Created on 2012-09-09 22:50 by christian.heimes, last changed 2012-09-24 05:46 by python-dev. This issue is now closed.
| Files | ||||
|---|---|---|---|---|
| File name | Uploaded | Description | Edit | |
| closeown.patch | christian.heimes, 2012-09-09 22:50 | review | ||
| Messages (12) | |||
|---|---|---|---|
| msg170142 - (view) | Author: Christian Heimes (christian.heimes) *  |
Date: 2012-09-09 22:50 | |
In Python/pythonrun.c PyRun_SimpleFileExFlags() may leak a FILE pointer if closeit is False and maybe_pyc_file returns true. This happens when the filename ends with ".pyc" or ".pyc". In this case the file is opened a second time with fopen() a few lines lower and assigned to the same variable "fp". However "fp" is never closed. The bug was found by Coverity as CID 719689. Georg: I've set the priority to release blocker as a resource leak of a file pointer and therefore rare file descriptor is IMHO a severe issue. |
|||
| msg170149 - (view) | Author: Jesús Cea Avión (jcea) * |
Date: 2012-09-10 02:24 | |
LGTM. Instead of using a flag variable, could you use something like "if (fp!=NULL) fclose(fp);"? |
|||
| msg170169 - (view) | Author: Christian Heimes (christian.heimes) * |
Date: 2012-09-10 09:31 | |
No, that's not possible. 'fp' is part of the argument to PyRun_SimpleFileExFlags() and never NULL. We need some way to distinguish a externally provided fp from a self-opened fp and fclose(fp) in the latter case. |
|||
| msg170208 - (view) | Author: Jesús Cea Avión (jcea) * |
Date: 2012-09-10 17:58 | |
Look good to me, then. Are you committing the patch yourself?. You can patch 2.7, 3.2 and default, and bug Georg for 3.3.0 inclusion (patches to default will be in 3.3.1), unless Georg is notified and chooses to cherry pick it. |
|||
| msg170296 - (view) | Author: Roundup Robot (python-dev) | Date: 2012-09-11 12:12 | |
New changeset 4754c4a710e6 by Christian Heimes in branch 'default': Issue #15895: Fix FILE pointer leak in PyRun_SimpleFileExFlags() when filename points to a pyc/pyo file and closeit is false. http://hg.python.org/cpython/rev/4754c4a710e6 |
|||
| msg170298 - (view) | Author: Christian Heimes (christian.heimes) * |
Date: 2012-09-11 12:15 | |
The bug was a 3.3 regression, possibly introduced with the merge of importlib. I forgot to tick the 3,3regression keyword. I've applied the fix. I'm leaving this ticket open and assigned to Georg for 3.3.0. |
|||
| msg170306 - (view) | Author: Roundup Robot (python-dev) | Date: 2012-09-11 13:47 | |
New changeset c3539eb02470 by Christian Heimes in branch 'default': Issue #15895: my analysis was slightly off. The FILE pointer is only leaked when set_main_loader() fails for a pyc file with closeit=0. In the success case run_pyc_file() does its own cleanup of the fp. I've changed the code to use another FILE ptr for pyc files and moved the fclose() to PyRun_SimpleFileExFlags() to make it more obvious what's happening. http://hg.python.org/cpython/rev/c3539eb02470 |
|||
| msg170331 - (view) | Author: Roundup Robot (python-dev) | Date: 2012-09-11 17:28 | |
New changeset 6fea947edead by Christian Heimes in branch 'default': Updates NEWS for issue #15895 http://hg.python.org/cpython/rev/6fea947edead |
|||
| msg170932 - (view) | Author: STINNER Victor (haypo) * |
Date: 2012-09-21 21:52 | |
This issue is not a regression, I think that the fix can wait for Python 3.3.1. |
|||
| msg170933 - (view) | Author: STINNER Victor (haypo) * |
Date: 2012-09-21 21:56 | |
> This issue is not a regression My bad, Crys explained me on IRC that the leak does not exist in Python 3.2, it was introduced by the importlib. So it is a regression. |
|||
| msg170958 - (view) | Author: Georg Brandl (georg.brandl) * |
Date: 2012-09-22 06:53 | |
Now transplanted to release clone. |
|||
| msg171094 - (view) | Author: Roundup Robot (python-dev) | Date: 2012-09-24 05:46 | |
New changeset fe4e6e0d6b0f by Christian Heimes in branch 'default': Issue #15895: Fix FILE pointer leak in PyRun_SimpleFileExFlags() when filename points to a pyc/pyo file and closeit is false. http://hg.python.org/cpython/rev/fe4e6e0d6b0f New changeset 9251fb0440a0 by Christian Heimes in branch 'default': Issue #15895: my analysis was slightly off. The FILE pointer is only leaked when set_main_loader() fails for a pyc file with closeit=0. In the success case run_pyc_file() does its own cleanup of the fp. I've changed the code to use another FILE ptr for pyc files and moved the fclose() to PyRun_SimpleFileExFlags() to make it more obvious what's happening. http://hg.python.org/cpython/rev/9251fb0440a0 New changeset 61b1c25d55f2 by Christian Heimes in branch 'default': Updates NEWS for issue #15895 http://hg.python.org/cpython/rev/61b1c25d55f2 |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2012-09-24 05:46:45 | python-dev | set | messages: + msg171094 |
| 2012-09-22 06:53:35 | georg.brandl | set | status: open -> closed messages: + msg170958 |
| 2012-09-21 21:56:10 | haypo | set | messages: + msg170933 |
| 2012-09-21 21:52:45 | haypo | set | nosy:
+ haypo messages: + msg170932 |
| 2012-09-11 17:28:52 | python-dev | set | messages: + msg170331 |
| 2012-09-11 13:47:38 | python-dev | set | messages: + msg170306 |
| 2012-09-11 12:15:10 | christian.heimes | set | versions:
+ Python 3.3, Python 3.4 messages: + msg170298 keywords: + 3.3regression resolution: fixed stage: patch review -> resolved |
| 2012-09-11 12:12:32 | python-dev | set | nosy:
+ python-dev messages: + msg170296 |
| 2012-09-10 17:58:01 | jcea | set | messages: + msg170208 |
| 2012-09-10 09:31:34 | christian.heimes | set | messages: + msg170169 |
| 2012-09-10 02:24:49 | jcea | set | nosy:
+ jcea messages: + msg170149 |
| 2012-09-09 22:50:50 | christian.heimes | create | |