On 2023-08-01, at 00:15:13 +0200, Florian Westphal wrote: > Florian Westphal <fw@xxxxxxxxx> wrote: > > I'll apply it on Friday in case noone objects until then. > > Just for completeness: I did not apply it due to the unanswered > question from Pablo. Thanks, Florian. I had hoped to have reponded over the week-end, but I was busier than I expected and didn't pick this up again till yesterday evening. Answers should be forthcoming soon. > Also, don't we risk ambiguity here, or is there a guarantee that > kernel never emits mapped addresses? It would indeed mean that real ipv4-in-ipv6 addresses would be output in ipv4 format. It's been several months, so I can't quite remember my thinking, but I probably inclined towards regarding the ambiguity as benign. I have spotted a problem with the existing patch, however, so another version will definitely be needed, and I can look at eliminating the ambiguity at the same time. J.
Attachment:
signature.asc
Description: PGP signature
