Re: [OE-core] [PATCH] systemd: stop enabling non-standard MAC policy when using the 'pni-names' DISTRO_FEATURE

["Antonin Godard" <antonin.godard@bootlin.com>]

On Fri Sep 5, 2025 at 3:09 PM CEST, Koen Kooi via lists.openembedded.org wrote:
>
>
>> Op 5 sep 2025, om 14:02 heeft Koen Kooi via lists.openembedded.org <koen.kooi=oss.qualcomm.com@lists.openembedded.org> het volgende geschreven:
>> 
>> 
>> 
>>> Op 5 sep 2025, om 11:09 heeft Ross Burton <ross.burton@arm.com> het volgende geschreven:
>>> 
>>> On 5 Sep 2025, at 09:33, Antonin Godard <antonin.godard@bootlin.com> wrote:
>>>>> + if ${@ 'true' if not bb.utils.contains('PACKAGECONFIG', 'pni-names', 'true', 'false', d) else 'false'}; then
>>>> 
>>>> This could be simplified to 
>>>> 
>>>> if ${@bb.utils.contains('PACKAGECONFIG', 'pni-names', 'false', 'true', d)}; then
>>>> 
>>>> no?
>>> 
>>> I dislike contains() calls that return false,true because that subverts expectations,
>> 
>> Same here, which is why I lifted this variant from an existing OE-core bbclass :) 
>> 
>>> but this seems like a compromise to me:
>>> 
>>> if not ${@bb.utils.contains('PACKAGECONFIG', 'pni-names', ’true', ‘false', d)}; then
>> 
>> I can rework it to that. If that works, a v2 will best sent!
>
> So that *technically* works because 'not' is both not a bash builtin nor a command installed in the sysroot. This ends up giving us a 
>
> not: not found 
>
> in log.do_install. I don't think this is an improvement over the original patch. Thoughts?

It should be 

if ! ${@bb.utils.contains('PACKAGECONFIG', 'pni-names', ’true', ‘false', d)}; then

We mixed up python and shell syntax here :)

Antonin

-- 
Antonin Godard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com