Message 287286 - Python tracker

➜

This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author	methane
Recipients	audric, methane, serhiy.storchaka, vstinner, xiang.zhang
Date	2017-02-08.08:55:07
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1486544108.07.0.017956021725.issue29438@psf.upfronthosting.co.za>
In-reply-to

Content
> Why res == 0 is added? If PyDict_SetItem() triggers recursive calling of _PyObjectDict_SetItem() which calls PyDict_SetItem() it may be possible that the first PyDict_SetItem() is failed while the dict is changed by the second PyDict_SetItem() and CACHED_KEYS(tp) becomes outdated. To avoid hiding error raised in PyDict_SetItem(). But it seems I was too nervous. The error will be hidden only when make_keys_shared() raise exception. I'll remove the check. BTW, how about -py35.patch? It is minimum patch to avoid "use after free". It skip CACHED_KEYS(tp) = NULL entirely. But I think I can apply same patch to Python 3.5 too.

> Why res == 0 is added? If PyDict_SetItem() triggers recursive calling of _PyObjectDict_SetItem() which calls PyDict_SetItem() it may be possible that the first PyDict_SetItem() is failed while the dict is changed by the second PyDict_SetItem() and CACHED_KEYS(tp) becomes outdated.

To avoid hiding error raised in PyDict_SetItem().
But it seems I was too nervous.  The error will be hidden only when make_keys_shared() raise exception.
I'll remove the check.

BTW, how about -py35.patch?  It is minimum patch to avoid "use after free".  It skip 
CACHED_KEYS(tp) = NULL entirely.  But I think I can apply same patch to Python 3.5 too.

History
Date	User	Action	Args
2017-02-08 08:55:08	methane	set	recipients: + methane, vstinner, serhiy.storchaka, xiang.zhang, audric
2017-02-08 08:55:08	methane	set	messageid: <1486544108.07.0.017956021725.issue29438@psf.upfronthosting.co.za>
2017-02-08 08:55:08	methane	link	issue29438 messages
2017-02-08 08:55:07	methane	create