[oe] [PATCH][meta-oe] systemd: add postrm u-a calls

[Anders Darander]

* Koen Kooi <koen at dominion.thruhere.net> [111121 10:34]:
> Op 21-11-11 09:46, Martin Jansa schreef:
> > On Wed, Nov 16, 2011 at 03:15:17PM +0100, Martin Jansa wrote:
> >> On Thu, Nov 10, 2011 at 09:22:12AM +0100, Anders Darander wrote:
> >>> * Martin Jansa <martin.jansa at gmail.com> [111110 08:57]:
> >>>> On Thu, Nov 10, 2011 at 08:44:49AM +0100, Anders Darander wrote:
> >>>>> * Martin Jansa <martin.jansa at gmail.com> [111109 20:25]:
> >>>>>> Signed-off-by: Martin Jansa <Martin.Jansa at gmail.com> --- 
> >>>>>> meta-oe/recipes-core/systemd/systemd_git.bb |   11 +++++++++-- 
> >>>>>> 1 files changed, 9 insertions(+), 2 deletions(-)

> >>>>>> diff --git a/meta-oe/recipes-core/systemd/systemd_git.bb
> >>>>>> b/meta-oe/recipes-core/systemd/systemd_git.bb index
> >>>>>> 90862a3..d1d4306 100644 ---
> >>>>>> a/meta-oe/recipes-core/systemd/systemd_git.bb +++
> >>>>>> b/meta-oe/recipes-core/systemd/systemd_git.bb @@ -14,7 +14,7 @@
> >>>>>> inherit gitpkgv # TODO: # u-a for runlevel and telinit -# u-a
> >>>>>> prerm

> >>>>> Here on your todo list, it stated u-a for prerm, but you applied
> >>>>> it to the postrm. In this case, it probably doesn't really
> >>>>> matter, but I'd still guess that it would make more sense to
> >>>>> remove the u-a before removing the binaries. Although, I don't
> >>>>> think it'll actually cause any problems in this case.

> >>>> I did what u-a.bbclass usually do.. if you think it should be in
> >>>> prerm then it would be best to first fix in u-a.bbclass and then we
> >>>> can update the rest of metadata.

> >>> Yeah, like I said, is't likely not a real problem. There'll probably 
> >>> just be a tiny window during which, in this case, e.g. reboot isn't 
> >>> available, before u-a has been told to change the reboot link.

> >>> I'm currently just a little bit catious as I've recently stumbled
> >>> upon post-/prerm issues using rpm... (Although the ones I know of
> >>> (busybox updates, rsyslog removal) are either fixed or there has been
> >>> patches submitted).

> >> So should I change postrm to prerm in this patch or is there other 
> >> reason why it wasn't applied yet?

> I was waiting for Anders to comment on it :) Changing to prerm would be better.

Sorry, for the delay. I've been out of the loop for almost a week
(hardly read any emails at all).

Well, prerm would probably be better, as Koen stated, not least
long-ternm. But postrm has been working as far as I see, so I'll leave
it to you to decide.

Cheers,
/Anders

-- 
Anders Darander
ChargeStorm AB / eStorm AB