[Buildroot] [PATCH v4 3/3] .gitlab-ci.yml: add trigger per job
Arnout Vandecappelle
arnout at mind.be
Mon Apr 22 07:19:40 UTC 2019
On 22/04/2019 03:27, Ricardo Martincoski wrote:
> Hello,
>
> On Sat, Apr 13, 2019 at 10:46 AM, Arnout Vandecappelle wrote:
>
>> On 08/04/2019 05:22, Ricardo Martincoski wrote:
>>> The check-* jobs are fast, so there is no need to add a per job trigger
>>> for them.
>>
>> This bit is no longer true, right?
I'm sorry, I had not read it correctly. I had read "The check-* jobs are fast,
so there is no need to exclude them from the defconfig and test branches."
Indeed, there is no need to add a per job trigger. The main reason for that is
though that they'll run for every push anyway, so adding a separate trigger is
kind of useless.
>
> I still think:
> - The check-* jobs are fast
> - there is no *need* to add a per job trigger for them.
> But with v4 of this series indeed it should be easy to add a per job trigger
> for them too, for consistence or some other reason.
>
>>
>> I can fix that up while applying, no need to resend.
>
> So feel free to add them *if* you keep current behavior. For example, now when
> one pushes a branch to GitLab the 4 check-* jobs run.
> I guess you would need to do this for each of the 4 jobs:
>
> check-DEVELOPERS:
> extends: .check_base
> + only:
> + - triggers
> + - tags
> + - branches
> + - /-check-DEVELOPERS$/
>
> If you intend to change the current behavior, please do this in a follow up
> patch so we can discuss a little more during its review.
So no, that was not at all my intention.
>> Also, I'll squash patches 2 and 3. Patch 2 is really not worth much on its own
>> IMO. It would make more sense to introduce the _base keys in a separate patch
>> (i.e. moving the artifacts to _base) and include the variable name in that
>> patch. But I don't think it's worth the effort to try to split like that.
>
> OK. Please feel free to squash them.
I'll do that when I get around to applying them :-)
Regards,
Arnout
More information about the buildroot
mailing list