pgsql: Propagate trigger arguments to partitions

Hi Alvaro,

On Tue, Jul 09, 2019 at 09:23:45PM +0000, Alvaro Herrera wrote:

Propagate trigger arguments to partitions

We were creating the cloned triggers with an empty list of arguments,
losing the ones that had been specified by the user when creating the
trigger in the partitioned table. Repair.

This was forgotten in commit 86f575948c77.

This commit has broken the buildfarm on REL_11_STABLE for a couple of
animals:
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=prion&dt=2019-07-10%2006%3A27%3A02

/home/ec2-user/bf/root/REL_11_STABLE/pgsql.build/../pgsql/src/backend/commands/tablecmds.c:
In function ‘CloneRowTriggersToPartition’:
/home/ec2-user/bf/root/REL_11_STABLE/pgsql.build/../pgsql/src/backend/commands/tablecmds.c:15283:4:
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < trigForm->tgnargs; i++)

This gets fixed with the attached. I would personally keep the code
consistent for v12 and HEAD for this code path, but I am fine to let
you decide the way you prefer.
--
Michael

On 2019-Jul-10, Michael Paquier wrote:

Hello

This commit has broken the buildfarm on REL_11_STABLE for a couple of
animals:
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=prion&amp;dt=2019-07-10%2006%3A27%3A02

Argh. You say "a couple", you mean "half the buildfarm for that branch"
:-(

error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < trigForm->tgnargs; i++)

This gets fixed with the attached. I would personally keep the code
consistent for v12 and HEAD for this code path, but I am fine to let
you decide the way you prefer.

I kinda like block-scoped variables, braindead though they are in
C, so I'll just patch pg11. It's not like this code changes too much
anyway (famous last words).

Thanks for pinging

--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

On Wed, Jul 10, 2019 at 08:53:36AM -0400, Alvaro Herrera wrote:

I kinda like block-scoped variables, braindead though they are in
C, so I'll just patch pg11. It's not like this code changes too much
anyway (famous last words).

Thanks! That's fine by me.
--
Michael