Message353521
| Author | nascheme |
|---|---|
| Recipients | Mark.Shannon, christian.heimes, jdemeyer, lukasz.langa, methane, miss-islington, nascheme, pablogsal, petr.viktorin, pitrou, tim.peters, vstinner |
| Date | 2019-09-30.01:06:31 |
| SpamBayes Score | -1.0 |
| Marked as misclassified | Yes |
| Message-id | <1569805591.48.0.0147991400448.issue38006@roundup.psfhosted.org> |
| In-reply-to |
| Content | |
|---|---|
Since W is in the unreachable set, we should not be executing its callback. Would the attached rough patch (gc_disable_wr_callback.txt) be a possible fix? When we find W inside handle_weakrefs(), we mark it as trash and will not execute the callback. This is similar to Pablo's bpo-38009 but I think attacks the problem in a better way. Having func_clear() being run is only one possible bad outcome of executing the callback. We want to avoid executing *any* user level Python code if the weakref has access to objects found by the GC and which have tp_clear called on them. |
|
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2019-09-30 01:06:31 | nascheme | set | recipients: + nascheme, tim.peters, pitrou, vstinner, christian.heimes, petr.viktorin, methane, lukasz.langa, Mark.Shannon, jdemeyer, pablogsal, miss-islington |
| 2019-09-30 01:06:31 | nascheme | set | messageid: <1569805591.48.0.0147991400448.issue38006@roundup.psfhosted.org> |
| 2019-09-30 01:06:31 | nascheme | link | issue38006 messages |
| 2019-09-30 01:06:31 | nascheme | create | |