Author vstinner
Recipients Yury.Selivanov, casevh, josh.r, lemburg, mark.dickinson, pitrou, rhettinger, serhiy.storchaka, skrah, vstinner, yselivanov, zbyrne
Date 2016-02-06.01:31:55
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1454722317.71.0.39458019109.issue21955@psf.upfronthosting.co.za>
In-reply-to
Content
inline-2.patch: more complete version of inline.patch.

Optimize the same instructions than Python 2: BINARY_ADD, INPLACE_ADD, BINARY_SUBSTRACT, INPLACE_SUBSTRACT.


Quick & *dirty* microbenchmark:

$ ./python -m timeit -s 'x=1' 'x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x+x'

* Original: 287 ns
* fastint5_2.patch: 261 ns (-9%)
* inline-2.patch: 212 ns (-26%)


$ ./python -m timeit -s 'x=1000; y=1' 'x-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y-y'

* Original: 517 ns
* fastint5_2.patch: 469 ns (-9%)
* inline-2.patch: 442 ns (-15%)


Ok. Now I'm lost. We have so many patches :-) Which one do you prefer?

In term of speedup, I expect that Python 2 design (inline-2.patch) cannot be beaten in term of performance by another another option since it doesn't need any C code and does everything in ceval.c.
History
Date User Action Args
2016-02-06 01:31:58vstinnersetrecipients: + vstinner, lemburg, rhettinger, mark.dickinson, pitrou, casevh, skrah, Yury.Selivanov, serhiy.storchaka, yselivanov, josh.r, zbyrne
2016-02-06 01:31:57vstinnersetmessageid: <1454722317.71.0.39458019109.issue21955@psf.upfronthosting.co.za>
2016-02-06 01:31:57vstinnerlinkissue21955 messages
2016-02-06 01:31:56vstinnercreate