Message 364429 - Python tracker

Message364429

Author	Mark.Shannon
Recipients	Mark.Shannon, eric.snow, jeremy.kloth, jkloth, larry, maciej.szulik, nanjekyejoannah, ncoghlan, phsilva, rhettinger, shihai1991, vstinner
Date	2020-03-17.14:10:55
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1584454255.51.0.514415647375.issue39511@roundup.psfhosted.org>
In-reply-to

Content
Consider the case where a thread that doesn't hold the GIL attempts to get a reference on `None`. The problem with having a single immortal `None`, is that it will cause data cache thrashing as two different CPUs modify the refcount on the shared `None` object. Each subinterpreter needs its own distinct `None`. `None` could be made immortal, it just can't be shared between sub-interpreters.

Content

Consider the case where a thread that doesn't hold the GIL attempts to get a reference on `None`.

The problem with having a single immortal `None`, is that it will cause data cache thrashing as two different CPUs modify the refcount on the shared `None` object.
Each subinterpreter needs its own distinct `None`.

`None` could be made immortal, it just can't be shared between sub-interpreters.

History
Date	User	Action	Args
2020-03-17 14:10:55	Mark.Shannon	set	recipients: + Mark.Shannon, rhettinger, ncoghlan, vstinner, larry, jkloth, phsilva, jeremy.kloth, eric.snow, maciej.szulik, nanjekyejoannah, shihai1991
2020-03-17 14:10:55	Mark.Shannon	set	messageid: <1584454255.51.0.514415647375.issue39511@roundup.psfhosted.org>
2020-03-17 14:10:55	Mark.Shannon	link	issue39511 messages
2020-03-17 14:10:55	Mark.Shannon	create