This partly reverts e6acde1a9 commit where is no clear
justification why the change was actually done. Because the
pg_config.h was not ready for ppc64p7 and aarch64 -- it was clear
breakage. In 1242873 discussion, it appeared that ppc64p7 needs
special handling as it is compatible/equivalent with ppc64.
Resolves: rhbz#1242873
Version: 9.4.4-3
.. to help users who face issues like rhbz#1118392.
Also, testsuite consists of several 'make check' runs. From now,
continue with checking if some of 'make check' runs failed; just
to be sure we have complete test results. Make the test failed
once all testing is done.
Resolves: #1118392
Version: 9.3.5-8
.. together with other scripts and files. This should make the
development and testing of those files easier. Also packaging
should benefit from this change.
Version: 9.3.5-6
Resolves: #1145129
The %{name}.macros file is not loaded by rpm. Spotted by jmlich
through regular dependency checker in Fedora & postgresql-ip4r
package.
Version: 9.3.5-4
Each package having some *.so plugin for postgresql-server should
be rebuilt against each major version bump of PostgreSQL. Using
this statements in F21+
BuildRequires: postgresql-devel
Require: postgresql-server(:MODULE_COMPAT_%{postgresql_major})
then automatically blocks installation of postgresql-server and
incompatible plugin at the same time and thus indirectly forces
plugin maintainers to rebuild their packages against newer server.
File macros.postgresql now starts providing correct
%{postgresql_major} macro.
Resolves: #1008939
Version: 9.3.4-8
.. by setting $PATH to default value. Also check for command
existence by 'command' built-in instead of checking for file
existence.
Version: 9.3.4-2
Resolves: #1097317
This generated the manual page badly before as the output of
--version option depended on system's postgresql.service (not
available at a build time). So handle options rather at the
beginning of the script.
Version: 9.3.2-5
Generate postgresql-setup and postgresql-check-db-dir in %prep
rather then in %install. It is (a) better place and (b) it speeds
up its testing ('rpmbuild -bp *.spec' is able to generate them).
Version: 9.3.2-4
Users may now specify by this variable parameters passed to initdb
run. See `man initdb(1)`. This may help if user gets stacked
with bug #1007802.
Related: #1007802
Version: 9.3.2-4
