Re: Confusion on GPT wedges

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

    Date:        Wed, 9 Jan 2019 08:44:30 -0600
    From:        Robert Nestor <rnestor%mac.com@localhost>
    Message-ID:  <54A34E8D-120E-47DA-B2E6-E00C113E03A8%mac.com@localhost>

  | Is this a bug or oversight in wedge processing by the components used for booting?

Yes.

If you were asking which of those, I'd ask why you'd care, or what is the
difference -- except I suppose it would be slassified as a bug if the code was
there to attempt to do it, but didn't work properly, and an oversight if it is just
that no-one thought  of that issue yet ... but that is really a bug too.)

Wedge support for booting is relatively recent, so not everything has been
done yet .. code for this gets added as someone requires it (hint hint ...)

  | My work-around at this point is to uniquely name the wedges on both disk volumes,

That's what I do, I have things like ROOT_0 and ROOT_1 which all works well 
enough.

kre

ps: there are more places we need proper wedge support - like in dump(8).
I recently did a level 0 dump of a filesys, followed later by an incremental.
The incremental turned ot to be (effectively) another level 0 (though not
called that) as the filesystem was on a different wedge the second time than
it had been the first.   It would have been much worse had the wedge it was
on the 2nd time been a wedge that had been used for a filesystem dump for
a different filesyste,.   That would have generated a real mess - so I was
lucky.   [And aside: no, I am not going to switch from dump(8) to some other
method of doing backups, it is far superior to anything else around.]