When shutting down the regression test postmaster after 'make check',

tglsfdc · tglsfdc · commit 5c7d04d24b96 · 2003-07-30T17:08:47.000Z
wait for the postmaster to actually exit.  Otherwise running repeated
'make check's tends to misbehave, because we try to remove and recreate
the data directory while the old PM is still alive.
diff --git a/src/test/regress/pg_regress.sh b/src/test/regress/pg_regress.sh
@@ -1,5 +1,5 @@
 #! /bin/sh
-# $Header: /cvsroot/pgsql/src/test/regress/Attic/pg_regress.sh,v 1.32 2003/07/29 00:03:19 tgl Exp $
+# $Header: /cvsroot/pgsql/src/test/regress/Attic/pg_regress.sh,v 1.33 2003/07/30 17:08:47 tgl Exp $
 
 me=`basename $0`
 : ${TMPDIR=/tmp}
@@ -652,6 +652,7 @@ done | tee "$result_summary_file" 2>&1
 if [ -n "$postmaster_pid" ]; then
     message "shutting down postmaster"
     kill -15 "$postmaster_pid"
+    wait "$postmaster_pid"
     unset postmaster_pid
 fi
 
