Re: Re: Re: [patch] cgd

[Taylor R Campbell <campbell+netbsd%mumble.net@localhost>]

   Date: Thu, 2 Dec 2010 05:07:10 +0000
   From: "Roland C. Dowdeswell" <elric%imrryr.org@localhost>

   So, the part of your patch:

   [...]

   Your use case here is to generate a new params file which has in it
   only a stored key?

Partially, yes -- for that case, the patch makes cgdconfig produce a
single storedkey entry with the actual key, deterministically.  I had
a few ideas in mind for this (which had to do with writing test cases
for cgdconfig, verifying the pass phrase with `cgdconfig -G', &c.),
but it's not important.

The other purpose of that part of the patch is to call params_verify
before keygen_generate.  As is, `cgdconfig -G -k pkcs5_pbkdf2/sha1'
emits a spurious warning, and there may be other related problems too.