#996 rpmdiff result for different count of hardlinks in rpm should fail
Closed: Fixed by tkopecek. Opened by jcupova.

Steps to Reproduce:

Run scratch build on arch x86_64 and i386 with this spec file:
Name: package-test
Version: 1
Release: 1%{?dist}
Summary: Summary
BuildArch: noarch

Group: a
License: a
URL: a

%description
desc

%prep

%build

%install
%{__mkdir_p} ${RPM_BUILD_ROOT}/bin
echo "echo ok" > ${RPM_BUILD_ROOT}/bin/test
%if %{__isa_bits} == 64
ln ${RPM_BUILD_ROOT}/bin/test ${RPM_BUILD_ROOT}/bin/test2
ln ${RPM_BUILD_ROOT}/bin/test ${RPM_BUILD_ROOT}/bin/test3
%else
ln ${RPM_BUILD_ROOT}/bin/test ${RPM_BUILD_ROOT}/bin/test2
cp ${RPM_BUILD_ROOT}/bin/test ${RPM_BUILD_ROOT}/bin/test3
%endif

%files
/bin/test
/bin/test2
/bin/test3

%changelog

Actual result:

BuildArch on x86_64 and i386 is closed and parent task build method is also closed.

Expected result:

BuildArch on x86_64 and i386 is closed but parent task build method should be failed and return result that count of hardlink is not same in rpm on x86_64 and i386.


Metadata Update from @tkopecek:
- Custom field Size adjusted to None
- Issue set to the milestone: 1.19

Metadata Update from @dgregor:
- Issue set to the milestone: 1.20 (was: 1.19)

Chatted with @tkopecek about this one... extending the existing rpmdiff tests for noarch package selection is a bit of a rabbit hole. Instead of changing the code, I would ask that we just document the existing behavior and leave the onus on package maintainers to address noarch package discrepancies that aren't covered by Koji.

PR #1875

Metadata Update from @tkopecek:
- Issue tagged with: doc

Metadata Update from @jcupova:
- Issue tagged with: no_qe

Commit 0e932095 fixes this issue

Commit f9553070 fixes this issue

This issue has been migrated to Fedora Forge:
https://forge.fedoraproject.org/koji/koji/issues/996

Please continue any further discussion there.

Metadata
Related Pull Requests