wg(4): switch to the new if_clone interface

[asomers]

Summary

MFC after: 2 weeks

Test Plan

manual testing

[zlei]

Glad to see this feature!

[kevans]

sys/dev/wg/if_wg.c

2925

What's the practical difference between if_clone_simple and ifc_attach_cloner that's changing something here? I note that I don't seem to have any problems creating a wg interface in a vnet jail, so I'm not sure I actually understand the problem.

[zlei]

sys/dev/wg/if_wg.c

2925

I think this is a follow-up of https://reviews.freebsd.org/D36636 , and should be re-titled .

[jhb]

sys/dev/wg/if_wg.c

2925

Yes, this is just moving to the "newer" clone API it seems, but no functional change. The title of the review should be updated to reflect that.

[asomers]

sys/dev/wg/if_wg.c

2925

When I tried to create a wg interface in a jail, it failed, and dtrace indicated that the failure was in ifc_create_ifp . This patch fixed it. Or at least, this patch "fixed" it. But it's possible that that failure was due to a different reason. Perhaps because my build was a few weeks old, or perhaps I ran into https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=264981 . I'll retest again without this patch, probably next weekend.

[asomers]

@jhb @kevans you're right, this change isn't necessary to do what I said. I think I got confused by two other problems:

• The jail can't load the if_wg kernel module, and
• https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=264981

Is there any value to this change anyway, as an act of maintenance? Or should I just abandon it?

[kevans]

In D37740#860416, @asomers wrote:

@jhb @kevans you're right, this change isn't necessary to do what I said. I think I got confused by two other problems:

• The jail can't load the if_wg kernel module, and
• https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=264981

Is there any value to this change anyway, as an act of maintenance? Or should I just abandon it?

Ahh!

If we're switching other stuff away from the if_clone_{simple,advanced} KPI anyways, I don't really see any reason not to proceed (modulo commit message, of course).

[asomers]

• Also replace if_clone_detach with ifc_detach_cloner

[jhb]

Agree that moving to the new KPI is the right thing to do.

[emaste]

sounds good to me