Author rhettinger
Recipients Nitapol, lukasz.langa, ned.deily, rhettinger, ronaldoussoren, vstinner
Date 2019-08-25.06:21:55
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1566714116.19.0.39066650446.issue36205@roundup.psfhosted.org>
In-reply-to
Content
Something in the release build process is triggering this behavior (perhaps PGO).

I observe the bug on the python.org official 3.8b3 release, official 3.7 release, but not in the official 3.6 release:

$ python3.8 ~/time_compare.py
Python ('v3.8.0b3:4336222407', 'Jul 29 2019 09:46:03') Clang 6.0 (clang-600.0.57)
CPU Time: 14.632004 Wall Clock: 7.318585415  Distance: 750
$ python3.7 ~/time_compare.py
Python ('v3.7.4:e09359112e', 'Jul  8 2019 14:54:52') Clang 6.0 (clang-600.0.57)
CPU Time: 16.405296 Wall Clock: 8.208222215000001  Distance: 750
$ python3.6 ~/time_compare.py
Python ('v3.6.8:3c6b436a57', 'Dec 24 2018 02:04:31') GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.57)
CPU Time: 8.645299 Wall Clock: 8.647321333999571  Distance: 7


However, when I do my own fresh build for 3.8b3, the problem disappears:

$ py ~/time_compare.py
Python ('tags/v3.8.0b3:4336222407', 'Aug 24 2019 23:08:57') Clang 10.0.1 (clang-1001.0.46.4)
CPU Time: 7.446087 Wall Clock: 7.447327639  Distance: 750

I build with the following process:

    make distclean
    ./configure --with-openssl=$(brew --prefix openssl)
    make
History
Date User Action Args
2019-08-25 06:21:56rhettingersetrecipients: + rhettinger, ronaldoussoren, vstinner, ned.deily, lukasz.langa, Nitapol
2019-08-25 06:21:56rhettingersetmessageid: <1566714116.19.0.39066650446.issue36205@roundup.psfhosted.org>
2019-08-25 06:21:56rhettingerlinkissue36205 messages
2019-08-25 06:21:55rhettingercreate