All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ondrej Kozina <okozina@redhat.com>
To: Yoann CONGAL <yoann.congal@smile.fr>, dm-crypt@saout.de
Subject: [dm-crypt] Re: LUKS container creation without device mapper or loop device access
Date: Wed, 30 Jun 2021 11:10:17 +0200	[thread overview]
Message-ID: <f733c8c7-ea1b-9acd-3212-df2fca306500@redhat.com> (raw)
In-Reply-To: <CAMSfU+5R5ZBeVjCkbeqRh33g_rJpZRrygs89zyf59s2Nb7QXrQ@mail.gmail.com>

Hi,

On 6/18/21 10:22 AM, Yoann CONGAL wrote:
> 
>  From what I understood of the internals of cryptsetup, it knows how to
> build the LUKS header but rely on the dm-crypt module of the kernel to
> do the actual data encryption. (Please correct me if I'm wrong)

Yes, dm-crypt is usually only necessary to access data when LUKS device 
is activated (unlocked). That said, there are some exceptions. For 
example when crypto backend used in libcryptsetup (or kernel crypto API) 
does not support used cipher/mode for some reason. In that case we 
fallback to use dm-crypt to perform encryption/decryption of LUKS 
keyslots. It also requires root privs in this corner case.

> 
> So, I have two questions :
> * Do you know of a tool that does the full LUKS image (header and
> data) fully in userland? (I did search for it and found nothing)
> * If the above answer is "It does not exist yet", would you be open to
> its inclusion in cryptsetup? My guess is that a tightly managed intern
> may handle this.
With default cipher (aes) you can use new LUKS2 reencryption code for 
that. LUKS2 header (cryptsetup format) can be created fully without need 
to use dm-crypt already, but If you need to encrypt existing data you 
can use following command:

This should work without root privs. It will create separate detached 
LUKS2 header in <new_detached_LUKS2_header>:

cryptsetup reencrypt --encrypt <my_data_file> --header 
<new_detached_LUKS2_header> --disable-locks

For header put in the beginning of the data file you can use:

cryptsetup reencrypt --encrypt <my_data_file> --reduce-device-size 32M 
--disable-locks

just bear in mind that my_data_file must have 32MiB spare space at the 
end (iow there should be no useful data at the end of the file).

With root privs, you can drop --disable-locks parameter and also use 
block devices in place of <my_data_file>.

Look for more information related to online encryption under "reencrypt" 
action of cryptsetup.

Kind regards
Ondrej K.

_______________________________________________
dm-crypt mailing list -- dm-crypt@saout.de
To unsubscribe send an email to dm-crypt-leave@saout.de

  reply	other threads:[~2021-06-30  9:19 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-18  8:22 [dm-crypt] LUKS container creation without device mapper or loop device access Yoann CONGAL
2021-06-30  9:10 ` Ondrej Kozina [this message]
2021-07-01 15:57   ` [dm-crypt] " Yoann CONGAL

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=f733c8c7-ea1b-9acd-3212-df2fca306500@redhat.com \
    --to=okozina@redhat.com \
    --cc=dm-crypt@saout.de \
    --cc=yoann.congal@smile.fr \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.