Author ncoghlan
Recipients Rhamphoryncus, effbot, georg.brandl, jcea, ncoghlan, timehorse, vstinner
Date 2009-01-02.00:08:58
SpamBayes Score 0.000159847
Marked as misclassified No
Message-id <1230854939.23.0.581472186321.issue3299@psf.upfronthosting.co.za>
In-reply-to
Content
Ah, now that I go to implement it, I remember why I didn't like the idea
of doing anything directly in PyObject_Del. While the Python memory
allocator is primarily designed for allocation of Python objects, it
isn't actually *limited* to that. So there is no guarantee that the void
pointer passed in to PyObject_Del can be safely cast to a PyObject pointer.

The idea could still be implemented by scanning the list of active
objects to see if the passed in pointer refers to one of them, but that
would make object deallocation in pydebug builds extraordinarily slow.
History
Date User Action Args
2009-01-02 00:08:59ncoghlansetrecipients: + ncoghlan, effbot, georg.brandl, jcea, Rhamphoryncus, vstinner, timehorse
2009-01-02 00:08:59ncoghlansetmessageid: <1230854939.23.0.581472186321.issue3299@psf.upfronthosting.co.za>
2009-01-02 00:08:58ncoghlanlinkissue3299 messages
2009-01-02 00:08:58ncoghlancreate