Message 64373 - 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	pitrou
Recipients	christian.heimes, gregory.p.smith, pitrou
Date	2008-03-23.19:05:32
SpamBayes Score	0.062273283
Marked as misclassified	No
Message-id	<1206299134.66.0.0612444192968.issue2013@psf.upfronthosting.co.za>
In-reply-to

Content
The problem with choosing a sensible freelist size is that we don't have any reference workloads. However, I just tested with 10000 and it doesn't seem to slow anything down anyway. It doesn't make our microbenchmarks I thought the patch to compact freelists at each full gc collection had been committed, but it doesn't seem there. Perhaps it will change matters quite a bit. On the one hand, it will allow for bigger freelists with less worries of degrading memory footprint (but still, potential cache pollution). On the other hand, the bigger the freelists, the more expensive it is to deallocate them.

The problem with choosing a sensible freelist size is that we don't have
any reference workloads. However, I just tested with 10000 and it
doesn't seem to slow anything down anyway. It doesn't make our
microbenchmarks

I thought the patch to compact freelists at each full gc collection had
been committed, but it doesn't seem there. Perhaps it will change
matters quite a bit. On the one hand, it will allow for bigger freelists
with less worries of degrading memory footprint (but still, potential
cache pollution). On the other hand, the bigger the freelists, the more
expensive it is to deallocate them.

History
Date	User	Action	Args
2008-03-23 19:05:35	pitrou	set	spambayes_score: 0.0622733 -> 0.062273283 recipients: + pitrou, gregory.p.smith, christian.heimes
2008-03-23 19:05:34	pitrou	set	spambayes_score: 0.0622733 -> 0.0622733 messageid: <1206299134.66.0.0612444192968.issue2013@psf.upfronthosting.co.za>
2008-03-23 19:05:33	pitrou	link	issue2013 messages
2008-03-23 19:05:32	pitrou	create