[dm-crypt] [ANNOUNCE] cryptsetup 1.6.4

Sven Eschenberg sven at whgl.uni-frankfurt.de
Fri Feb 28 23:06:30 CET 2014


On Fri, February 28, 2014 22:46, Arno Wagner wrote:
> On Fri, Feb 28, 2014 at 22:26:03 CET, Sven Eschenberg wrote:
>> Just out of curiosity,
>>
>> Isn't it possible (yet) to override header fields during luksopen? If
>> not,
>> wouldn't it make sense to add something like that in future versions? I
>> think it could be of great help when the header is partly damaged, to be
>> able to override things without using a hex editor.
>
> I doubt this makes much sense. From what I have seen,
> usually the magic string at the start is gone as well,
> and then there is a real risk that people try this with
> the wrong data. Using a hex-editor is not that hard and
> using a hex-dumper is basically required to get any
> reasonable form of diagnostics. Even the keyslot-
> checker is basically a specialized hexdump tool.

Okay, that's true. I personally do use a hexeditor too, I have to admit,
just thought it could help less experienced people. Then again, if people
fail to use a hex editor chances are big they make things worse. I guess I
didn't think this through long enough.

>
>> I am aware that one could use the non-LUKS mode to open a LUKS device by
>> passing all required parameters, admitted. But a mode where one can use
>> what's in the header and override single fields could be interesting.
>> Once
>> the correct params are determinde this way, one could maybe add an
>> option
>> to repair the header with the given replacements (Maybe by adding the
>> option to reencryt?).
>>
>> Just some thoughts that crossed my mind.
>
> I doubt this really helps. Also remember that finding out what
> actually broke the header is important, so fiddeling around
> with an opaque header and commandline arguments to cryptsetup
> after you have analyzed a hexdump strikes me as not that effective.

That's absolutely true, indeed.

>
> I do understand that hex-editing is akward for many people,
> but I do not think this makes it any better or clearer.
> One thing that would help a bit is a header layout with
> hex offsets. I think I am going to add that to the FAQ.

Good point, I'd propose adding this to the luks on disk format document as
well. If you cannot convert HEX<->DEC inside your head having the hex
offsets at disposal helps alot. I admit, I used a converter when I edited
my headers lately.

>
> Admittedly, the whirlpool problem (BTW, now documented in
> FAQ Item 8.3) would be easy to solve with your proposal.
>
> Arno
> --
> Arno Wagner,     Dr. sc. techn., Dipl. Inform.,    Email: arno at wagner.name
> GnuPG: ID: CB5D9718  FP: 12D6 C03B 1B30 33BB 13CF  B774 E35C 5FA1 CB5D
> 9718
> ----
> A good decision is based on knowledge and not on numbers. -  Plato
> _______________________________________________
> dm-crypt mailing list
> dm-crypt at saout.de
> http://www.saout.de/mailman/listinfo/dm-crypt
>

Looking at the steps to solve the whirlpool issue basicly triggered the
idea (Even though I am not affected). And thanks for keeping the FAQ
current and for your work on it in general, I am sure it helps a lot of
people.

-Sven




More information about the dm-crypt mailing list