Author vstinner
Recipients josh.r, llllllllll, martin.panter, python-dev, serhiy.storchaka, vstinner, yselivanov
Date 2016-04-19.21:28:36
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1461101316.66.0.532562367455.issue26802@psf.upfronthosting.co.za>
In-reply-to
Content
> FYI, there is a proposition about constructing arguments tuple and dict in bytecode instead of ceval.c. This will significantly simplify CALL_FUNCTION (which will get just one optional tuple and one optional dict). Likely this idea will be implemented after changing to wordcode.

Wordcode is the issue #26647.

Do you have a reference to the more efficient implementation of CALL_FUNCTION? I recall vaguely this idea.

--

It was also proposed to pass keyword arguments as positional arguments: replace func(a=1, b=2) with func(1, 2) with "def func(a, b): ...". But this optimization requires something like FAT Python to disable the optimization if the function is replaced at runtime.

This optimization is more complex, maybe it's not worth.
History
Date User Action Args
2016-04-19 21:28:36vstinnersetrecipients: + vstinner, python-dev, martin.panter, serhiy.storchaka, yselivanov, josh.r, llllllllll
2016-04-19 21:28:36vstinnersetmessageid: <1461101316.66.0.532562367455.issue26802@psf.upfronthosting.co.za>
2016-04-19 21:28:36vstinnerlinkissue26802 messages
2016-04-19 21:28:36vstinnercreate