1
$\begingroup$

If the LUKS header is removed from the LUKS partition, will this make the brute-force attack much more difficult? Can software tell which type of partition is it?

$\endgroup$

1 Answer 1

1
$\begingroup$

If we look at a sample Luks header ( The tables taken from here )


  1. Luks Partition Header (PHDR) consists of the necessary information to decrypt/encrypt the luks device as followings,
-------------------------------+-----------------------------
|   Partition Header Contents  |      Output by luksDump    |
-------------------------------+-----------------------------
|  Magic Byte                  | LUKS                       |
-------------------------------+-----------------------------
| Luks version                 | 1                          |
-------------------------------+-----------------------------
| Cipher Name                  | serpent                    |
-------------------------------+-----------------------------
| Cipher Mode                  | xts-plain64                |
-------------------------------+-----------------------------
| Hash Type                    | whirpool                   |
-------------------------------+-----------------------------
| Payload offset               | 4096                       |
-------------------------------+-----------------------------
| Master Key Bits              | 512                        |
-------------------------------+-----------------------------
| Master Key Digest            | 8a 90 77 7e ...            |
-------------------------------+-----------------------------
| Master Key Salt              | ca f3 40 fc ...            |
-------------------------------+-----------------------------
| Master Key Digest Iteration  | 23250                      |
-------------------------------+-----------------------------
| Partition UUID               | 46927c7d-7f17-4312-...     |
-------------------------------+-----------------------------
                          Table - 1

2. 2nd portion of partition header (PHDR) consists of 8 key-slot descriptors i.e information about key-slot as following,
------------------------------+------------------------------
| Key Descriptors             |      Output by luksDump     |
+-----------------------------+------------------------------
| Activation status           | Enabled                     |
+-----------------------------+------------------------------
| Iteration                   | 146285                      |
+-----------------------------+------------------------------
| Salt                        | 1e c1 94 24 36 4c 98..      |
+-----------------------------+------------------------------
| Key Material Offset         | 8                           |
+-----------------------------+------------------------------
| AF Stripes                  | 4000                        |
------------------------------+------------------------------
                         Table - 2

will this make the bruteforce crack much more difficult?

One will see that an attacker will lose a lot of information about the header like cipher name, cipher mode, iteration, hash, key slots, etc.

The easy attack on the LUKS is brute-forcing the password (hashcat can do this) if the password is weak. When the LUKS header is lost, the attacker cannot test the password anymore; the user passwords decrypt the related key slot to release the Master Key (MK). If there is no data to decrypt, it is a random process that is not better than brute0force.

The rest is brute-forcing the MK. For 128-bit encryption algorithms, except for a foreseeable Cryptographic Quantum Computer (CQC), it is infeasible. When one uses a cipher with a 256-bit key it is safe from all targets including CQC.

Can softwares tell which type of the partition is?

No!

And, make sure that you removed the header without leaving a trace, i.e. secure erase.

$\endgroup$
5
  • 1
    $\begingroup$ Note that this particular header indicates a 256-bit cipher in XTS mode, which means the effective security is actually 384 bits (not 512, but that's due to a meet-in-the-middle attack). $\endgroup$
    – forest
    Commented Apr 9, 2022 at 20:54
  • $\begingroup$ So I guess if the attacker gets the LUKS disk without LUKS header, he/she/they have to guess to rebuild the LUKS header first, but it is hard to test every try, and every try leads so many tries... Is it also work for quantum computer in the future? They say quantum computer crack the algorithm and encryption. $\endgroup$ Commented Apr 10, 2022 at 4:48
  • $\begingroup$ @kytebucks000 Rebuilding is not really possible since there is no encrypted MK anymore. As I said, quantum attacks really depend on the key size though even 128-bit is arguable since it is not clear how one can run the sequential calls of Grover's machine.. $\endgroup$
    – kelalaka
    Commented Apr 10, 2022 at 7:47
  • $\begingroup$ brilliant answers! $\endgroup$ Commented Apr 15, 2022 at 9:15
  • $\begingroup$ @kytebucks000 well, in SO network, upvote (10Pts)is used to show that ( you need 15 points for this and If you edit a post, you will get +2 points) and accepting the answer (15pts to the answerer and 20pts to you) is performed by clicking the checkmark. This indicates the other that the problem is solved. Have fun. $\endgroup$
    – kelalaka
    Commented Apr 15, 2022 at 13:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.