Issue 33622: Fix and improve errors handling in the garbage collector (original) (raw)
Issue33622
Created on 2018-05-23 17:41 by serhiy.storchaka, last changed 2022-04-11 14:59 by admin. This issue is now closed.
Pull Requests | |||
---|---|---|---|
URL | Status | Linked | Edit |
PR 7078 | merged | serhiy.storchaka,2018-05-23 17:50 | |
PR 7094 | merged | miss-islington,2018-05-24 12:19 | |
PR 7095 | merged | serhiy.storchaka,2018-05-24 12:30 | |
PR 7096 | merged | serhiy.storchaka,2018-05-24 12:30 | |
PR 7126 | merged | serhiy.storchaka,2018-05-26 10:07 |
Messages (8) | ||
---|---|---|
msg317431 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-23 17:41 |
There are following bugs in the garbage collector. * If the garbage collector fails to add an object with __del__ or referenced by an object with __del__ to gc.garbage (in handle_legacy_finalizers()), it leaks it and other not added objects with __del__ and referenced by them. They become no longer accessible by the garbage collector. * PyGC_Collect() is not documented, but it is a public C API. And it can be called by user with an exception set. PyGC_Collect() then can either crash or just silent the exception. It is safer to safe possible exception and restore it after collecting. * A pointer to released member can be used (compared with NULL) in invoke_gc_callback(). This is an undefined behavior. | ||
msg317567 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-24 12:19 |
New changeset 301e3cc8a5bc68c5347ab6ac6f83428000d31ab2 by Serhiy Storchaka in branch 'master': bpo-33622: Fix issues with handling errors in the GC. (GH-7078) https://github.com/python/cpython/commit/301e3cc8a5bc68c5347ab6ac6f83428000d31ab2 | ||
msg317568 - (view) | Author: miss-islington (miss-islington) | Date: 2018-05-24 12:39 |
New changeset 2fe940c727802ad54cff9486c658bc38743f7bfc by Miss Islington (bot) in branch '3.7': bpo-33622: Fix issues with handling errors in the GC. (GH-7078) https://github.com/python/cpython/commit/2fe940c727802ad54cff9486c658bc38743f7bfc | ||
msg317569 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-24 13:05 |
New changeset f0e04b2ae27a4da940a76ae522f0438f80ce5d5a by Serhiy Storchaka in branch '3.6': [3.6] bpo-33622: Fix issues with handling errors in the GC. (GH-7078) (GH-7095) https://github.com/python/cpython/commit/f0e04b2ae27a4da940a76ae522f0438f80ce5d5a | ||
msg317617 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-24 20:27 |
New changeset 9044cd675902b438bc500908e410382ff48299d8 by Serhiy Storchaka in branch '2.7': [2.7] bpo-33622: Fix issues with handling errors in the GC. (GH-7078) (#7096) https://github.com/python/cpython/commit/9044cd675902b438bc500908e410382ff48299d8 | ||
msg317732 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-26 10:13 |
PR 7126 adds several assert(!PyErr_Occurred()) for earlier detection of leaked exceptions. And the failure in adding to gc.garbage is no longer fatal. It is better to not add an object to gc.garbage in this iteration than crash when we run garbage collection when have no memory. | ||
msg318048 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-29 15:50 |
New changeset c4653c9bf159c3919a50f4ced32eef713e7e764e by Serhiy Storchaka in branch 'master': bpo-33622: Add checks for exceptions leaks in the garbage collector. (GH-7126) https://github.com/python/cpython/commit/c4653c9bf159c3919a50f4ced32eef713e7e764e | ||
msg318290 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * ![]() |
Date: 2018-05-31 12:30 |
See , and for three cases in the stdlib where an exception can be set in the tp_clear handler. This caused a crash in the garbage collector, or can be just silenced if failed at the shutdown stage. In the master branch it will cause writing a traceback to the stderr. It would be better to handle exceptions locally in the tp_clear handlers. But perhaps it may be worth to handle leaked exceptions in the garbage collector too. I'm just not sure about a solution. Writing a traceback to the stderr can cause a regression if it was just silenced before. But silencing it can hide bugs. |
History | |||
---|---|---|---|
Date | User | Action | Args |
2022-04-11 14:59:00 | admin | set | github: 77803 |
2019-03-21 13:09:05 | serhiy.storchaka | set | status: open -> closedresolution: fixedstage: patch review -> resolved |
2018-05-31 12:30:05 | serhiy.storchaka | set | messages: + |
2018-05-29 15:50:13 | serhiy.storchaka | set | messages: + |
2018-05-26 10:13:29 | serhiy.storchaka | set | messages: + |
2018-05-26 10:07:43 | serhiy.storchaka | set | stage: patch reviewpull_requests: + <pull%5Frequest6761> |
2018-05-26 10:02:57 | serhiy.storchaka | set | status: closed -> opentitle: Fix errors handling in the garbage collector -> Fix and improve errors handling in the garbage collectorresolution: fixed -> (no value)stage: resolved -> (no value) |
2018-05-25 14:14:35 | serhiy.storchaka | set | status: open -> closedresolution: fixedstage: patch review -> resolved |
2018-05-24 20:27:11 | serhiy.storchaka | set | messages: + |
2018-05-24 13:05:43 | serhiy.storchaka | set | messages: + |
2018-05-24 12:39:11 | miss-islington | set | nosy: + miss-islingtonmessages: + |
2018-05-24 12:30:35 | serhiy.storchaka | set | pull_requests: + <pull%5Frequest6732> |
2018-05-24 12:30:23 | serhiy.storchaka | set | pull_requests: + <pull%5Frequest6731> |
2018-05-24 12:19:45 | miss-islington | set | pull_requests: + <pull%5Frequest6730> |
2018-05-24 12:19:37 | serhiy.storchaka | set | messages: + |
2018-05-24 08:05:16 | pitrou | set | priority: normal -> low |
2018-05-23 18:04:41 | serhiy.storchaka | set | title: Fix and improve errors handling in the garbage collector -> Fix errors handling in the garbage collector |
2018-05-23 17:50:39 | serhiy.storchaka | set | keywords: + patchstage: patch reviewpull_requests: + <pull%5Frequest6708> |
2018-05-23 17:41:20 | serhiy.storchaka | create |