obmc-op-control-power_git.bb refactoring

[Lei YU]

On Tue, Nov 15, 2022 at 3:49 AM Andrew Geissler <geissonator at gmail.com> wrote:
>
>
>
> > On Nov 10, 2022, at 5:41 PM, Andrew Jeffery <andrew at aj.id.au> wrote:
> >
> > On Fri, 11 Nov 2022, at 07:29, Andrew Geissler wrote:
> >>
> >> Here's what I'm thinking:
> >>
> >> Move these services to PSM and rename them as follows:
> >> - op-power-start at .service -> phosphor-power-start at .service
> >> - op-power-stop at .service -> phosphor-power-stop at .service
> >
> > This isn't terribly constructive but I feel like given s/op-/phosphor-/
> > results in names like 'phosphor-power*' it's going to be confusing with
> > respect to the existing phosphor-power subproject.
>
> Yeah, good point. We have another naming convention on some service
> files, “obmc”, so I’ve put that up instead of “phosphor” for the service
> files in question.
>
> Here’s all the reviews for anyone interested:
>   https://gerrit.openbmc.org/q/topic:op-power-control-ref
>
> Need to merge the phosphor-state-manager changes first and then
> I can officially put the openbmc/openbmc changes up for review.
>
> >
> > Not sure how to fix it though.
> >
> > Andrew

There are several situations for now:

* There are a set of `obmc-host-[start|stop|...]` and
`obmc-chassis-[poweron|poweroff|...]` targets
* Now we will have `phosphor-power-xxx` services as proposed by this change
* For intel-x86 systems, it removes all the host/chassis related
targets in meta-intel-openbmc/meta-common/recipes-core/systemd/obmc-targets.bbappend
* I am not sure how amd-x86 and arm systems handle the targets.


The targets and services are a bit confusing, especially for different systems.

Hopefully we could get a better naming convention so that we do know
which are generic and which are platform specific?