Message 402044 - Python tracker

Message402044

Author	vstinner
Recipients	Mark.Shannon, kj, malin, neonene, paul.moore, rhettinger, steve.dower, tim.golden, vstinner, zach.ware
Date	2021-09-17.12:11:52
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1631880712.12.0.248303840984.issue45116@roundup.psfhosted.org>
In-reply-to

Content
I added Py_ALWAYS_INLINE to run benchmarks more easily. Even if Py_INCREF() is converted back to a macro, there are now multiple static inline functions which are short and performance critical. Using Py_ALWAYS_INLINE may speed up the Python debug builds and the PGO build on Windows if I understood correctly. Right now, I'm not sure. The heuristic to decide if a function is inlined or not seems to depend a lot on the compiler and the compiler options.

Content

I added Py_ALWAYS_INLINE to run benchmarks more easily. Even if Py_INCREF() is converted back to a macro, there are now multiple static inline functions which are short and performance critical.

Using Py_ALWAYS_INLINE *may* speed up the Python debug builds and the PGO build on Windows if I understood correctly.

Right now, I'm not sure. The heuristic to decide if a function is inlined or not seems to depend a lot on the compiler and the compiler options.

History
Date	User	Action	Args
2021-09-17 12:11:52	vstinner	set	recipients: + vstinner, rhettinger, paul.moore, tim.golden, Mark.Shannon, zach.ware, steve.dower, malin, neonene, kj
2021-09-17 12:11:52	vstinner	set	messageid: <1631880712.12.0.248303840984.issue45116@roundup.psfhosted.org>
2021-09-17 12:11:52	vstinner	link	issue45116 messages
2021-09-17 12:11:52	vstinner	create