1 The dm-integrity target emulates a block device that has additional
2 per-sector tags that can be used for storing integrity information.
4 A general problem with storing integrity tags with every sector is that
5 writing the sector and the integrity tag must be atomic - i.e. in case of
6 crash, either both sector and integrity tag or none of them is written.
8 To guarantee write atomicity, the dm-integrity target uses journal, it
9 writes sector data and integrity tags into a journal, commits the journal
10 and then copies the data and integrity tags to their respective location.
12 The dm-integrity target can be used with the dm-crypt target - in this
13 situation the dm-crypt target creates the integrity data and passes them
14 to the dm-integrity target via bio_integrity_payload attached to the bio.
15 In this mode, the dm-crypt and dm-integrity targets provide authenticated
16 disk encryption - if the attacker modifies the encrypted device, an I/O
17 error is returned instead of random data.
19 The dm-integrity target can also be used as a standalone target, in this
20 mode it calculates and verifies the integrity tag internally. In this
21 mode, the dm-integrity target can be used to detect silent data
22 corruption on the disk or in the I/O path.
25 When loading the target for the first time, the kernel driver will format
26 the device. But it will only format the device if the superblock contains
27 zeroes. If the superblock is neither valid nor zeroed, the dm-integrity
28 target can't be loaded.
30 To use the target for the first time:
31 1. overwrite the superblock with zeroes
32 2. load the dm-integrity target with one-sector size, the kernel driver
33 will format the device
34 3. unload the dm-integrity target
35 4. read the "provided_data_sectors" value from the superblock
36 5. load the dm-integrity target with the the target size
37 "provided_data_sectors"
38 6. if you want to use dm-integrity with dm-crypt, load the dm-crypt target
39 with the size "provided_data_sectors"
44 1. the underlying block device
46 2. the number of reserved sector at the beginning of the device - the
47 dm-integrity won't read of write these sectors
49 3. the size of the integrity tag (if "-" is used, the size is taken from
50 the internal-hash algorithm)
53 D - direct writes (without journal) - in this mode, journaling is
54 not used and data sectors and integrity tags are written
55 separately. In case of crash, it is possible that the data
56 and integrity tag doesn't match.
57 J - journaled writes - data and integrity tags are written to the
58 journal and atomicity is guaranteed. In case of crash,
59 either both data and tag or none of them are written. The
60 journaled mode degrades write throughput twice because the
61 data have to be written twice.
62 R - recovery mode - in this mode, journal is not replayed,
63 checksums are not checked and writes to the device are not
64 allowed. This mode is useful for data recovery if the
65 device cannot be activated in any of the other standard
68 5. the number of additional arguments
72 journal_sectors:number
73 The size of journal, this argument is used only if formatting the
74 device. If the device is already formatted, the value from the
77 interleave_sectors:number
78 The number of interleaved sectors. This values is rounded down to
79 a power of two. If the device is already formatted, the value from
80 the superblock is used.
83 The number of sectors in one buffer. The value is rounded down to
86 The tag area is accessed using buffers, the buffer size is
87 configurable. The large buffer size means that the I/O size will
88 be larger, but there could be less I/Os issued.
90 journal_watermark:number
91 The journal watermark in percents. When the size of the journal
92 exceeds this watermark, the thread that flushes the journal will
96 Commit time in milliseconds. When this time passes, the journal is
97 written. The journal is also written immediatelly if the FLUSH
100 internal_hash:algorithm(:key) (the key is optional)
101 Use internal hash or crc.
102 When this argument is used, the dm-integrity target won't accept
103 integrity tags from the upper target, but it will automatically
104 generate and verify the integrity tags.
106 You can use a crc algorithm (such as crc32), then integrity target
107 will protect the data against accidental corruption.
108 You can also use a hmac algorithm (for example
109 "hmac(sha256):0123456789abcdef"), in this mode it will provide
110 cryptographic authentication of the data without encryption.
112 When this argument is not used, the integrity tags are accepted
113 from an upper layer target, such as dm-crypt. The upper layer
114 target should check the validity of the integrity tags.
116 journal_crypt:algorithm(:key) (the key is optional)
117 Encrypt the journal using given algorithm to make sure that the
118 attacker can't read the journal. You can use a block cipher here
119 (such as "cbc(aes)") or a stream cipher (for example "chacha20",
120 "salsa20", "ctr(aes)" or "ecb(arc4)").
122 The journal contains history of last writes to the block device,
123 an attacker reading the journal could see the last sector nubmers
124 that were written. From the sector numbers, the attacker can infer
125 the size of files that were written. To protect against this
126 situation, you can encrypt the journal.
128 journal_mac:algorithm(:key) (the key is optional)
129 Protect sector numbers in the journal from accidental or malicious
130 modification. To protect against accidental modification, use a
131 crc algorithm, to protect against malicious modification, use a
132 hmac algorithm with a key.
134 This option is not needed when using internal-hash because in this
135 mode, the integrity of journal entries is checked when replaying
136 the journal. Thus, modified sector number would be detected at
140 The size of a data block in bytes. The larger the block size the
141 less overhead there is for per-block integrity metadata.
142 Supported values are 512, 1024, 2048 and 4096 bytes. If not
143 specified the default block size is 512 bytes.
145 The journal mode (D/J), buffer_sectors, journal_watermark, commit_time can
146 be changed when reloading the target (load an inactive table and swap the
147 tables with suspend and resume). The other arguments should not be changed
148 when reloading the target because the layout of disk data depend on them
149 and the reloaded target would be non-functional.
152 The layout of the formatted block device:
153 * reserved sectors (they are not used by this target, they can be used for
154 storing LUKS metadata or for other purpose), the size of the reserved
155 area is specified in the target arguments
157 * magic string - identifies that the device was formatted
159 * log2(interleave sectors)
161 * the number of journal sections
162 * provided data sectors - the number of sectors that this target
163 provides (i.e. the size of the device minus the size of all
164 metadata and padding). The user of this target should not send
165 bios that access data beyond the "provided data sectors" limit.
166 * flags - a flag is set if journal_mac is used
168 The journal is divided into sections, each section contains:
169 * metadata area (4kiB), it contains journal entries
170 every journal entry contains:
171 * logical sector (specifies where the data and tag should
173 * last 8 bytes of data
174 * integrity tag (the size is specified in the superblock)
175 every metadata sector ends with
176 * mac (8-bytes), all the macs in 8 metadata sectors form a
177 64-byte value. It is used to store hmac of sector
178 numbers in the journal section, to protect against a
179 possibility that the attacker tampers with sector
180 numbers in the journal.
182 * data area (the size is variable; it depends on how many journal
183 entries fit into the metadata area)
184 every sector in the data area contains:
185 * data (504 bytes of data, the last 8 bytes are stored in
188 To test if the whole journal section was written correctly, every
189 512-byte sector of the journal ends with 8-byte commit id. If the
190 commit id matches on all sectors in a journal section, then it is
191 assumed that the section was written correctly. If the commit id
192 doesn't match, the section was written partially and it should not
194 * one or more runs of interleaved tags and data. Each run contains:
195 * tag area - it contains integrity tags. There is one tag for each
196 sector in the data area
197 * data area - it contains data sectors. The number of data sectors
198 in one run must be a power of two. log2 of this value is stored