* [PATCH] systemd: Add systemd package to PACKAGE var
@ 2013-02-12 8:22 Khem Raj
2013-02-12 9:08 ` Ross Burton
0 siblings, 1 reply; 8+ messages in thread
From: Khem Raj @ 2013-02-12 8:22 UTC (permalink / raw)
To: openembedded-core
If someone defines SYSTEMD_PACKAGES to be different
then ${PN} then we need to make sure that they get
added to PACKAGES variable
Signed-off-by: Khem Raj <raj.khem@gmail.com>
---
meta/classes/systemd.bbclass | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/meta/classes/systemd.bbclass b/meta/classes/systemd.bbclass
index 32cc5c2..672f304 100644
--- a/meta/classes/systemd.bbclass
+++ b/meta/classes/systemd.bbclass
@@ -46,6 +46,12 @@ def systemd_populate_packages(d):
val = (d.getVar(var, True) or "").strip()
return val
+ # prepend systemd-packages not already included
+ def systemd_create_package(pkg_systemd):
+ packages = d.getVar('PACKAGES', True)
+ if not pkg_systemd in packages:
+ d.appendVar('PACKAGES', " " + pkg_systemd)
+
# Add a runtime dependency on systemd to pkg
def systemd_add_rdepends(pkg):
@@ -144,6 +150,7 @@ def systemd_populate_packages(d):
# Run all modifications once when creating package
if os.path.exists(d.getVar("D", True)):
for pkg in d.getVar('SYSTEMD_PACKAGES', True).split():
+ systemd_create_package(pkg)
if d.getVar('SYSTEMD_SERVICE_' + pkg, True):
systemd_generate_package_scripts(pkg)
systemd_add_rdepends(pkg)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 8:22 [PATCH] systemd: Add systemd package to PACKAGE var Khem Raj
@ 2013-02-12 9:08 ` Ross Burton
2013-02-12 15:01 ` Anders Darander
2013-02-12 17:35 ` Khem Raj
0 siblings, 2 replies; 8+ messages in thread
From: Ross Burton @ 2013-02-12 9:08 UTC (permalink / raw)
To: Khem Raj; +Cc: openembedded-core
On Tuesday, 12 February 2013 at 08:22, Khem Raj wrote:
> If someone defines SYSTEMD_PACKAGES to be different
> then ${PN} then we need to make sure that they get
> added to PACKAGES variable
The only case it won't already be in PACKAGES is if you're creating a package which contains just the service file, which as I've said before isn't recommended - package the service files along with the binaries that they are executing.
Or is there another use-case I'm missing?
Ross
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 9:08 ` Ross Burton
@ 2013-02-12 15:01 ` Anders Darander
2013-02-12 15:06 ` Ross Burton
2013-02-12 17:35 ` Khem Raj
1 sibling, 1 reply; 8+ messages in thread
From: Anders Darander @ 2013-02-12 15:01 UTC (permalink / raw)
To: openembedded-core
* Ross Burton <ross.burton@intel.com> [130212 10:09]:
> On Tuesday, 12 February 2013 at 08:22, Khem Raj wrote:
> > If someone defines SYSTEMD_PACKAGES to be different
> > then ${PN} then we need to make sure that they get
> > added to PACKAGES variable
> The only case it won't already be in PACKAGES is if you're creating a package which contains just the service file, which as I've said before isn't recommended - package the service files along with the binaries that they are executing.
> Or is there another use-case I'm missing?
Well, there is always the possibillity that the install is split into
multiple packages, with binaries also in some sub-packages. I can't
recall right now if we have such packages, where the binaries in the
sub-packages should be started by init, though.
Still, I'd say that it might be a valid use case for some applications.
Cheers,
Anders
--
Anders Darander
ChargeStorm AB / eStorm AB
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 15:01 ` Anders Darander
@ 2013-02-12 15:06 ` Ross Burton
2013-02-12 16:49 ` Anders Darander
0 siblings, 1 reply; 8+ messages in thread
From: Ross Burton @ 2013-02-12 15:06 UTC (permalink / raw)
To: openembedded-core
On Tuesday, 12 February 2013 at 15:01, Anders Darander wrote:
> > Or is there another use-case I'm missing?
>
> Well, there is always the possibillity that the install is split into
> multiple packages, with binaries also in some sub-packages. I can't
> recall right now if we have such packages, where the binaries in the
> sub-packages should be started by init, though.
And those sub-packages are obviously already in PACKAGES, so this isn't required again. I'm just trying to keep the amount of magic in the class to a minimum unless it's required.
Ross
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 15:06 ` Ross Burton
@ 2013-02-12 16:49 ` Anders Darander
0 siblings, 0 replies; 8+ messages in thread
From: Anders Darander @ 2013-02-12 16:49 UTC (permalink / raw)
To: openembedded-core
* Ross Burton <ross.burton@intel.com> [130212 16:07]:
> On Tuesday, 12 February 2013 at 15:01, Anders Darander wrote:
> > > Or is there another use-case I'm missing?
> > Well, there is always the possibillity that the install is split into
> > multiple packages, with binaries also in some sub-packages. I can't
> > recall right now if we have such packages, where the binaries in the
> > sub-packages should be started by init, though.
> And those sub-packages are obviously already in PACKAGES, so this
> isn't required again. I'm just trying to keep the amount of magic in
> the class to a minimum unless it's required.
Ah, of course they are...
I completely mis-read the patch... As the patch only was concerned with
adding packages to PACKAGES, I completely agree with you. Lets get rid
of that extra magic.
Cheers,
Anders
--
Anders Darander
ChargeStorm AB / eStorm AB
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 9:08 ` Ross Burton
2013-02-12 15:01 ` Anders Darander
@ 2013-02-12 17:35 ` Khem Raj
2013-02-12 21:06 ` Burton, Ross
1 sibling, 1 reply; 8+ messages in thread
From: Khem Raj @ 2013-02-12 17:35 UTC (permalink / raw)
To: Ross Burton; +Cc: openembedded-core
On Tue, Feb 12, 2013 at 1:08 AM, Ross Burton <ross.burton@intel.com> wrote:
> On Tuesday, 12 February 2013 at 08:22, Khem Raj wrote:
>> If someone defines SYSTEMD_PACKAGES to be different
>> then ${PN} then we need to make sure that they get
>> added to PACKAGES variable
>
> The only case it won't already be in PACKAGES is if you're creating a package which contains just the service file, which as I've said before isn't recommended - package the service files along with the binaries that they are executing.
>
> Or is there another use-case I'm missing?
Thinking about it from different perspective, I agree that probably
adding an extra package is not right and having unitfiles as part of
package proper is fine. but we should definitely have a check where if
SYSTEMD_PACKAGES dont exist in PACKAGES then it should error out or
warn about it.
>
> Ross
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 17:35 ` Khem Raj
@ 2013-02-12 21:06 ` Burton, Ross
2013-02-12 22:42 ` Khem Raj
0 siblings, 1 reply; 8+ messages in thread
From: Burton, Ross @ 2013-02-12 21:06 UTC (permalink / raw)
To: Khem Raj; +Cc: openembedded-core
On 12 February 2013 17:35, Khem Raj <raj.khem@gmail.com> wrote:
> Thinking about it from different perspective, I agree that probably
> adding an extra package is not right and having unitfiles as part of
> package proper is fine. but we should definitely have a check where if
> SYSTEMD_PACKAGES dont exist in PACKAGES then it should error out or
> warn about it.
Yes, throwing an error/warning does make sense, as the alternative
would be files mysteriously disappearing. Can you send a patch for
that?
Ross
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] systemd: Add systemd package to PACKAGE var
2013-02-12 21:06 ` Burton, Ross
@ 2013-02-12 22:42 ` Khem Raj
0 siblings, 0 replies; 8+ messages in thread
From: Khem Raj @ 2013-02-12 22:42 UTC (permalink / raw)
To: Burton, Ross; +Cc: openembedded-core
On Tue, Feb 12, 2013 at 1:06 PM, Burton, Ross <ross.burton@intel.com> wrote:
>
> Yes, throwing an error/warning does make sense, as the alternative
> would be files mysteriously disappearing. Can you send a patch for
> that?
I have something for test.
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2013-02-12 22:58 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-02-12 8:22 [PATCH] systemd: Add systemd package to PACKAGE var Khem Raj
2013-02-12 9:08 ` Ross Burton
2013-02-12 15:01 ` Anders Darander
2013-02-12 15:06 ` Ross Burton
2013-02-12 16:49 ` Anders Darander
2013-02-12 17:35 ` Khem Raj
2013-02-12 21:06 ` Burton, Ross
2013-02-12 22:42 ` Khem Raj
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.