8 6. Backup and Data Recovery
9 7. Interoperability with other Disk Encryption Tools
10 8. Issues with Specific Versions of cryptsetup
19 This is the FAQ (Frequently Asked Questions) for cryptsetup. It
20 covers Linux disk encryption with plain dm-crypt (one passphrase,
21 no management, no metadata on disk) and LUKS (multiple user keys
22 with one master key, anti-forensic features, metadata block at
23 start of device, ...). The latest version of this FAQ should
24 usually be available at
25 http://code.google.com/p/cryptsetup/wiki/FrequentlyAskedQuestions
30 ATTENTION: If you are going to read just one thing, make it the
31 section on Backup and Data Recovery. By far the most questions on
32 the cryptsetup mailing list are from people that managed to damage
33 the start of their LUKS partitions, i.e. the LUKS header. In
34 most cases, there is nothing that can be done to help these poor
35 souls recover their data. Make sure you understand the problem and
36 limitations imposed by the LUKS security model BEFORE you face
37 such a disaster! In particular, make sure you have a current header
38 backup before doing any potentially dangerous operations.
40 BACKUP: Yes, encrypted disks die, just as normal ones do. A full
41 backup is mandatory, see Section "6. Backup and Data Recovery" on
42 options for doing encrypted backup.
44 DISTRIBUTION INSTALLERS: Some distribution installers offer to
45 create LUKS containers in a way that can be mistaken as activation
46 of an existing container. Creating a new LUKS container on top of
47 an existing one leads to permanent, complete and irreversible data
48 loss. It is strongly recommended to only use distribution
49 installers after a complete backup of all LUKS containers has been
52 NO WARNING ON NON-INERACTIVE FORMAT: If you feed cryptsetup from
53 STDIN (e.g. via GPG) on LUKS format, it does not give you the
54 warning that you are about to format (and e.g. will lose any
55 pre-existing LUKS container on the target), as it assumes it is
56 used form a script. In this scenario, the responsibility for
57 warning the user and possibly checking for an existing LUKS header
58 goes over to the script. This is a more general form of the
61 LUKS PASSPHRASE IS NOT THE MASTER KEY: The LUKS passphrase is not
62 used in deriving the master key. It is used in decrypting a master
63 key that is randomly selected on header creation. This means that
64 if you create a new LUKS header on top of an old one with
65 exactly the same parameters and exactly the same passphrase as the
66 old one, it will still have a different master key and your data
67 will be permanently lost.
69 PASSPHRASE CHARACTER SET: Some people have had difficulties with
70 this when upgrading distributions. It is highly advisable to only
71 use the 94 printable characters from the first 128 characters of
72 the ASCII table, as they will always have the same binary
73 representation. Other characters may have different encoding
74 depending on system configuration and your passphrase will not
75 work with a different encoding. A table of the standardized first
76 128 ASCII caracters can, e.g. be found on
77 http://en.wikipedia.org/wiki/ASCII
80 * System Specific warnings
82 - Ubuntu as of 4/2011: It seems the installer offers to create
83 LUKS partitions in a way that several people mistook for an offer
84 to activate their existing LUKS partition. The installer gives no
85 or an inadequate warning and will destroy your old LUKS header,
86 causing permanent data loss. See also the section on Backup and
89 This issue has been acknowledged by the Ubuntu dev team, see here:
90 http://launchpad.net/bugs/420080
95 Current FAQ maintainer is Arno Wagner <arno@wagner.name>. Other
96 contributors are listed at the end. If you want to contribute, send
97 your article, including a descriptive headline, to the maintainer,
98 or the dm-crypt mailing list with something like "FAQ ..." in the
99 subject. You can also send more raw information and have me write
100 the section. Please note that by contributing to this FAQ, you
101 accept the license described below.
103 This work is under the "Attribution-Share Alike 3.0 Unported"
104 license, which means distribution is unlimited, you may create
105 derived works, but attributions to original authors and this
106 license statement must be retained and the derived work must be
107 under the same license. See
108 http://creativecommons.org/licenses/by-sa/3.0/ for more details of
111 Side note: I did text license research some time ago and I think
112 this license is best suited for the purpose at hand and creates the
116 * Where is the project website?
118 There is the project website at http://code.google.com/p/cryptsetup/
119 Please do not post questions there, nobody will read them. Use
120 the mailing-list instead.
123 * Is there a mailing-list?
125 Instructions on how to subscribe to the mailing-list are at on the
126 project website. People are generally helpful and friendly on the
129 The question of how to unsubscribe from the list does crop up
130 sometimes. For this you need your list management URL, which is
131 sent to you initially and once at the start of each month. Go to
132 the URL mentioned in the email and select "unsubscribe". This page
133 also allows you to request a password reminder.
135 Alternatively, you can send an Email to dm-crypt-request@saout.de
136 with just the word "help" in the subject or message body. Make sure
137 to send it from your list address.
139 The mailing list archive is here:
140 http://dir.gmane.org/gmane.linux.kernel.device-mapper.dm-crypt
146 * What is the difference between "plain" and LUKS format?
148 Plain format is just that: It has no metadata on disk, reads all
149 paramters from the commandline (or the defaults), derives a
150 master-key from the passphrase and then uses that to de-/encrypt
151 the sectors of the device, with a direct 1:1 mapping between
152 encrypted and decrypted sectors.
154 Primary advantage is high resilience to damage, as one damaged
155 encrypted sector results in exactly one damaged decrypted sector.
156 Also, it is not readily apparent that there even is encrypted data
157 on the device, as an overwrite with crypto-grade randomness (e.g.
158 from /dev/urandom) looks exactly the same on disk.
160 Side-note: That has limited value against the authorities. In
161 civilized countries, they cannot force you to give up a crypto-key
162 anyways. In the US, the UK and dictatorships around the world,
163 they can force you to give up the keys (using imprisonment or worse
164 to pressure you), and in the worst case, they only need a
165 nebulous "suspicion" about the presence of encrypted data. My
166 advice is to either be ready to give up the keys or to not have
167 encrypted data when traveling to those countries, especially when
168 crossing the borders.
170 Disadvantages are that you do not have all the nice features that
171 the LUKS metadata offers, like multiple passphrases that can be
172 changed, the cipher being stored in the metadata, anti-forensic
173 properties like key-slot diffusion and salts, etc..
175 LUKS format uses a metadata header and 8 key-slot areas that are
176 being placed ath the begining of the disk, see below under "What
177 does the LUKS on-disk format looks like?". The passphrases are used
178 to decryt a single master key that is stored in the anti-forensic
181 Advantages are a higher usability, automatic configuration of
182 non-default crypto parameters, defenses against low-entropy
183 passphrases like salting and iterated PBKDF2 passphrase hashing,
184 the ability to change passhrases, and others.
186 Disadvantages are that it is readily obvious there is encrypted
187 data on disk (but see side note above) and that damage to the
188 header or key-slots usually results in permanent data-loss. See
189 below under "6. Backup and Data Recovery" on how to reduce that
190 risk. Also the sector numbers get shifted by the length of the
191 header and key-slots and there is a loss of that size in capacity
192 (1MB+4096B for defaults and 2MB for the most commonly used
193 non-default XTS mode).
196 * Can I encrypt an already existing, non-empty partition to use LUKS?
198 There is no converter, and it is not really needed. The way to do
199 this is to make a backup of the device in question, securely wipe
200 the device (as LUKS device initialization does not clear away old
201 data), do a luksFormat, optionally overwrite the encrypted device,
202 create a new filesystem and restore your backup on the now
203 encrypted device. Also refer to sections "Security Aspects" and
204 "Backup and Data Recovery".
206 For backup, plain GNU tar works well and backs up anything likely
207 to be in a filesystem.
210 * How do I use LUKS with a loop-device?
212 This can be very handy for experiments. Setup is just the same as
213 with any block device. If you want, for example, to use a 100MiB
214 file as LUKS container, do something like this:
216 head -c 100M /dev/zero > luksfile # create empty file
217 losetup /dev/loop0 luksfile # map luksfile to /dev/loop0
218 cryptsetup luksFormat /dev/loop0 # create LUKS on loop device
220 Afterwards just use /dev/loop0 as a you would use a LUKS partition.
221 To unmap the file when done, use "losetup -d /dev/loop0".
224 * When I add a new key-slot to LUKS, it asks for a passphrase but
225 then complains about there not being a key-slot with that
228 That is as intended. You are asked a passphrase of an existing
229 key-slot first, before you can enter the passphrase for the new
230 key-slot. Otherwise you could break the encryption by just adding a
231 new key-slot. This way, you have to know the passphrase of one of
232 the already configured key-slots in order to be able to configure a
236 * Encrytion on top of RAID or the other way round?
238 Unless you have special needs, place encryption between RAID and
239 filesystem, i.e. encryption on top of RAID. You can do it the other
240 way round, but you have to be aware that you then need to give the
241 pasphrase for each individual disk and RAID autotetection will not
242 work anymore. Therefore it is better to encrypt the RAID device,
246 * How do I read a dm-crypt key from file?
248 Note that the file will still be hashed first, just like keyboard
249 input. Use the --key-file option, like this:
251 cryptsetup create --key-file keyfile e1 /dev/loop0
254 * How do I read a LUKS slot key from file?
256 What you really do here is to read a passphrase from file, just as
257 you would with manual entry of a passphrase for a key-slot. You can
258 add a new passphrase to a free key-slot, set the passphrase of an
259 specific key-slot or put an already configured passphrase into a
260 file. In the last case make sure no trailing newline (0x0a) is
261 contained in the key file, or the passphrase will not work because
262 the whole file is used as input.
264 To add a new passphrase to a free key slot from file, use something
267 cryptsetup luksAddKey /dev/loop0 keyfile
269 To add a new passphrase to a specific key-slot, use something like
272 cryptsetup luksAddKey --key-slot 7 /dev/loop0 keyfile
274 To supply a key from file to any LUKS command, use the --key-file
275 option, e.g. like this:
277 cryptsetup luksOpen --key-file keyfile /dev/loop0 e1
280 * How do I read the LUKS master key from file?
282 The question you should ask yourself first is why you would want to
283 do this. The only legitimate reason I can think of is if you want
284 to have two LUKS devices with the same master key. Even then, I
285 think it would be preferable to just use key-slots with the same
286 passphrase, or to use plain dm-crypt instead. If you really have a
287 good reason, please tell me. If I am convinced, I will add how to
291 * What are the security requirements for a key read from file?
293 A file-stored key or passphrase has the same security requirements
294 as one entered interactively, however you can use random bytes and
295 thereby use bytes you cannot type on the keyboard. You can use any
296 file you like as key file, for example a plain text file with a
297 human readable passphrase. To generate a file with random bytes,
298 use something like this:
300 head -c 256 /dev/random > keyfile
303 * If I map a journaled file system using dm-crypt/LUKS, does it still
304 provide its usual transactional guarantees?
306 As far as I know it does (but I may be wrong), but please note that
307 these "guarantees" are far weaker than they appear to be. For
308 example, you may not get a hard flush to disk surface even on a
309 call to fsync. In addition, the HDD itself may do independent
310 write reordering. Some other things can go wrong as well. The
311 filesystem developers are aware of these problems and typically
312 can make it work anyways. That said, dm-crypt/LUKS should not make
315 Personally, I have several instances of ext3 on dm-crypt and have
316 not noticed any specific problems.
318 Update: I did run into frequent small freezes (1-2 sec) when putting
319 a vmware image on ext3 over dm-crypt. This does indicate that the
320 transactional guarantees are in place, but at a cost. When I went
321 back to ext2, the problem went away. This also seems to have gotten
322 better with kernel 2.6.36 and the reworking of filesystem flush
323 locking. Kernel 2.6.38 is expected to have more improvements here.
326 * Can I use LUKS or cryptsetup with a more secure (external) medium
327 for key storage, e.g. TPM or a smartcard?
329 Yes, see the answers on using a file-supplied key. You do have to
330 write the glue-logic yourself though. Basically you can have
331 cryptsetup read the key from STDIN and write it there with your
332 own tool that in turn gets the key from the more secure key
336 * Can I resize a dm-crypt or LUKS partition?
338 Yes, you can, as neither dm-crypt nor LUKS stores partition size.
339 Whether you should is a different question. Personally I recommend
340 backup, recreation of the encrypted partition with new size,
341 recreation of the filesystem and restore. This gets around the
342 tricky business of resizing the filesystem. Resizing a dm-crypt or
343 LUKS container does not resize the filesystem in it. The backup is
344 really non-optional here, as a lot can go wrong, resulting in
345 partial or complete data loss. Using something like gparted to
346 resize an encrypted partition is slow, but typicaly works. This
347 will not change the size of the filesystem hidden under the
350 You also need to be aware of size-based limitations. The one
351 currently relevant is that aes-xts-plain should not be used for
352 encrypted container sizes larger than 2TiB. Use aes-xts-plain64
359 * My dm-crypt/LUKS mapping does not work! What general steps are
360 there to investigate the problem?
362 If you get a specific error message, investigate what it claims
363 first. If not, you may want to check the following things.
365 - Check that "/dev", including "/dev/mapper/control" is there. If it
366 is missing, you may have a problem with the "/dev" tree itself or
367 you may have broken udev rules.
369 - Check that you have the device mapper and the crypt target in your
370 kernel. The output of "dmsetup targets" should list a "crypt"
371 target. If it is not there or the command fails, add device mapper
372 and crypt-target to the kernel.
374 - Check that the hash-functions and ciphers you want to use are in
375 the kernel. The output of "cat /proc/crypto" needs to list them.
378 * My dm-crypt mapping suddenly stopped when upgrading cryptsetup.
380 The default cipher, hash or mode may have changed (the mode changed
381 from 1.0.x to 1.1.x). See under "Issues With Specific Versions of
385 * When I call cryptsetup from cron/CGI, I get errors about unknown
388 If you get errors about unknown parameters or the like that are not
389 present when cryptsetup is called from the shell, make sure you
390 have no older version of cryptsetup on your system that then gets
391 called by cron/CGI. For example some distributions install
392 cryptsetup into /usr/sbin, while a manual install could go to
393 /usr/local/sbin. As a debugging aid, call "cryptsetup --version"
394 from cron/CGI or the non-shell mechanism to be sure the right
398 * Unlocking a LUKS device takes very long. Why?
400 The iteration time for a key-slot (see Section 5 for an explanation
401 what iteration does) is calculated when setting a passphrase. By
402 default it is 1 second on the machine where the passphrase is set.
403 If you set a passphrase on a fast machine and then unlock it on a
404 slow machine, the unlocking time can be much longer. Also take into
405 account that up to 8 key-slots have to be tried in order to find the
408 If this is problem, you can add another key-slot using the slow
409 machine with the same passphrase and then remove the old key-slot.
410 The new key-slot will have an iteration count adjusted to 1 second
411 on the slow machine. Use luksKeyAdd and then luksKillSlot or
414 However, this operation will not change volume key iteration count
415 (MK iterations in output of "cryptsetup luksDump"). In order to
416 change that, you will have to backup the data in the LUKS
417 container (i.e. your encrypted data), luksFormat on the slow
418 machine and restore the data. Note that in the original LUKS
419 specification this value was fixed to 10, but it is now derived
420 from the PBKDF2 benchmark as well and set to iterations in 0.125
421 sec or 1000, whichever is larger. Also note that MK iterations
422 are not very security relevant. But as each key-slot already takes
423 1 second, spending the additional 0.125 seconds really does not
427 * "blkid" sees a LUKS UUID and an ext2/swap UUID on the same device.
430 Some old versions of cryptsetup have a bug where the header does
431 not get completely wiped during LUKS format and an older ext2/swap
432 signature remains on the device. This confuses blkid.
434 Fix: Wipe the unused header areas by doing a backup and restore of
435 the header with cryptsetup 1.1.x:
437 cryptsetup luksHeaderBackup --header-backup-file <file> <device>
438 cryptsetup luksHeaderRestore --header-backup-file <file> <device>
441 * cryptsetup segfaults on Gentoo amd64 hardened ...
443 There seems to be some inteference between the hardening and and
444 the way cryptsetup benchmarks PBKDF2. The solution to this is
445 currently not quite clear for an encrypted root filesystem. For
446 other uses, you can apparently specify USE="dynamic" as compile
447 flag, see http://bugs.gentoo.org/show_bug.cgi?id=283470
453 * I get the error "LUKS keyslot x is invalid." What does that mean?
455 This means that the given keyslot has an offset that points
456 outside the valid keyslot area. Typically, the reason is a
457 corrupted LUKS header because something was written to the start of
458 the device the LUKS contaner is on. Refer to Section "Backup and
459 Data Recovery" and ask on the mailing list if you have trouble
460 diagnosing and (if still possible) repairing this.
463 * Can a bad RAM module cause problems?
465 LUKS and dm-crypt can give the RAM quite a workout, especially when
466 combined with software RAID. In particular the combination RAID5 +
467 LUKS + XFS seems to uncover RAM problems that never caused obvious
468 problems before. Symptoms vary, but often the problem manifest
469 itself when copying large amounts of data, typically several times
470 larger than your main memory.
472 Side note: One thing you should always do on large data
473 copy/movements is to run a verify, for example with the "-d"
474 option of "tar" or by doing a set of MD5 checksums on the source
477 find . -type f -exec md5sum \{\} \; > checksum-file
479 and then a "md5sum -c checksum-file" on the other side. If you get
480 mismatches here, RAM is the primary suspect. A lesser suspect is
481 an overclocked CPU. I have found countless hardware problems in
482 verify runs after copying or making backups. Bit errors are much
483 more common than most people think.
485 Some RAM issues are even worse and corrupt structures in one of the
486 layers. This typically results in lockups, CPU state dumps in the
487 system logs, kernel panic or other things. It is quite possible to
488 have the problem with an encrypted device, but not with an
489 otherwise the same unencrypted device. The reason for that is that
490 encryption has an error amplification property: You flip one bit
491 in an encrypted data block, and the decrypted version has half of
492 its bits flipped. This is an important security property for modern
493 ciphers. With the usual modes in cryptsetup (CBC, ESSIV, XTS), you
494 get up to a completely changed 512 byte block per bit error. A
495 corrupt block causes a lot more havoc than the occasionally
496 flipped single bit and can result in various obscure errors.
498 Note, that a verify run on copying between encrypted or
499 unencrypted devices will reliably detect corruption, even when the
500 copying itself did not report any problems. If you find defect
501 RAM, assume all backups and copied data to be suspect, unless you
507 First you should know that overclocking often makes memory
508 problems worse. So if you overclock (which I strongly recommend
509 against in a system holding data that has some worth), run the
510 tests with the overclocking active.
512 There are two good options. One is Memtest86+ and the other is
513 "memtester" by Charles Cazabon. Memtest86+ requires a reboot and
514 then takes over the machine, while memtester runs from a
515 root-shell. Both use different testing methods and I have found
516 problems fast with each one that the other needed long to find. I
517 recommend running the following procedure until the first error is
520 - Run Memtest86+ for one cycle
522 - Run memterster for one cycle (shut down as many other applications
525 - Run Memtest86+ for 24h or more
527 - Run memtester for 24h or more
529 If all that does not produce error messages, your RAM may be sound,
530 but I have had one weak bit that Memtest86+ needed around 60 hours
531 to find. If you can reproduce the original problem reliably, a good
532 additional test may be to remove half of the RAM (if you have more
533 than one module) and try whether the problem is still there and if
534 so, try with the other half. If you just have one module, get a
535 different one and try with that. If you do overclocking, reduce
536 the settings to the most conservative ones available and try with
543 * Is LUKS insecure? Everybody can see I have encrypted data!
545 In practice it does not really matter. In most civilized countries
546 you can just refuse to hand over the keys, no harm done. In some
547 countries they can force you to hand over the keys, if they suspect
548 encryption. However the suspicion is enough, they do not have to
549 prove anything. This is for practical reasons, as even the presence
550 of a header (like the LUKS header) is not enough to prove that you
551 have any keys. It might have been an experiment, for example. Or it
552 was used as encrypted swap with a key from /dev/random. So they
553 make you prove you do not have encrypted data. Of course that is
554 just as impossible as the other way round.
556 This means that if you have a large set of random-looking data,
557 they can already lock you up. Hidden containers (encryption hidden
558 within encryption), as possible with Truecrypt, do not help
559 either. They will just assume the hidden container is there and
560 unless you hand over the key, you will stay locked up. Don't have
561 a hidden container? Though luck. Anybody could claim that.
563 Still, if you are concerned about the LUKS header, use plain
564 dm-crypt with a good passphrase. See also Section 2, "What is the
565 difference between "plain" and LUKS format?"
568 * Should I initialize (overwrite) a new LUKS/dm-crypt partition?
570 If you just create a filesystem on it, most of the old data will
571 still be there. If the old data is sensitive, you should overwrite
572 it before encrypting. In any case, not initializing will leave the
573 old data there until the specific sector gets written. That may
574 enable an attacker to determine how much and where on the
575 partition data was written. If you think this is a risk, you can
576 prevent this by overwriting the encrypted device (here assumed to
577 be named "e1") with zeros like this:
579 dd_rescue -w /dev/zero /dev/mapper/e1
581 or alternatively with one of the following more standard commands:
583 cat /dev/zero > /dev/mapper/e1
584 dd if=/dev/zero of=/dev/mapper/e1
587 * How do I securely erase a LUKS (or other) partition?
589 For LUKS, if you are in a desperate hurry, overwrite the LUKS
590 header and key-slot area. This means overwriting the first
591 (keyslots x stripes x keysize) + offset bytes. For the default
592 parameters, this is the 1'052'672 bytes, i.e. 1MiB + 4096 of the
593 LUKS partition. For 512 bit key length (e.g. for aes-xts-plain with
594 512 bit key) this is 2MiB. (The diferent offset stems from
595 differences in the sector alignment of the key-slots.) If in doubt,
596 just be generous and overwrite the first 10MB or so, it will likely
597 still be fast enough. A single overwrite with zeros should be
598 enough. If you anticipate being in a desperate hurry, prepare the
599 command beforehand. Example with /dev/sde1 as the LUKS partition
600 and default parameters:
602 head -c 1052672 /dev/zero > /dev/sde1; sync
604 A LUKS header backup or full backup will still grant access to
605 most or all data, so make sure that an attacker does not have
606 access to backups or destroy them as well.
608 If you have time, overwrite the whole LUKS partition with a single
609 pass of zeros. This is enough for current HDDs. For SSDs or FLASH
610 (USB sticks) you may want to overwrite the whole drive several
611 times to be sure data is not retained by wear leveling. This is
612 possibly still insecure as SSD technology is not fully understood
613 in this regard. Still, due to the anti-forensic properties of the
614 LUKS key-slots, a single overwrite of an SSD or FLASH drive could
615 be enough. If in doubt, use physical destruction in addition. Here
616 is a link to some current reseach results on erasing SSDs and FLASH
618 http://www.usenix.org/events/fast11/tech/full_papers/Wei.pdf
620 Keep in mind to also erase all backups.
622 Example for a zero-overwrite erase of partition sde1 done with
625 dd_rescue -w /dev/zero /dev/sde1
628 * How do I securely erase a backup of a LUKS partition or header?
630 That depends on the medium it is stored on. For HDD and SSD, use
631 overwrite with zeros. For an SSD or FLASH drive (USB stick), you
632 may want to overwrite the complete SSD several times and use
633 physical destruction in addition, see last item. For re-writable
634 CD/DVD, a single overwrite should also be enough, due to the
635 anti-forensic properties of the LUKS keyslots. For write-once
636 media, use physical destruction. For low security requirements,
637 just cut the CD/DVD into several parts. For high security needs,
638 shred or burn the medium. If your backup is on magnetic tape, I
639 advise physical destruction by shredding or burning, after
640 overwriting . The problem with magnetic tape is that it has a
641 higher dynamic range than HDDs and older data may well be
642 recoverable after overwrites. Also write-head alignment issues can
643 lead to data not actually being deleted at all during overwrites.
646 * What about backup? Does it compromise security?
648 That depends. See next section.
651 * Why is all my data permanently gone if I overwrite the LUKS header?
653 Overwriting the LUKS header in part or in full is the most common
654 reason why access to LUKS containers is lost permanently.
655 Overwriting can be done in a number of fashions, like creating a
656 new filesystem on the raw LUKS partition, making the raw partition
657 part of a raid array and just writing to the raw partition.
659 The LUKS header contains a 256 bit "salt" value and without that no
660 decryption is possible. While the salt is not secret, it is
661 key-grade material and cannot be reconstructed. This is a
662 cryptographically strong "cannot". From observations on the
663 cryptsetup mailing-list, people typically go though the usual
664 stages of grief (Denial, Anger, Bargaining, Depression, Acceptance)
665 when this happens to them. Observed times vary between 1 day and 2
666 weeks to complete the cycle. Seeking help on the mailing-list is
667 fine. Even if we usually cannot help with getting back your data,
668 most people found the feedback comforting.
670 If your header does not contain an intact salt, best go directly
671 to the last stage ("Acceptance") and think about what to do now.
672 There is one exception that I know of: If your LUKS container is
673 still open, then it may be possible to extract the master key from
674 the running system. See Item "How do I recover the master key from
675 a mapped LUKS container?" in Section "Backup and Data Recovery".
680 A salt is a random key-grade value added to the passphrase before
681 it is processed. It is not kept secret. The reason for using salts
682 is as follows: If an attacker wants to crack the password for a
683 single LUKS container, then every possible passphrase has to be
684 tried. Typically an attacker will not try every binary value, but
685 will try words and sentences from a dictionary.
687 If an attacker wants to attack several LUKS containers with the
688 same dictionary, then a different approach makes sense: Compute the
689 resulting slot-key for each dictionary element and store it on
690 disk. Then the test for each entry is just the slow unlocking with
691 the slot key (say 0.00001 sec) instead of calculating the slot-key
692 first (1 sec). For a single attack, this does not help. But if you
693 have more than one container to attack, this helps tremendously,
694 also because you can prepare your table before you even have the
695 container to attack! The calculation is also very simple to
696 parallelize. You could, for example, use the night-time unused CPU
697 power of your desktop PCs for this.
699 This is where the salt comes in. If the salt is combined with the
700 passphrase (in the simplest form, just appended to it), you
701 suddenly need a separate table for each salt value. With a
702 reasonably-sized salt value (256 bit, e.g.) this is quite
706 * Is LUKS secure with a low-entropy (bad) passphrase?
708 Note: You should only use the 94 printable characters from 7 bit
709 ASCII code to prevent your passphrase from failing when the
710 character encoding changes, e.g. because of a system upgrade, see
711 also the note at the very start of this FAQ under "WARNINGS".
713 This needs a bit of theory. The quality of your passphrase is
714 directly related to its entropy (information theoretic, not
715 thermodynamic). The entropy says how many bits of "uncertainty" or
716 "randomness" are in you passphrase. In other words, that is how
717 difficult guessing the passphrase is.
719 Example: A random English sentence has about 1 bit of entropy per
720 character. A random lowercase (or uppercase) character has about
723 Now, if n is the number of bits of entropy in your passphrase and t
724 is the time it takes to process a passphrase in order to open the
725 LUKS container, then an attacker has to spend at maximum
727 attack_time_max = 2^n * t
729 time for a successful attack and on average half that. There is no
730 way getting around that relationship. However, there is one thing
731 that does help, namely increasing t, the time it takes to use a
732 passphrase, see next FAQ item.
734 Still, if you want good security, a high-entropy passphrase is the
735 only option. Use at least 64 bits for secret stuff. That is 64
736 characters of English text (but only if randomly chosen) or a
737 combination of 12 truly random letters and digits.
739 For passphrase generation, do not use lines from very well-known
740 texts (religious texts, Harry potter, etc.) as they are to easy to
741 guess. For example, the total Harry Potter has about 1'500'000
742 words (my estimation). Trying every 64 character sequence starting
743 and ending at a word boundary would take only something like 20
744 days on a single CPU and is entirely feasible. To put that into
745 perspective, using a number of Amazon EC2 High-CPU Extra Large
746 instances (each gives about 8 real cores), this tests costs
747 currently about 50USD/EUR, but can be made to run arbitrarily fast.
749 On the other hand, choosing 1.5 lines from, say, the Wheel of Time
750 is in itself not more secure, but the book selection adds quite a
751 bit of entropy. (Now that I have mentioned it here, don't use tWoT
752 either!) If you add 2 or 3 typos or switch some words around, then
753 this is good passphrase material.
756 * What is "iteration count" and why is decreasing it a bad idea?
758 Iteration count is the number of PBKDF2 iterations a passphrase is
759 put through before it is used to unlock a key-slot. Iterations are
760 done with the explicit purpose to increase the time that it takes
761 to unlock a key-slot. This provides some protection against use of
762 low-entropy passphrases.
764 The idea is that an attacker has to try all possible passphrases.
765 Even if the attacker knows the passphrase is low-entropy (see last
766 item), it is possible to make each individual try take longer. The
767 way to do this is to repeatedly hash the passphrase for a certain
768 time. The attacker then has to spend the same time (given the same
769 computing power) as the user per try. With LUKS, the default is 1
770 second of PBKDF2 hashing.
772 Example 1: Lets assume we have a really bad passphrase (e.g. a
773 girlfriends name) with 10 bits of entropy. With the same CPU, an
774 attacker would need to spend around 500 seconds on average to
775 break that passphrase. Without iteration, it would be more like
776 0.0001 seconds on a modern CPU.
778 Example 2: The user did a bit better and has 32 chars of English
779 text. That would be about 32 bits of entropy. With 1 second
780 iteration, that means an attacker on the same CPU needs around 136
781 years. That is pretty impressive for such a weak passphrase.
782 Without the iterations, it would be more like 50 days on a modern
783 CPU, and possibly far less.
785 In addition, the attacker can both parallelize and use special
786 hardware like GPUs to speed up the attack. The attack can also
787 happen quite some time after the luksFormat operation and CPUs can
788 have become faster and cheaper. For that reason you want a bit
789 of extra security. Anyways, in Example 1 your are screwed. In
790 example 2, not necessarily. Even if the attack is faster, it still
791 has a certain cost associated with it, say 10000 EUR/USD with
792 iteration and 1 EUR/USD without iteration. The first can be
793 prohibitively expensive, while the second is something you try
794 even without solid proof that the decryption will yield something
797 The numbers above are mostly made up, but show the idea. Of course
798 the best thing is to have a high-entropy passphrase.
800 Would a 100 sec iteration time be even better? Yes and no.
801 Cryptographically it would be a lot better, namely 100 times better.
802 However, usability is a very important factor for security
803 technology and one that gets overlooked surprisingly often. For
804 LUKS, if you have to wait 2 minutes to unlock the LUKS container,
805 most people will not bother and use less secure storage instead. It
806 is better to have less protection against low-entropy passphrases
807 and people actually use LUKS, than having them do without
808 encryption altogether.
810 Now, what about decreasing the iteration time? This is generally a
811 very bad idea, unless you know and can enforce that the users only
812 use high-entropy passphrases. If you decrease the iteration time
813 without ensuring that, then you put your users at increased risk,
814 and considering how rarely LUKS containers are unlocked in a
815 typical work-flow, you do so without a good reason. Don't do it.
816 The iteration time is already low enough that users with entropy
817 low passphrases are vulnerable. Lowering it even further increases
818 this danger significantly.
821 * What about iteration count with plain dm-crypt?
823 Simple: There is none. There is also no salting. If you use plain
824 dm-crypt, the only way to be secure is to use a high entropy
825 passphrase. If in doubt, use LUKS instead.
828 * Is LUKS with default parameters less secure on a slow CPU?
830 Unfortunately, yes. However the only aspect affected is the
831 protection for low-entropy passphrase or master-key. All other
832 security aspects are independent of CPU speed.
834 The master key is less critical, as you really have to work at it
835 to give it low entropy. One possibility is to supply the master key
836 yourself. If that key is low-entropy, then you get what you
837 deserve. The other known possibility is to use /dev/urandom for
838 key generation in an entropy-startved situation (e.g. automatic
839 installation on an embedded device without network and other entropy
842 For the passphrase, don't use a low-entropy passphrase. If your
843 passphrase is good, then a slow CPU will not matter. If you insist
844 on a low-entropy passphrase on a slow CPU, use something like
845 "--iter-time=10" or higher and wait a long time on each LUKS unlock
846 and pray that the attacker does not find out in which way exactly
847 your passphrase is low entropy. This also applies to low-entropy
848 passphrases on fast CPUs. Technology can do only so much to
849 compensate for problems in front of the keyboard.
852 * Why was the default aes-cbc-plain replaced with aes-cbc-essiv?
854 The problem is that cbc-plain has a fingerprint vulnerability, where
855 a specially crafted file placed into the crypto-container can be
856 recognized from the outside. The issue here is that for cbc-plain
857 the initialization vector (IV) is the sector number. The IV gets
858 XORed to the first data chunk of the sector to be encrypted. If you
859 make sure that the first data block to be stored in a sector
860 contains the sector number as well, the first data block to be
861 encrypted is all zeros and always encrypted to the same ciphertext.
862 This also works if the first data chunk just has a constant XOR
863 with the sector number. By having several shifted patterns you can
864 take care of the case of a non-power-of-two start sector number of
867 This mechanism allows you to create a pattern of sectors that have
868 the same first ciphertext block and signal one bit per sector to the
869 outside, allowing you to e.g. mark media files that way for
870 recognition without decryption. For large files this is a
871 practical attack. For small ones, you do not have enough blocks to
872 signal and take care of different file starting offsets.
874 In order to prevent this attack, the default was changed to
875 cbc-essiv. ESSIV uses a keyed hash of the sector number, with the
876 encryption key as key. This makes the IV unpredictable without
877 knowing the encryption key and the watermarking attack fails.
880 * Are there any problems with "plain" IV? What is "plain64"?
882 First, "plain" and "plain64" are both not secure to use with CBC,
883 see previous FAQ item.
885 However there are modes, like XTS, that are secure with "plain" IV.
886 The next limit is that "plain" is 64 bit, with the upper 32 bit set
887 to zero. This means that on volumes larger than 2TiB, the IV
888 repeats, creating a vulnerability that potentially leaks some
889 data. To avoid this, use "plain64", which uses the full sector
890 number up to 64 bit. Note that "plain64" requires a kernel >=
891 2.6.33. Also note that "plain64" is backwards compatible for
892 volume sizes <= 2TiB, but not for those > 2TiB. Finally, "plain64"
893 does not cause any performance penalty compared to "plain".
896 * What about XTS mode?
898 XTS mode is potentially even more secure than cbc-essiv (but only if
899 cbc-essiv is insecure in your scenario). It is a NIST standard and
900 used, e.g. in Truecrypt. At the moment, if you want to use it, you
901 have to specify it manually as "aes-xts-plain", i.e.
903 cryptsetup -c aes-xts-plain luksFormat <device>
905 For volumes >2TiB and kernels >= 2.6.33 use "plain64" (see FAQ
906 item on "plain" and "plain64"):
908 cryptsetup -c aes-xts-plain64 luksFormat <device>
910 There is a potential security issue with XTS mode and large blocks.
911 LUKS and dm-crypt always use 512B blocks and the issue does not
915 6. Backup and Data Recovery
918 * Why do I need Backup?
920 First, disks die. The rate for well-treated (!) disk is about 5%
921 per year, which is high enough to worry about. There is some
922 indication that this may be even worse for some SSDs. This applies
923 both to LUKS and plain dm-crypt partitions.
925 Second, for LUKS, if anything damages the LUKS header or the
926 key-stripe area then decrypting the LUKS device can become
927 impossible. This is a frequent occuurence. For example an
928 accidental format as FAT or some software overwriting the first
929 sector where it suspects a partition boot sector typically makes a
930 LUKS partition permanently inacessible. See more below on LUKS
933 So, data-backup in some form is non-optional. For LUKS, you may
934 also want to store a header backup in some secure location. This
935 only needs an update if you change passphrases.
938 * How do I backup a LUKS header?
940 While you could just copy the appropriate number of bytes from the
941 start of the LUKS partition, the best way is to use command option
942 "luksHeaderBackup" of cryptsetup. This protects also against
943 errors when non-standard parameters have been used in LUKS
944 partition creation. Example:
947 cryptsetup luksHeaderBackup --header-backup-file h /dev/mapper/c1
949 To restore, use the inverse command, i.e.
951 cryptsetup luksHeaderRestore --header-backup-file h /dev/mapper/c1
954 * How do I backup a LUKS or dm-crypt partition?
956 There are two options, a sector-image and a plain file or
957 filesystem backup of the contents of the partition. The sector
958 image is already encrypted, but cannot be compressed and contains
959 all empty space. The filesystem backup can be compressed, can
960 contain only part of the encrypted device, but needs to be
961 encrypted separately if so desired.
963 A sector-image will contain the whole partition in encrypted form,
964 for LUKS the LUKS header, the keys-slots and the data area. It can
965 be done under Linux e.g. with dd_rescue (for a direct image copy)
966 and with "cat" or "dd". Example:
968 cat /dev/sda10 > sda10.img
969 dd_rescue /dev/sda10 sda10.img
971 You can also use any other backup software that is capable of making
972 a sector image of a partition. Note that compression is
973 ineffective for encrypted data, hence it does not make sense to
976 For a filesystem backup, you decrypt and mount the encrypted
977 partition and back it up as you would a normal filesystem. In this
978 case the backup is not encrypted, unless your encryption method
979 does that. For example you can encrypt a backup with "tar" as
982 tar cjf - <path> | gpg --cipher-algo AES -c - > backup.tbz2.gpg
984 And verify the backup like this if you are at "path":
986 cat backup.tbz2.gpg | gpg - | tar djf -
988 Note: Allways verify backups, especially encrypted ones.
990 In both cases GnuPG will ask you interactively for your symmetric
991 key. The verify will only output errors. Use "tar dvjf -" to get
992 all comparison results. To make sure no data is written to disk
993 unencrypted, turn off swap if it is not encrypted before doing the
996 You can of course use different or no compression and you can use
997 an asymmetric key if you have one and have a backup of the secret
998 key that belongs to it.
1000 A second option for a filestem-level backup that can be used when
1001 the backup is also on local disk (e.g. an external USB drive) is
1002 to use a LUKS container there and copy the files to be backed up
1003 between both mounted containers. Also see next item.
1006 * Do I need a backup of the full partition? Would the header and
1007 key-slots not be enough?
1009 Backup protects you against two things: Disk loss or corruption
1010 and user error. By far the most questions on the dm-crypt mailing
1011 list about how to recover a damaged LUKS partition are related
1012 to user error. For example, if you create a new filesystem on a
1013 LUKS partition, chances are good that all data is lost
1016 For this case, a header+key-slot backup would often be enough. But
1017 keep in mind that a well-treated (!) HDD has roughly a failure
1018 risk of 5% per year. It is highly advisable to have a complete
1019 backup to protect against this case.
1022 * *What do I need to backup if I use "decrypt_derived"?
1024 This is a script in Debian, intended for mounting /tmp or swap with
1025 a key derived from the master key of an already decrypted device.
1026 If you use this for an device with data that should be persistent,
1027 you need to make sure you either do not lose access to that master
1028 key or have a backup of the data. If you derive from a LUKS
1029 device, a header backup of that device would cover backing up the
1030 master key. Keep in mind that this does not protect against disk
1033 Note: If you recreate the LUKS header of the device you derive from
1034 (using luksFormat), the master key changes even if you use the same
1035 passphrase(s) and you will not be able to decrypt the derived
1036 device with the new LUKS header.
1039 * Does a backup compromise security?
1041 Depends on how you do it. However if you do not have one, you are
1042 going to eventually lose your encrypted data.
1044 There are risks introduced by backups. For example if you
1045 change/disable a key-slot in LUKS, a binary backup of the partition
1046 will still have the old key-slot. To deal with this, you have to
1047 be able to change the key-slot on the backup as well, securely
1048 erase the backup or do a filesystem-level backup instead of a binary
1051 If you use dm-crypt, backup is simpler: As there is no key
1052 management, the main risk is that you cannot wipe the backup when
1053 wiping the original. However wiping the original for dm-crypt
1054 should consist of forgetting the passphrase and that you can do
1055 without actual access to the backup.
1057 In both cases, there is an additional (usually small) risk with
1058 binary backups: An attacker can see how many sectors and which
1059 ones have been changed since the backup. To prevent this, use a
1060 filesystem level backup methid that encrypts the whole backup in
1061 one go, e.g. as described above with tar and GnuPG.
1063 My personal advice is to use one USB disk (low value data) or
1064 three disks (high value data) in rotating order for backups, and
1065 either use independent LUKS partitions on them, or use encrypted
1066 backup with tar and GnuPG.
1068 If you do network-backup or tape-backup, I strongly recommend to
1069 go the filesystem backup path with independent encryption, as you
1070 typically cannot reliably delete data in these scenarios,
1071 especially in a cloud setting. (Well, you can burn the tape if it
1072 is under your control...)
1075 * What happens if I overwrite the start of a LUKS partition or damage
1076 the LUKS header or key-slots?
1078 There are two critical components for decryption: The salt values
1079 in the header itself and the key-slots. If the salt values are
1080 overwritten or changed, nothing (in the cryptographically strong
1081 sense) can be done to access the data, unless there is a backup
1082 of the LUKS header. If a key-slot is damaged, the data can still
1083 be read with a different key-slot, if there is a remaining
1084 undamaged and used key-slot. Note that in order to make a key-slot
1085 unrecoverable in a cryptographically strong sense, changing about
1086 4-6 bits in random locations of its 128kiB size is quite enough.
1089 * What happens if I (quick) format a LUKS partition?
1091 I have not tried the different ways to do this, but very likely you
1092 will have written a new boot-sector, which in turn overwrites the
1093 LUKS header, including the salts, making your data permanently
1094 irretrivable, unless you have a LUKS header backup. You may also
1095 damage the key-slots in part or in full. See also last item.
1098 * How do I recover the master key from a mapped LUKS container?
1100 This is typically only needed if you managed to damage your LUKS
1101 header, but the container is still mapped, i.e. "luksOpen"ed. It
1102 also helps if you have a mapped container that you forgot or do not
1103 know a passphrase for (e.g. on a long running server.)
1105 WARNING: Things go wrong, do a full backup before trying this!
1107 WARNING: This exposes the master key of the LUKS container. Note
1108 that both ways to recreate a LUKS header with the old master key
1109 described below will write the master key to disk. Unless you are
1110 sure you have securely erased it afterwards, e.g. by writing it to
1111 an encrypted partition, RAM disk or by erasing the filesystem you
1112 wrote it to by a complete overwrite, you should change the master
1113 key afterwards. Changing the master key requires a full data
1114 backup, luksFormat and then restore of the backup.
1116 First, there is a script by Milan that automatizes the whole
1117 process, except generating a new LUKS header with the old master
1118 key (it prints the command for that though):
1120 http://code.google.com/p/cryptsetup/source/browse/trunk/misc/luks-header-from-active
1122 You can also do this manually. Here is how:
1124 - Get the master key from the device mapper. This is done by the
1125 following command. Substitute c5 for whatever you mapped to:
1127 # dmsetup table --target crypt --showkey /dev/mapper/c5
1129 0 200704 crypt aes-cbc-essiv:sha256
1130 a1704d9715f73a1bb4db581dcacadaf405e700d591e93e2eaade13ba653d0d09
1133 The result is actually one line, wrapped here for clarity. The long
1134 hex string is the master key.
1136 - Convert the master key to a binary file representation. You can
1137 do this manually, e.g. with hexedit. You can also use the tool
1138 "xxd" from vim like this:
1140 echo "a1704d9....53d0d09" | xxd -r -p > <master-key-file>
1142 - Do a luksFormat to create a new LUKS header.
1144 NOTE: If your header is intact and you just forgot the
1145 passphrase, you can just set a new passphrase, see next subitem.
1147 Unmap the device before you do that (luksClose). Then do
1149 cryptsetup luksFormat --master-key-file=<master-key-file> <luks device>
1151 Note that if the container was created with other than the default
1152 settings of the cryptsetup version you are using, you need to give
1153 additional parameters specifying the deviations. If in doubt, try
1154 the script by Milan. It does recover the other parameters as well.
1156 Side note: This is the way the decrypt_derived script gets at the
1157 master key. It just omits the conversion and hashes the master key
1160 - If the header is intact and you just forgot the passphrase, just
1161 set a new passphrase like this:
1163 cryptsetup luksAddKey --master-key-file=<master-key-file> <luks device>
1165 You may want to disable the old one afterwards.
1168 * What does the on-disk structure of dm-crypt look like?
1170 There is none. dm-crypt takes a block device and gives encrypted
1171 access to each of its blocks with a key derived from the passphrase
1172 given. If you use a cipher different than the default, you have to
1173 specify that as a parameter to cryptsetup too. If you want to
1174 change the password, you basically have to create a second
1175 encrypted device with the new passphrase and copy your data over.
1176 On the plus side, if you accidentally overwrite any part of a
1177 dm-crypt device, the damage will be limited to the are you
1181 * What does the on-disk structure of LUKS look like?
1183 A LUKS partition consists of a header, followed by 8 key-slot
1184 descriptors, followed by 8 key slots, followed by the encrypted
1187 Header and key-slot descriptors fill the first 592 bytes. The
1188 key-slot size depends on the creation parameters, namely on the
1189 number of anti-forensic stripes, key material offset and master
1192 With the default parameters, each key-slot is a bit less than
1193 128kiB in size. Due to sector alignment of the key-slot start,
1194 that means the key block 0 is at offset 0x1000-0x20400, key
1195 block 1 at offset 0x21000-0x40400, and key block 7 at offset
1196 0xc1000-0xe0400. The space to the next full sector address is
1197 padded with zeros. Never used key-slots are filled with what the
1198 disk originally contained there, a key-slot removed with
1199 "luksRemoveKey" or "luksKillSlot" gets filled with 0xff. Due to
1200 2MiB default alignment, start of the data area for cryptsetup 1.3
1201 and later is at 2MiB, i.e. at 0x200000. For older versions, it is
1202 at 0x101000, i.e. at 1'052'672 bytes, i.e. at 1MiB + 4096 bytes
1203 from the start of the partition. Incidentally, "luksHeaderBackup"
1204 for a LUKS container created with default parameters dumps exactly
1205 the first 2MiB (or 1'052'672 bytes for headers created with
1206 cryptsetup versions < 1.3) to file and "luksHeaderRestore" restores
1209 For non-default parameters, you have to figure out placement
1210 yourself. "luksDump" helps. See also next item. For the most common
1211 non-default settings, namely aes-xts-plain with 512 bit key, the
1212 offsets are: 1st keyslot 0x1000-0x3f800, 2nd keyslot
1213 0x40000-0x7e000, 3rd keyslot 0x7e000-0xbd800, ..., and start of
1214 bulk data at 0x200000.
1216 The exact specification of the format is here:
1217 http://code.google.com/p/cryptsetup/wiki/Specification
1220 * What is the smallest possible LUKS container?
1222 Note: From cryptsetup 1.3 onwards, alignment is set to 1MB. With
1223 modern Linux partitioning tools that also align to 1MB, this will
1224 result in aligmnet to 2k secors and typical Flash/SSD sectors,
1225 which is highly desirable for a number of reasons. Changing the
1226 alignment is not recomended.
1228 That said, with default parameters, the data area starts at
1229 exactly 2MB offset (at 0x101000 for crptsetup versions before 1.3).
1230 The smallest data area you can have is one sector of 512 bytes.
1231 Data areas of 0 bytes can be created, but fail on mapping.
1233 While you cannot put a filesystem into something this small, it may
1234 still be used to contain, for eamcple, key. Note that with current
1235 formatting tools, a partition for a container this size will be
1236 3MiB anyways. If you put the LUKS container into a file (via
1237 losetup and a loopback device), the file needs to be 2097664 bytes
1238 in size, i.e. 2MiB + 512B.
1240 The two ways to influence the start of the data area are key-size
1243 For alignment, you can go down to 1 on the parameter. This will
1244 still leave you with a data-area starting at 0x101000, i.e.
1245 1MiB+4096B (default parameters) as alignment will be rounded up to
1246 the next multiple of 8 (i.e. 4096 bytes) (TODO: need to verify
1249 For key-size, you can use 128 bit (e.g. AES-128 with CBC), 256 bit
1250 (e.g. AES-256 with CBC) or 512 bit (e.g. AES-256 with XTS mode).
1251 You can do 64 bit (e.g. blofish-64 with CBC), but anything below
1252 128 bit has to be considered insecure today.
1254 Example 1 - AES 128 bit with CBC:
1256 cryptsetup luksFormat -s 128 --align-payload=8 <device>
1258 This results in a data offset of 0x81000, i.e. 516KiB or 528384
1259 bytes. Add one 512 byte sector and the smallest LUKS container size
1260 with these parameters is 516KiB + 512B or 528896 bytes.
1262 Example 2 - Blowfish 64 bit with CBC (WARNING: insecure):
1264 cryptsetup luksFormat -c blowfish -s 64 --align-payload=8 /dev/loop0
1266 This results in a data offset of 0x41000, i.e. 260kiB or 266240
1267 bytes, with a minimal LUKS conatiner size of 260kiB + 512B or
1271 * I think this is overly complicated. Is there an alternative?
1273 Not really. Encryption comes at a price. You can use plain
1274 dm-crypt to simplify things a bit. It does not allow multiple
1275 passphrases, but on the plus side, it has zero on disk description
1276 and if you overwrite some part of a plain dm-crypt partition,
1277 exactly the overwritten parts are lost (rounded up to sector
1281 7. Interoperability with other Disk Encryption Tools
1284 * What is this section about?
1286 Cryptsetup for plain dm-crypt can be used to access a number of
1287 on-disk formats created by tools like loop-aes patched into
1288 losetup. This somtimes works and sometimes does not. This section
1289 collects insights into what works, what does not and where more
1290 information is required.
1292 Additional information may be found in the mailing-list archives,
1293 mentioned at the start of this FAQ document. If you have a
1294 solution working that is not yet documented here and think a wider
1295 audience may be intertested, please email the FAQ maintainer.
1298 * loop-aes: General observations.
1300 One problem is that there are different versions of losetup around.
1301 loop-aes is a patch for losetup. Possible problems and deviations
1302 from cryptsetup option syntax include:
1304 - Offsets specifed in bytes (cryptsetup: 512 byte sectors)
1306 - The need to specify an IV offset
1308 - Encryption mode needs specifying (e.g. "-c twofish-cbc-plain")
1310 - Key size needs specifying (e.g. "-s 128" for 128 bit keys)
1312 - Passphrase hash algorithm needs specifying
1314 Also note that because plain dm-crypt and loop-aes format does not
1315 have metadata, autodetection, while feasible in most cases, would
1316 be a lot of work that nobody really wants to do. If you still have
1317 the old set-up, using a verbosity option (-v) on mapping with the
1318 old tool or having a look into the system logs after setup could
1319 give you the information you need.
1322 * loop-aes patched into losetup on debian 5.x, kernel 2.6.32
1324 In this case, the main problem seems to be that this variant of
1325 losetup takes the offset (-o option) in bytes, while cryptsetup
1326 takes it in sectors of 512 bytes each. Example: The losetupp
1329 losetup -e twofish -o 2560 /dev/loop0 /dev/sdb1
1330 mount /dev/loop0 mountpoint
1334 cryptsetup create -c twofish -o 5 --skip 5 e1 /dev/sdb1
1335 mount /dev/mapper/e1 mountpoint
1338 * loop-aes with 160 bit key
1340 This seems to be sometimes used with twofish and blowfish and
1341 represents a 160 bit ripemed160 hash output padded to 196 bit key
1342 length. It seems the corresponding options for cryptsetup are
1344 --cipher twofish-cbc-null -s 192 -h ripemd160:20
1347 8. Issues with Specific Versions of cryptsetup
1350 * When using the create command for plain dm-crypt with cryptsetup
1351 1.1.x, the mapping is incompatible and my data is not accessible
1354 With cryptsetup 1.1.x, the distro maintainer can define different
1355 default encryption modes for LUKS and plain devices. You can check
1356 these compiled-in defaults using "cryptsetup --help". Moreover, the
1357 plain device default changed because the old IV mode was
1358 vulnerable to a watermarking attack.
1360 If you are using a plain device and you need a compatible mode, just
1361 specify cipher, key size and hash algorithm explicitly. For
1362 compatibility with cryptsetup 1.0.x defaults, simple use the
1365 cryptsetup create -c aes-cbc-plain -s 256 -h ripemd160 <name> <dev>
1367 LUKS stores cipher and mode in the metadata on disk, avoiding this
1371 * cryptsetup on SLED 10 has problems...
1373 SLED 10 is missing an essential kernel patch for dm-crypt, which
1374 is broken in its kernel as a result. There may be a very old
1375 version of cryptsetup (1.0.x) provided by SLED, which should also
1376 not be used anymore as well. My advice would be to drop SLED 10.
1378 A. Contributors In no particular order: