So apparently, %patch -F does not actually *work* in the current Fedora

infrastructure. Grumble.
This commit is contained in:
Tom Lane 2008-07-29 01:50:10 +00:00
parent 09bd3b3257
commit b2169aa2ae
2 changed files with 5 additions and 9 deletions

View File

@ -4,12 +4,12 @@ for PostgreSQL, but it's not very tenable in the Fedora/RHEL world.
Dike out the check.
diff -Naur postgresql-8.3RC2.orig/configure.in postgresql-8.3RC2/configure.in
--- postgresql-8.3RC2.orig/configure.in 2008-01-17 23:47:32.000000000 -0500
+++ postgresql-8.3RC2/configure.in 2008-01-18 11:39:58.000000000 -0500
diff -Naur postgresql-8.3.3.orig/configure.in postgresql-8.3.3/configure.in
--- postgresql-8.3.3.orig/configure.in 2008-06-08 20:38:40.000000000 -0400
+++ postgresql-8.3.3/configure.in 2008-07-28 21:45:08.000000000 -0400
@@ -19,10 +19,6 @@
AC_INIT([PostgreSQL], [8.3RC2], [pgsql-bugs@postgresql.org])
AC_INIT([PostgreSQL], [8.3.3], [pgsql-bugs@postgresql.org])
-m4_if(m4_defn([m4_PACKAGE_VERSION]), [2.59], [], [m4_fatal([Autoconf version 2.59 is required.
-Untested combinations of 'autoconf' and PostgreSQL versions are not

View File

@ -351,8 +351,7 @@ system, including regression tests and benchmarks.
%prep
%setup -q
%patch1 -p1
# intentionally allowing fuzz here because of adjacent version number
%patch2 -p1 -F2
%patch2 -p1
%patch3 -p1
%patch4 -p1
# patch5 is applied later
@ -862,9 +861,6 @@ rm -rf $RPM_BUILD_ROOT
%changelog
* Mon Jul 28 2008 Tom Lane <tgl@redhat.com> 8.3.3-3
- Fix build failure caused by new default patch fuzz = 0 policy in rawhide.
The only patch that actually broke was postgresql-ac-version.patch, and
for that it seems better to allow fuzz than to have a guaranteed patch
failure on every upstream minor version update.
* Fri Jun 20 2008 Tom Lane <tgl@redhat.com> 8.3.3-2
- Install Pgtcl in /usr/lib/tcl$TCL_VERSION, not directly in /usr/lib.