Add mode where contrib installcheck runs each module in a separately named database.

Normally each module is tested in aq database named contrib_regression,
which is dropped and recreated at the beginhning of each pg_regress run.
This mode, enabled by adding USE_MODULE_DB=1 to the make command line,
runs most modules in a database with the module name embedded in it.

This will make testing pg_upgrade on clusters with the contrib modules
a lot easier.

Still to be done: adapt to the MSVC build system.

Backpatch to 9.0, which is the earliest version it is reasonably
possible to test upgrading from.

contrib/dblink/Makefile

@@ -10,6 +10,9 @@ DATA = uninstall_dblink.sql
 REGRESS = dblink
 
 
+# the db name is hard-coded in the tests
+override undefine USE_MODULE_DB
+
 ifdef USE_PGXS
 PG_CONFIG = pg_config
 PGXS := $(shell $(PG_CONFIG) --pgxs)

src/Makefile.global.in

@@ -284,6 +284,15 @@ BZIP2	= bzip2
 
 PL_TESTDB = pl_regression
 CONTRIB_TESTDB = contrib_regression
+ifneq ($(MODULE_big),)
+  CONTRIB_TESTDB_MODULE = contrib_regression_$(MODULE_big)
+else 
+  ifneq ($(MODULES),)
+    CONTRIB_TESTDB_MODULE = contrib_regression_$(MODULES)
+  else
+    CONTRIB_TESTDB_MODULE = contrib_regression
+  endif
+endif
 
 # Installation.
 

src/makefiles/pgxs.mk

@@ -233,7 +233,11 @@ ifdef REGRESS
 
 # Calling makefile can set REGRESS_OPTS, but this is the default:
 ifndef REGRESS_OPTS
-REGRESS_OPTS = --dbname=$(CONTRIB_TESTDB)
+ ifdef USE_MODULE_DB
+   REGRESS_OPTS = --dbname=$(CONTRIB_TESTDB_MODULE)
+ else
+   REGRESS_OPTS = --dbname=$(CONTRIB_TESTDB)
+ endif
 endif
 
 # where to find psql for running the tests