Bug 612123 - rpmbuild rarely fails on s390x
Summary: rpmbuild rarely fails on s390x
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: rpm
Version: 6.0
Hardware: All
OS: Linux
Target Milestone: rc
: ---
Assignee: Panu Matilainen
QA Contact: BaseOS QE Security Team
Keywords: Regression
Depends On:
Blocks: 582655
TreeView+ depends on / blocked
Reported: 2010-07-07 11:12 UTC by Šimon Lukašík
Modified: 2011-03-15 14:01 UTC (History)
1 user (show)

Clone Of:
Last Closed: 2010-07-19 05:48:48 UTC

Attachments (Terms of Use)
full rpmbuild log (2.00 KB, text/plain)
2010-07-07 11:12 UTC, Šimon Lukašík
no flags Details
Spec file (663 bytes, text/plain)
2010-07-07 11:13 UTC, Šimon Lukašík
no flags Details
rpmbuild.log (1.84 KB, text/plain)
2010-07-12 21:05 UTC, Šimon Lukašík
no flags Details

Description Šimon Lukašík 2010-07-07 11:12:09 UTC
Created attachment 430033 [details]
full rpmbuild log

Description of problem:
When building 1000 same rpms, which differs only by name. Time-to-time one of build fails. It happens only on rhel6 s390x.

Version-Release number of selected component (if applicable):

How reproducible:
very rare

Steps to Reproduce:
1. run /CoreOS/rpm/Sanity/long-list several times
Actual results:
+ /usr/lib/rpm/redhat/brp-python-hardlink
/var/tmp/rpm-tmp.I8yduU: line 46: 22815 Segmentation fault      (core dumped) /usr/lib/rpm/redhat/brp-python-hardlink
error: Bad exit status from /var/tmp/rpm-tmp.I8yduU (%install)

RPM build errors:
    Bad exit status from /var/tmp/rpm-tmp.I8yduU (%install)
Executing(%prep): /bin/sh -e /var/tmp/rpm-tmp.nMsLWY

Expected results:

Additional info:

Comment 1 Šimon Lukašík 2010-07-07 11:13:06 UTC
Created attachment 430035 [details]
Spec file

Comment 5 Panu Matilainen 2010-07-07 13:51:15 UTC
It's not rpm that's crashing but something inside a script which rpm invokes during %install, based on the output that'd be inside brp-python-hardlink.

There's not a whole lot going on in brp-python-hardlink: it's a /bin/sh script which calls "find", comparing the results with "sha1sum" and calling "ln" on matches. One of those commands is segfaulting. Try adding 'ulimit -c unlimited' to the beginning of %install and see if you can catch a core dump or two, that'd tell us what exactly is crashing.

Comment 6 Šimon Lukašík 2010-07-12 21:05:10 UTC
Created attachment 431272 [details]

This is what I get when 'ulimit -c unlimited' added to %install section.

This log comes from task: https://beaker.engineering.redhat.com/recipes/12453#task147136

Comment 7 Šimon Lukašík 2010-07-15 01:14:45 UTC
Please note, that I cannot reproduce it on RHEL6.0-Snapshot-7.

There are beaker recipes, each with 32 runs of original reproducer. None of them caught traceback.


Comment 8 Šimon Lukašík 2010-07-15 01:17:37 UTC
s/traceback/seg fault/

Comment 9 Šimon Lukašík 2010-07-18 17:52:15 UTC
(In reply to comment #7)
> Please note, that I cannot reproduce it on RHEL6.0-Snapshot-7.
Problem was reported against Snap-6.

Comment 10 Jindrich Novy 2010-07-19 05:48:48 UTC
Closing as it is no more reproducible since snap 7.

Note You need to log in before you can comment on or make changes to this bug.