Bug 196229 - lang/python27: Build (ctypes) fails on i386 with LIBFFI option off
Summary: lang/python27: Build (ctypes) fails on i386 with LIBFFI option off
Status: Closed Overcome By Events
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: i386 Any
: --- Affects Only Me
Assignee: freebsd-python (Nobody)
URL: http://bugs.python.org/issue22521
Keywords: needs-patch, needs-qa
Depends on:
Blocks:
 
Reported: 2014-12-23 15:57 UTC by Adam Weinberger
Modified: 2018-05-21 08:59 UTC (History)
2 users (show)

See Also:
koobs: maintainer-feedback+


Attachments
showing just the FFI failure (2.43 KB, text/plain)
2014-12-23 15:57 UTC, Adam Weinberger
no flags Details
showing the full build with LIBFFI off (528.19 KB, text/plain)
2014-12-23 15:58 UTC, Adam Weinberger
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Adam Weinberger freebsd_committer freebsd_triage 2014-12-23 15:57:30 UTC
Created attachment 150912 [details]
showing just the FFI failure

If the LIBFFI option is off (i.e., use python's libffi) the _ctypes.so module gets a build failure. The port still builds, but then it fails during packaging.

I'm attaching a log of just the failure and a full build log.

I'm making an assumption and assigning the PR to you, but please forgive me if I should have assigned it to python@ instead.
Comment 1 Bugzilla Automation freebsd_committer freebsd_triage 2014-12-23 15:57:30 UTC
Auto-assigned to maintainer python@FreeBSD.org
Comment 2 Adam Weinberger freebsd_committer freebsd_triage 2014-12-23 15:58:12 UTC
Created attachment 150913 [details]
showing the full build with LIBFFI off
Comment 3 Adam Weinberger freebsd_committer freebsd_triage 2014-12-23 15:58:51 UTC
Oh hey it assigned it to python@ anyway. Bugzilla is smarter than me!
Comment 4 Marcus von Appen freebsd_committer freebsd_triage 2014-12-23 17:04:45 UTC
Looks like some of the backports from 3.4 will cause 2.7 to break now, too. Bug 194684 and http://bugs.python.org/issue22521 contain those information. Unfortunately there is no fix for that issue as of now.

The current recommendation is to keep LIBFFI enabled on the i386 arch. If there are no news within the next few days, we will add a BROKEN message for i386, if LIBFFI is disabled.
Comment 5 Walter Schwarzenfeld 2018-01-10 21:20:08 UTC
The Broken statement is in the Makefile. The PR on python.org is closed with broken.
I think it is nothing to do, and this can be closed.
Comment 6 Adam Weinberger freebsd_committer freebsd_triage 2018-05-18 15:44:50 UTC
(In reply to w.schwarzenfeld from comment #5)

I agree. Closing this out.