Re: 9.6 TAP tests and extensions - Mailing list pgsql-hackers

From Craig Ringer
Subject Re: 9.6 TAP tests and extensions
Date
Msg-id CAMsr+YFkYm3+2hqQvGY-=TWCBHSQL9FGALv6sehTY7SEpK7d_Q@mail.gmail.com
Whole thread Raw
In response to Re: 9.6 TAP tests and extensions  (Craig Ringer <craig@2ndquadrant.com>)
Responses Re: 9.6 TAP tests and extensions
Re: 9.6 TAP tests and extensions
List pgsql-hackers
On 13 September 2016 at 13:27, Craig Ringer <craig@2ndquadrant.com> wrote:
> This was wrong for out-of-tree builds, updated.
>
> Still pending fix for PG_REGRESS path when invoked using
> $(prove_check) from PGXS

Looking further at this, I think a pgxs-specific patch to add support
for prove tests and isolation tests would be best, and can be done
separately. Possibly even snuck into a point release, but if not, at
least extension authors can invoke prove in their own Makefile if the
required modules get installed. It just needs an adaptation of the
command used in the $(prove_check) definition.

Extension makefiles run tests by listing the tests in REGRESS .
Something similar would make sense for isolation checks. For prove,
probably just a macro that can be invoked to enable prove tests in
pgxs makefiles.

I suggest that the above patches be applied to 9.6 and v10. Then for
v10 I'll look at enhancing PGXS to make it easier to use isolation
tests and prove tests; extensions that want to use them in 9.6 can
just add something like:


prove_check:       rm -rf $(CURDIR)/tmp_check/log       cd $(srcdir) && TESTDIR='$(CURDIR)' PATH="$(shell pg_config
--bindir):$$PATH" PGPORT='6$(DEF_PGPORT)'
top_builddir='$(CURDIR)/$(top_builddir)' PG_REGRESS='pg_regress'
$(PROVE) $(PG_PROVE_FLAGS) $(PROVE_FLAGS) t/*.pl

.PHONY: prove_check



to their Makefile , so it's not necessary to have PGXS support for
this for it to be useful in 9.6.

-- Craig Ringer                   http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services



pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: WAL consistency check facility
Next
From: Michael Paquier
Date:
Subject: Re: WAL consistency check facility