Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.
Sign upbpo-39395: Clear env vars set by Python at exit #18078
Conversation
Environment variables set by os.environ and os.putenv() are now cleared at exit. Python manages their memory which is released at exit.
This comment has been minimized.
This comment has been minimized.
cc @eduardo-elizondo @ericsnowcurrently @serhiy-storchaka @encukou (author and reviewers of PR #10854 and PR #15892). |
static int | ||
_posix_clear(PyObject *module) | ||
{ | ||
Py_CLEAR(_posixstate(module)->billion); | ||
#ifdef HAVE_UNSETENV | ||
posix_unset_envvars(_posixstate(module)->posix_putenv_garbage); | ||
#endif | ||
Py_CLEAR(_posixstate(module)->posix_putenv_garbage); |
This comment has been minimized.
This comment has been minimized.
eduardo-elizondo
Jan 20, 2020
•
Contributor
Given that you are already doing these fixes around this, posix_putenv_garbage
should also be guarded by HAVE_PUTENV
(check its initialization in the init function).
This comment has been minimized.
This comment has been minimized.
Good catch, looks good to me |
This comment has been minimized.
This comment has been minimized.
Should the docs for os.environ and os.putenv be updated to note this explicit deletion? Is it possible that, on some platforms, existing programs might be using the current undefined behavior without crashing and thus might break with this change so it should be documented? |
This comment has been minimized.
This comment has been minimized.
Yes, it is possible. Python 3.8 was safe: the memory was never released, so environment variables remained available and didn't crash. I will first implement https://bugs.python.org/issue39406 before updating this PR to document the incompatible change. |
vstinner commentedJan 20, 2020
•
edited by bedevere-bot
Environment variables set by os.environ and os.putenv() are now
cleared at exit. Python manages their memory which is released
at exit.
https://bugs.python.org/issue39395