Messages (8) msg301882 - (view) Author: STINNER Victor (vstinner) * Date: 2017-09-11 16:29

The following commands show memory leaks introduced by bpo-30860:

$ ./python -m test -R 3:3 test_atexit -m test.test_atexit.SubinterpreterTest.test_callbacks_leak

$ ./python -m test -R 3:3 test_atexit -m test.test_atexit.SubinterpreterTest.test_callbacks_leak_refcycle

$ ./python -m test -R 3:3 test_threading -m test.test_threading.SubinterpThreadingTests.test_threads_join

$ ./python -m test -R 3:3 test_capi -m test.test_capi.SubinterpreterTest.test_subinterps

msg301883 - (view) Author: STINNER Victor (vstinner) * Date: 2017-09-11 16:30

> The following commands show memory leaks introduced by bpo-30860:

Oh, I'm talking about the commit 2ebc5ce42a8a9e047e790aefbf9a94811569b2b6.

msg301886 - (view) Author: Stefan Krah (skrah) * Date: 2017-09-11 16:36

Could be the same as #31408.

msg301889 - (view) Author: STINNER Victor (vstinner) * Date: 2017-09-11 17:14

The leak comes from the following lines of _PySys_EndInit():
---
    PyObject *warnoptions = get_warnoptions();
    if (warnoptions == NULL)
        return -1;
    SET_SYS_FROM_STRING_BORROW_INT_RESULT("warnoptions", warnoptions);

    PyObject *xoptions = get_xoptions();
    if (xoptions == NULL)
        return -1;
    SET_SYS_FROM_STRING_BORROW_INT_RESULT("_xoptions", xoptions);
---

It's not the first time that I have an issue with these attributes. The last reference weak caused by multiple interpreters was also related to this one if I recall correctly.

See bpo-30598 and my commit 865de27dd79571a4a5c7a7d22a07fb909c4a9f8e.

msg301891 - (view) Author: STINNER Victor (vstinner) * Date: 2017-09-11 17:16

Extract of my msg295399: "The problem is that warnoptions is stored in a C global variable *and* in sys.warnoptions of each interpreter. The ownership of this variable is unclear."

Maybe we need a change similar to bpo-28411 (commit 86b7afdfeee77993fe896a2aa13b3f4f95973f16) which removed the "modules" field from Py_InterpreterState.

msg301904 - (view) Author: Eric Snow (eric.snow) * Date: 2017-09-11 21:21

I'm looking into this.

msg301906 - (view) Author: Eric Snow (eric.snow) * Date: 2017-09-11 21:31

Moving warnoptions (and xoptions) out of PyInterpreterState seems like a good idea to me for the same reasons as applied to sys.modules.  The case for these two is even stronger since they are only used in sysmodule.c.

msg302154 - (view) Author: Eric Snow (eric.snow) * Date: 2017-09-14 07:36

New changeset dae0276bb6bc7281d59fb0b8f1aab31634ee80dc by Eric Snow in branch 'master':
bpo-30860: Fix a refleak. (#3567)
https://github.com/python/cpython/commit/dae0276bb6bc7281d59fb0b8f1aab31634ee80dc