Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.

Bug 1078354

Summary: numpy do not build for ppc64le archi
Product: [Fedora] Fedora Reporter: Michel Normand <normand>
Component: numpyAssignee: Gwyn Ciesla <gwync>
Status: CLOSED CURRENTRELEASE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: medium Docs Contact:
Priority: unspecified    
Version: rawhideCC: gwync, jcapik, jspaleta, orion, rdieter, sochotni, tomspur, ttomecek
Target Milestone: ---Keywords: Patch
Target Release: ---   
Hardware: ppc64le   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-08-20 09:32:43 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1051573    
Attachments:
Description Flags
numpy_long_double_representation.patch
none
numpy-1.8.1.ppc64le.patch none

Description Michel Normand 2014-03-19 16:21:40 UTC
Created attachment 876412 [details]
numpy_long_double_representation.patch

Description of problem: numpy do not build for ppc64le archi


Version-Release number of selected component (if applicable):
numpy-1.8.0-5.fc21


Steps to Reproduce:
1. fedpkg clone -a numpy
3. fedpkg mockbuild

Actual results:
...
 Traceback (most recent call last):
      File "setup.py", line 192, in <module>
        setup_package()
    ...
      File "numpy/core/setup_common.py", line 305, in long_double_representation
        raise ValueError("Unrecognized format (%s)" % saw)
    ValueError: Unrecognized format (['001', '043', '105', '147', '211', '253', '315', '357', '000', '000', '000', '124', '064', '157', '235', '301', '000', '000', '000', '000', '000', '000', '000', '0


Additional info:
In attachement numpy_long_double_representation.patch (not yet upstream) is suggested code change to avoid build error.

Comment 1 Michel Normand 2014-03-19 18:21:44 UTC
still pending discussion upstream for the patch to be applied:
https://github.com/numpy/numpy/issues/4516

Comment 2 Michel Normand 2014-04-24 06:25:56 UTC
Created attachment 889192 [details]
numpy-1.8.1.ppc64le.patch

This patch is the backport in 1.8.1 of git commit available in 1.9.0 futur release:
2ba74c3 Adding support of ppc64le architecture

Comment 3 Rex Dieter 2014-05-07 14:24:43 UTC
Patch looks good to me offhand, I'd say it's worth committing a FTBFS fix sooner rather than later (and I'd be happy to help sort out any problems, if there are any).

Comment 4 Jaromír Cápík 2014-05-07 14:44:24 UTC
Thanks Rex ... backport patch pushed and works.