Todd Vierling wrote:
> 
> I'm rather annoyed with the fact that vnd cannot stack atop a raw disk
> device, so I'd much prefer the ccd version.  (If I want an encrypted
> partition, it should be a *partition*, dammit, not a container file.  :)

It did seem like a mess, but a good place to start with the crypt stuff,
as I could much more easily check the on disk data, and ofcourse test it
on systems without re-partitioning. :)


> 
> [for the ccd version]
> 
> If you want to put the diffs up for FTP, after you've verified that striping
> and so forth still work, please do, and post the URL on tech-kern.  (Though
> we do tend to prefer uuencoded tarballs or raw "diff -u" for submitting via
> a PR.)

If that is what they prefer, then I will do that. I don't want to rattle
the cage :) Do they prefer diff's from top of source tree, or each file
induvidually from '.'?


> 
> Well, I'd presume that RAIDframe would be more reliable for this.  Also,
> NetBSD's vfs layer has changed its locking behavior significantly from
> FreeBSD, so you need to be sure to get that right.  So "feel free to merge
> it," if you want.  Good code submissions are always welcome....

Of course. I prefer RAIDframe to ccd generally as the raidctl program is
really easy to use. It is my intention to bring my patches from vnd/ccd
into raid as well, as they all have a similar "frame work". RAIDframe
could be abit more awkward as I (believe) it is work upon by a group,
not neccessarily in NetBSD so I would rather have their permission. Not
sure how it hangs together in those terms.

Lund

-- 
Jorgen "Lord" Lundman <lundman@lundman.net>
Technology Manager, Unix Administrator
Phone: +44 (0)20-86591860  Mobile: +44 (0)79-58642918
Pager: 07958642918@one2one.net
"Rare is the person who can weigh the faults of others 
 without putting his thumb on the scales": Byron J. Langenfeld