Bug 985582
Summary: | FTBFS: self check failures | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Karsten Hopp <karsten> |
Component: | subversion | Assignee: | Joe Orton <jorton> |
Status: | CLOSED RAWHIDE | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
Severity: | high | Docs Contact: | |
Priority: | medium | ||
Version: | 20 | CC: | andreas.stieger, jorton, karsten, vanmeeuwen+fedora |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | powerpc | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2013-12-02 13:31:35 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: |
Description
Karsten Hopp
2013-07-17 19:44:13 UTC
Is that reproducible, does it fail the same way every time? Yes, here's another attempt: http://ppc.koji.fedoraproject.org/koji/taskinfo?taskID=1238790 If there's no quick fix, would you mind disabling those tests on ppc and %{power64} until you've got a fix ? According to my build logs not having subversion-1.8.0 blocks at least 900 packages from being built on PPC secondary arch and any temporary fix/workaround would be welcome. There is a test case which fails if running as root. Are you running builds as root on ppc*? (Surely a bad idea?) I've pushed a commit/build which ignores test suite failures on ppc & ppc64 (hopefully correctly). $ make check TESTS=subversion/tests/cmdline/wc_tests.py [1/1] wc_tests.py..............................................................................................................success The tests in question pass if run as a non-root user. (kernel 3.9.5-301.fc19.ppc64p7) This bug appears to have been reported against 'rawhide' during the Fedora 20 development cycle. Changing version to '20'. More information and reason for this action is here: https://fedoraproject.org/wiki/BugZappers/HouseKeeping/Fedora20 As a friendly service from your fellow Subversion packagers over at openSUSE... We worked with upstream to fix this: https://svn.apache.org/viewvc?view=revision&revision=1543145 ------------------------------------------------------------------------ r1543145 | rhuijben | 2013-11-18 20:43:23 +0000 (Mon, 18 Nov 2013) | 9 lines Changed paths: M /subversion/trunk/subversion/svn/status-cmd.c Resolve a strange error that somehow occurred only on some specific powerpc architectures. The real problem is that we didn't properly detect the end of the variable argument list, so it can fail on all architectures. Found by: astieger * subversion/svn/status-cmd.c (svn_cl__status): Add missing end of warning status values marker. Andreas, thanks a lot, nice find! (In reply to Joe Orton from comment #8) > Andreas, thanks a lot, nice find! And while you are at it, your current build also fails due to dynamically linking against SQLite 3.8. Link statically against the recommended 3.7.15.1 or fetch fixes from trunkL https://svn.apache.org/viewvc?view=revision&revision=1542774 (configure.ac) SQLITE_RECOMMENDED_VER="3.7.15.1" Please note that svn with SQLite 3.8 has some performance issues. Now I owe you two beers, Andreas! Thanks so much for that, yes, it was breaking in that test. In fact these three trunk revisions depend on each other: r1542774, r1542767, r1542765 drink_of_choice_owed++ ... and finally a working svn build again - thanks Andreas :) I've put the fix for this in f20 & rawhide builds. |