Обсуждение: Re: pgsql: Clean up after TAP tests in oid2name and vacuumlo.
On 2018-Sep-04, Michael Paquier wrote:
> OK. I have dug into that, and finished with the attached. What do you
> think? One thing is that the definition of submake is moving out of
> REGRESS, and .PHONY gets defined.
Should this be used in src/test/modules/{brin,test_commit_ts} also?
Why do you guys design Makefile rules in pgsql-committers?
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
On Wed, Sep 05, 2018 at 11:39:50AM -0300, Alvaro Herrera wrote:
> Should this be used in src/test/modules/{brin,test_commit_ts} also?
Hmm, I have not thought those two ones. commit_ts relies on REGRESS to
be defined so as it does its cleanup. brin is more interesting, it
directly quotes that it needs to tweak its Makefile to do the cleanup,
and it uses isolation tests. Wouldn't it be more adapted to add a
second extra switch for isolation tests similar to REGRESS? We could
call it ISOLATION, and it would list the set of isolation tests you
could run from an extension. That would be useful for a lot of folks.
Thoughts? It would be better to start a new thread rather than posting
a new patch, but I'd rather take the temperature first.
> Why do you guys design Makefile rules in pgsql-committers?
That was a bad idea, and a reaction to what Tom has committed for the
cleanup of the new TAP tests I have added previously.
--
Michael
Вложения
Michael Paquier <michael@paquier.xyz> writes:
> On Wed, Sep 05, 2018 at 11:39:50AM -0300, Alvaro Herrera wrote:
>> Should this be used in src/test/modules/{brin,test_commit_ts} also?
> Hmm, I have not thought those two ones. commit_ts relies on REGRESS to
> be defined so as it does its cleanup. brin is more interesting, it
> directly quotes that it needs to tweak its Makefile to do the cleanup,
> and it uses isolation tests. Wouldn't it be more adapted to add a
> second extra switch for isolation tests similar to REGRESS? We could
> call it ISOLATION, and it would list the set of isolation tests you
> could run from an extension. That would be useful for a lot of folks.
> Thoughts? It would be better to start a new thread rather than posting
> a new patch, but I'd rather take the temperature first.
Yeah, if we're going to introduce infrastructure for TAP tests, it'd
make sense to do so for isolation tests as well.
regards, tom lane
On 2018-Sep-05, Michael Paquier wrote:
> On Wed, Sep 05, 2018 at 11:39:50AM -0300, Alvaro Herrera wrote:
> > Should this be used in src/test/modules/{brin,test_commit_ts} also?
>
> Hmm, I have not thought those two ones. commit_ts relies on REGRESS to
> be defined so as it does its cleanup. brin is more interesting, it
> directly quotes that it needs to tweak its Makefile to do the cleanup,
> and it uses isolation tests. Wouldn't it be more adapted to add a
> second extra switch for isolation tests similar to REGRESS? We could
> call it ISOLATION, and it would list the set of isolation tests you
> could run from an extension. That would be useful for a lot of folks.
>
> Thoughts?
Yeah, +1 for that.
> It would be better to start a new thread rather than posting
> a new patch, but I'd rather take the temperature first.
Slightly feverish, it appears.
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
On Wed, Sep 05, 2018 at 03:20:03PM -0300, Alvaro Herrera wrote: > On 2018-Sep-05, Michael Paquier wrote: > > It would be better to start a new thread rather than posting > > a new patch, but I'd rather take the temperature first. > > Slightly feverish, it appears. As far as I can see, the proposal is not cold to death and could have a future, so I'll begin a new thread with a more complete patch. -- Michael
Вложения
On Wed, Sep 05, 2018 at 01:05:42PM -0700, Michael Paquier wrote: > As far as I can see, the proposal is not cold to death and could have a > future, so I'll begin a new thread with a more complete patch. Done. Let's move the discussion here then: https://www.postgresql.org/message-id/20180906014849.GG2726@paquier.xyz -- Michael