[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: gitlab-ci check-dco test

From: Daniel P . Berrangé
Subject: Re: gitlab-ci check-dco test
Date: Mon, 12 Apr 2021 10:05:39 +0100
User-agent: Mutt/2.0.5 (2021-01-21)

On Fri, Apr 09, 2021 at 07:58:48PM -0700, Richard Henderson wrote:
> On development branches, it's not uncommon to push
> temporary --fixup patches, and normally one doesn't
> sign those.  But then of course one get hate-mail
> from the gitlab-ci job about the failing test.
> Is there a way to make it fatal on staging, but
> merely a warning on other branches (a-la checkpatch)?

Note,  checkpatch is *never* fatal on any branch because there are always
scenarios in which checkpatch gives false positives that we have to allow.

We can of course set 'allow_failure: true' on the DCO check, on non-staging
branches, but that loose some of its value. In general I think users should
see this as a mandatory check, because we don't want them ever sending
patches without this passing. Your scenario of sometimes needing to push
temporary fix patches is valid too of course.

So this is a no-win scenario and we have to decide what the least worst
option is. When I added this check I decided the least worst was to have
developers see failures when they have temp fixup patches, because I was
optimizing for the ensuring developers see & fix problems before they
submit to qemu-devel.

In libvirt we have the same check, but we moved it to a separate stage
in the pipeline "sanity_checks" instead of "build", so developers can
see at a glance in the UI that the build jobs all passed, and only the
sanity check(s) failed.

|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|

reply via email to

[Prev in Thread] Current Thread [Next in Thread]