Change new items to use only one sentence per line; other touchups.

I was also wondering whether I should change the older entries from the imperative "Add ..." to "... was added" etc. That should be a separate commit though. Comments?

I moved the initialization of nifp into the loop construct. I'm also going to correct the comment at line 833 to say that we release our request for promiscuous multicast. If this sounds OK, I'll commit, otherwise push a new version for the review.

I looked again at the race. I don't think there is one. As ALLMULTI is reference-counted, it doesn't really matter if we drop it after another mrouter has added it.

I'll mention the race in the comment.

Any comments or approvals?

There is a potential race here, if a multicast router socket is opened while this is still closing, as the if_allmulti() calls are done with no lock. I don't see a good solution, and that seems highly unlikely.

Am I allowed/able to commit this with only a src commit bit, or can someone else do it?

I used the previous URL because it was cited in the commit, but I suppose it makes sense to use the newer one.

Update URL.

Any comments from the doc folks?

Where do we stand on this? Is the change acceptable?

Oops, you are right; I tried to combine tests and it failed to check for a final NUL properly. Fixed, and I changed to isprint(); that should be liberal enough.

Fix test for missing NUL, switch to isprint().

Check for ASCII string for serial-number, check len of hw.fdt strings.

Would it be sufficient to check for ASCII with a NUL termination at the end?

Added copyright; touched up hardware info

Good point. I had more hardware information in the Description section originally, but moved it to Hardware. I'll try to come up with something more generic. I still want to mention the RPi 4, as that is the most common.

Abandoning since there are no reviewers, will commit without review.

Add missing newline

Fix delete case

No one seems to have done structural work on ifconfig for a while. Suggestions for other reviewers welcome, but I think it is pretty well tested.

Change getifflags() err_ok parameter to bool

Closing review, patch in progress upstream.

Thanks, sounds good.

The current code works for now, but there is a proposal for a change that would make the old code fail. Note that IN_LOOPBACK() is not in Linux; not sure about other BSDs. The change could be conditional on #ifdef IN_LOOPBACK.