Re: [OE-core] [openembedded-core][scarthgap][PATCH] tcp-wrappers: extend recipe for native/nativesdk builds

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Wed, 2026-03-18 at 17:51 +0100, Yoann Congal via lists.openembedded.org wrote:
> On Mon Mar 9, 2026 at 3:56 PM CET, Deepak Rathore via lists.openembedded.org wrote:
> > From: Deepak Rathore <deeratho@cisco.com>
> > 
> > Some SDK dependency chains require tcp-wrappers to be available
> > as host/SDK artifacts. The current metadata only partially provides this,
> > which can lead to dependency resolution failures when this recipe is pulled
> > into SDK-oriented builds.
> > 
> > This change does not alter target package behavior; it only enables required
> > native/nativesdk variants for build and SDK integration paths.
> > 
> > Signed-off-by: Deepak Rathore <deeratho@cisco.com>
> > 
> > diff --git a/meta/recipes-extended/tcp-wrappers/tcp-wrappers_7.6.bb b/meta/recipes-extended/tcp-wrappers/tcp-wrappers_7.6.bb
> > index bcd1d6f792..26eaeab1ac 100644
> > --- a/meta/recipes-extended/tcp-wrappers/tcp-wrappers_7.6.bb
> > +++ b/meta/recipes-extended/tcp-wrappers/tcp-wrappers_7.6.bb
> > @@ -132,3 +132,4 @@ do_install () {
> >  }
> >  
> >  FILES:${PN} += "${sysconfdir}/hosts.allow ${sysconfdir}/hosts.deny"
> > +BBCLASSEXTEND = "native nativesdk"
> 
> That effectively create 2 new recipes tcp-wrappers-native and
> nativesdk-tcp-wrappers. That is not a trivial change that I can take
> without a clear and strong justification.
> 
> What "dependency resolution failures" did you see?
> 
> Also, can you check that tcp-wrapper-native builds on Fedora 41? This is
> the scarthgap supported distro with the most recent GCC : gcc-14 (and it
> usually trigger build failures on old/unmaintained packages)
> 
> Finally, OE-Core master did remove the recipe for it being really
> unmaintained[0]. This is not a good sign :/

The other question that needs to be asked is why would you even need
tcp-wrappers-native? Why are you creating TCP services using the native
tools in a build? It doesn't make any sense.

We've had a flood of these "extend for native/nativesdk" patches and
I'm afraid I'm leaning toward saying no to all of them until we better
understand the actual use cases and make sure this isn't just a random
dependency explosion as people haven't actually thought about the
dependencies.

Cheers,

Richard