On Tue, Dec 27, 2022 at 01:58:14AM -0500, Tom Lane wrote:
> Justin Pryzby <pryzby@telsasoft.com> writes:
> > FYI: this causes meson test running ("installcheck") fail when run
> > twice. I guess that's expected to work, per:
>
> We do indeed expect that to work ... but I don't see any problem
> with repeat "make installcheck" on HEAD. Can you provide more
> detail about what you're seeing?
This fails when run more than once:
time meson test --setup running --print test_pg_db_role_setting-running/regress
@@ -1,12 +1,13 @@
CREATE EXTENSION test_pg_db_role_setting;
CREATE USER regress_super_user SUPERUSER;
+ERROR: role "regress_super_user" already exists
CREATE USER regress_regular_user;
+ERROR: role "regress_regular_user" already exists
...
It didn't fail for you because it says:
./src/test/modules/test_pg_db_role_setting/Makefile
+# disable installcheck for now
+NO_INSTALLCHECK = 1
It also says:
+# and also for now force NO_LOCALE and UTF8
+ENCODING = UTF8
+NO_LOCALE = 1
which was evidently copied from the "oat" tests, which have said that
since March (5b29a9f77, 7c51b7f7c).
It fails the same way with "make" if you change it to not disable
installcheck:
EXTRA_REGRESS_OPTS="--bindir=`pwd`/tmp_install/usr/local/pgsql/bin" PGHOST=/tmp make installcheck -C
src/test/modules/test_pg_db_role_setting
--
Justin