[issue36696] possible multiple regressions on AIX

Change by STINNER Victor : -- components: +Build resolution: -> fixed stage: patch review -> resolved status: open -> closed versions: +Python 3.8 ___ Python tracker ___ _

[issue36696] possible multiple regressions on AIX

STINNER Victor added the comment: New changeset 574913479f26b5ff48827861bce68281be01d16e by Victor Stinner in branch 'master': bpo-36635, bpo-36696: Fix setup.py on AIX (GH-12922) https://github.com/python/cpython/commit/574913479f26b5ff48827861bce68281be01d16e -- _

[issue36696] possible multiple regressions on AIX

STINNER Victor added the comment: > Should I open PR - or is this better fixed in the original PR? What keeps > things cleaner? I wrote PR 12922. I really hate -Ddefine syntax, but it seems like we have to use it just for AIX :-( I tried to use "-D Py_BUILD_CORE_..." in Modules/Setup, but

[issue36696] possible multiple regressions on AIX

Michael Felt added the comment: On 22/04/2019 14:15, Inada Naoki wrote: > Inada Naoki added the comment: > > Maybe, XLC doesn't support -D name. -Dname should be used instead. Excellent hint: the diff between bot run 1013 and run 1014 reveals: diff --git a/setup.py b/setup.py index 9c83914fd

[issue36696] possible multiple regressions on AIX

Change by STINNER Victor : -- keywords: +patch pull_requests: +12851 stage: -> patch review ___ Python tracker ___ ___ Python-bugs-

[issue36696] possible multiple regressions on AIX

Inada Naoki added the comment: Maybe, XLC doesn't support -D name. -Dname should be used instead. -- nosy: +inada.naoki ___ Python tracker ___ ___

[issue36696] possible multiple regressions on AIX

New submission from Michael Felt : My AIX bot has been very consistent - only the multiprocessing tests failing when run by bot, but 4 or 5 days ago 3 to 5 additional tests - that, afaik, had never failed before, are now failing. These may also be compiler related specifics, or the presence (