-
Notifications
You must be signed in to change notification settings - Fork 763
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
use FFI calls for refcounting on all abi3 versions #4324
Conversation
Our benchmarking infra isn't particularly high fidelity, but this change looks like it's within the noise floor for us: https://github.com/pyca/cryptography/actions/runs/9856363207?pr=11232 |
👍 cc @mgorny are you willing to test one of the problematic packages with this branch? If so, I'm tempted to merge it for the simplification it brings. |
I'll try on rustworkx in a minute. |
And confirmed! Without the patch, |
Great! In which case if fellow PyO3 maintainers are happy, I propose to merge this and ship as 0.22.2 (maybe also with #4321 when that's done). |
Having heard no concerns, I think let's move ahead with this. |
* use FFI calls for refcounting on all abi3 versions * fix implementation on PyPy
* use FFI calls for refcounting on all abi3 versions * fix implementation on PyPy
* use FFI calls for refcounting on all abi3 versions * fix implementation on PyPy
As per #4311 (comment)
The idea is that we can mitigate issues around 3.12's immortal objects and wonky CPython assertions by backporting the stable ABI change to use FFI calls for refcounting from 3.12 all the way back to our supported version of 3.8.
cc @alex - you might be interested in this for the possible performance impact on cryptography (though this impact would be felt eventually anyway).