Re: Allow CI to only run the compiler warnings task

From: "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>
To: Rustam ALLAKOV <rustamallakov(at)gmail(dot)com>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org, Bertrand Drouvot <bertranddrouvot(dot)pg(at)gmail(dot)com>
Subject: Re: Allow CI to only run the compiler warnings task
Date: 2025-06-04 18:08:17
Message-ID: CAKFQuwa6BPoW3Cg6r=JGFxwCp_TGYBoou8KTu1+5hyNTp8Be2g@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, May 21, 2025 at 10:32 AM Rustam ALLAKOV <rustamallakov(at)gmail(dot)com>
wrote:

> The following review has been posted through the commitfest application:
> make installcheck-world: not tested
> Implements feature: not tested
> Spec compliant: tested, failed
> Documentation: tested, failed
>
> Hi Hackers,
>
> it might be better to list tasks in alphabetical order, and if I
> understand correctly
> initially they were alphabetically sorted.
>
> > ci-task-only:
> {(compilerwarnings|freebsd|linux|macos|mingw|netbsd|openbsd|windows)}

Appending the option and matching the yml file is proper for this patch
given its goal.

I do feel like there is something a bit off. Specifically:

I'd keep ci-os-only as a deprecated alias for a while - let it accept the
new compilerwarnings task name. I'd rather not break people's muscle
memory and the fact they are unlikely to realize this change until they run
one through and see everything running despite their directive. The "mess"
is quite minimal for the benefit.
(this is the only one I'd say is a - soft - patch blocker for me)

I'd like to communicate the always-runs versus the manual runs here in some
way.
Keeping the flavors together would seem to have merit.
I feel like the yml file itself should put these in a comment block.

> also it may increase readability if you sort tasks alphabetically within
> the .cirrust.tasks.yml
>
>
That should be done as a dedicated refactoring patch. I wouldn't consider
it high priority though. I'd love to find something that makes my outliner
display the task names as maximum collapse though (VS Code).

I'm apt to propose a refactoring patch here at some point to address these
things - but this patch to extend the yml file in an obvious way need not
take on such responsibilities.

David J.

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tomas Vondra 2025-06-04 18:15:58 Re: [WIP]Vertical Clustered Index (columnar store extension) - take2
Previous Message Florents Tselai 2025-06-04 18:03:03 Re: like pg_shmem_allocations, but fine-grained for DSM registry ?