On Thu, Jul 23, 2020 at 1:06 AM Dan Callaghan
<[email protected]> wrote:
>
> Excerpts from Khem Raj's message of 2020-07-22 23:09:23 -07:00:
> > there is stress recipe in some layers which could provide stress
> > utility as well, therefore it should be using update-alternatives
> > mechanism so it fills in when stress is not provided by say stress
> > recipe.
>
> Hmm... The stress-ng recipe already RPROVIDES stress though. Doesn't
> that make it incompatible with any other existing stress recipe and
> update-alternatives?
>
> Personally I think alternatives is overkill here, and stress-ng should
> not RPROVIDES stress at all (that is, revert commit 72641004). In that
> case, we would have noticed that the stress recipe had been removed from
> oe-core and we could have just updated our image accordingly to depend
> on stress-ng instead.
>
yeah I think its a fair point, we already are providing it in that
case we can live with unconditional
symlink and let older stress recipe deal with conflicts if someone needs it
> > secondly please use ${D}${bindir}/stress instead of hardcoding the path
>
> Oops, yes of course.
>
> --
> Dan Callaghan <[email protected]>
> Software Engineer
> Opengear <https://opengear.com/>
>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#140891):
https://lists.openembedded.org/g/openembedded-core/message/140891
Mute This Topic: https://lists.openembedded.org/mt/75740083/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
