Re: netbsd-7 kernel and older vnconfig(8)

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Mon, 9 Nov 2015 17:11:11 +0000 (UTC)
    From:        christos%astron.com@localhost (Christos Zoulas)
    Message-ID:  <n1qk3e$935$1%ger.gmane.org@localhost>

  | I think that the best solution is to make the kernel return ENXIO again,
  | fix the userland to handle this in current, and issue a pullup for -7
  | for both userland and kernel. At least this makes things more stable in
  | the long term.

Agreed.   Pity this wasn't caught before 7 was released, then there'd
be no question that is the right way, but now we must either break
netbsd-6 (and before) userland, or netbsd 7 userland.

Aside from it being the better way, it is also the less intrusive to fix
it that way ... a netbsd-6 userland with 7 (or current) kernel will loop
(I expect) until it runs out of integers...   the only effect on a netbsd-7
vndconfig on an updated kernel would be to print an error message when it
reaches the end of the kernel-configured vnd's, instead of either just
printing "unconfigured" or terminating the loop.

Ie: annoying, rather than evil.

kre