Re: [patch] Adding EXTRA_REGRESS_OPTS to all pg_regress invocations - Mailing list pgsql-hackers

From Christoph Berg
Subject Re: [patch] Adding EXTRA_REGRESS_OPTS to all pg_regress invocations
Date
Msg-id 20140327170324.GI12703@msgid.df7cb.de
Whole thread Raw
In response to Re: [patch] Adding EXTRA_REGRESS_OPTS to all pg_regress invocations  (Bruce Momjian <bruce@momjian.us>)
Responses Re: [patch] Adding EXTRA_REGRESS_OPTS to all pg_regress invocations
List pgsql-hackers
Re: Bruce Momjian 2013-12-04 <20131204151533.GB17114@momjian.us>
> On Mon, May  6, 2013 at 11:51:47PM -0700, Christoph Berg wrote:
> > "make check" supports EXTRA_REGRESS_OPTS to pass extra options to
> > pg_regress, but all the other places where pg_regress is used do not
> > allow this. The attached patch adds EXTRA_REGRESS_OPTS to
> > Makefile.global.in (for contrib modules) and two more special
> > Makefiles (isolation and pg_upgrade).
> > 
> > The use case here is that Debian needs to be able to redirect the unix
> > socket directory used to /tmp, because /var/run/postgresql isn't
> > writable for the buildd user. The matching part for this inside
> > pg_regress is still in discussion here, but the addition of
> > EXTRA_REGRESS_OPTS is an independent step that is also useful for
> > others, so I'd like to propose it for inclusion.
> 
> Thanks, patch applied.  This will appear in PG 9.4.  I suppose we could
> backpatch this but I would need community feedback on that.

Thanks for pushing this. In the meantime, a new bit has appeared:
The new contrib/test_decoding checks make use of the
pg_isolation_regress_check macros (which the isolation test itself
doesn't). These macros also need EXTRA_REGRESS_OPTS, on top of
86ef4796f5120c55d1a48cfab52e51df8ed271b5:

diff --git a/src/Makefile.global.in b/src/Makefile.global.in
new file mode 100644
index cdddf49..8d08d19
*** a/src/Makefile.global.in
--- b/src/Makefile.global.in
*************** pg_regress_installcheck = $(top_builddir
*** 468,475 ****  pg_regress_clean_files = results/ regression.diffs regression.out tmp_check/ log/ 
! pg_isolation_regress_check = $(top_builddir)/src/test/isolation/pg_isolation_regress --inputdir=$(srcdir)
--temp-install=./tmp_check--top-builddir=$(top_builddir) $(pg_regress_locale_flags)
 
! pg_isolation_regress_installcheck = $(top_builddir)/src/test/isolation/pg_isolation_regress --inputdir=$(srcdir)
--top-builddir=$(top_builddir)$(pg_regress_locale_flags)
###########################################################################
 
--- 468,475 ----  pg_regress_clean_files = results/ regression.diffs regression.out tmp_check/ log/ 
! pg_isolation_regress_check = $(top_builddir)/src/test/isolation/pg_isolation_regress --inputdir=$(srcdir)
--temp-install=./tmp_check--top-builddir=$(top_builddir) $(pg_regress_locale_flags) $(EXTRA_REGRESS_OPTS)
 
! pg_isolation_regress_installcheck = $(top_builddir)/src/test/isolation/pg_isolation_regress --inputdir=$(srcdir)
--top-builddir=$(top_builddir)$(pg_regress_locale_flags) $(EXTRA_REGRESS_OPTS)
###########################################################################
 


Christoph
-- 
cb@df7cb.de | http://www.df7cb.de/



pgsql-hackers by date:

Previous
From: Pavel Stehule
Date:
Subject: Re: four minor proposals for 9.5
Next
From: Tom Lane
Date:
Subject: Re: Useless "Replica Identity: NOTHING" noise from psql \d