage's weak
coercion dictionaries. It is possible that `gc.collect()` collects some
objects, and by this collection *other* objects become collectable, which
however will only be collected during a subsequent cyclic garbage
collection. One can construct examples such that n cyclic garbage
collections will not collect a certain object, but the (n+1)-st cyclic
garbage collection *will* collect it.
It could be that we see this phenomenon here in real life. I.e., it could
be that the two `gc.collect()` commands in the test do not make a certain
field collectable that was created in a different doctest. However, a
third collection step that may or may not happen in the for-loop does make
it collectable.
Thus, depending on whether or not a third collection happens, we see a
failure or not.
If my explanation is correct, then disabling the garbage collection till
we really want it to happen should fix the test permanently.
----
New commits:
||[http://git.sagemath.org/sage.git/commit/?id=0d97c7db960aa058b7cb1c60aea839dbc3d161e9
0d97c7d]||{{{Make garbage collection deterministic during a test against a
memory leak}}}||
--
Ticket URL: <http://trac.sagemath.org/ticket/19244#comment:2>
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 unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/d/optout.
