250 lines · plain
1=========2dm-verity3=========4 5Device-Mapper's "verity" target provides transparent integrity checking of6block devices using a cryptographic digest provided by the kernel crypto API.7This target is read-only.8 9Construction Parameters10=======================11 12::13 14 <version> <dev> <hash_dev>15 <data_block_size> <hash_block_size>16 <num_data_blocks> <hash_start_block>17 <algorithm> <digest> <salt>18 [<#opt_params> <opt_params>]19 20<version>21 This is the type of the on-disk hash format.22 23 0 is the original format used in the Chromium OS.24 The salt is appended when hashing, digests are stored continuously and25 the rest of the block is padded with zeroes.26 27 1 is the current format that should be used for new devices.28 The salt is prepended when hashing and each digest is29 padded with zeroes to the power of two.30 31<dev>32 This is the device containing data, the integrity of which needs to be33 checked. It may be specified as a path, like /dev/sdaX, or a device number,34 <major>:<minor>.35 36<hash_dev>37 This is the device that supplies the hash tree data. It may be38 specified similarly to the device path and may be the same device. If the39 same device is used, the hash_start should be outside the configured40 dm-verity device.41 42<data_block_size>43 The block size on a data device in bytes.44 Each block corresponds to one digest on the hash device.45 46<hash_block_size>47 The size of a hash block in bytes.48 49<num_data_blocks>50 The number of data blocks on the data device. Additional blocks are51 inaccessible. You can place hashes to the same partition as data, in this52 case hashes are placed after <num_data_blocks>.53 54<hash_start_block>55 This is the offset, in <hash_block_size>-blocks, from the start of hash_dev56 to the root block of the hash tree.57 58<algorithm>59 The cryptographic hash algorithm used for this device. This should60 be the name of the algorithm, like "sha1".61 62<digest>63 The hexadecimal encoding of the cryptographic hash of the root hash block64 and the salt. This hash should be trusted as there is no other authenticity65 beyond this point.66 67<salt>68 The hexadecimal encoding of the salt value.69 70<#opt_params>71 Number of optional parameters. If there are no optional parameters,72 the optional parameters section can be skipped or #opt_params can be zero.73 Otherwise #opt_params is the number of following arguments.74 75 Example of optional parameters section:76 1 ignore_corruption77 78ignore_corruption79 Log corrupted blocks, but allow read operations to proceed normally.80 81restart_on_corruption82 Restart the system when a corrupted block is discovered. This option is83 not compatible with ignore_corruption and requires user space support to84 avoid restart loops.85 86panic_on_corruption87 Panic the device when a corrupted block is discovered. This option is88 not compatible with ignore_corruption and restart_on_corruption.89 90ignore_zero_blocks91 Do not verify blocks that are expected to contain zeroes and always return92 zeroes instead. This may be useful if the partition contains unused blocks93 that are not guaranteed to contain zeroes.94 95use_fec_from_device <fec_dev>96 Use forward error correction (FEC) to recover from corruption if hash97 verification fails. Use encoding data from the specified device. This98 may be the same device where data and hash blocks reside, in which case99 fec_start must be outside data and hash areas.100 101 If the encoding data covers additional metadata, it must be accessible102 on the hash device after the hash blocks.103 104 Note: block sizes for data and hash devices must match. Also, if the105 verity <dev> is encrypted the <fec_dev> should be too.106 107fec_roots <num>108 Number of generator roots. This equals to the number of parity bytes in109 the encoding data. For example, in RS(M, N) encoding, the number of roots110 is M-N.111 112fec_blocks <num>113 The number of encoding data blocks on the FEC device. The block size for114 the FEC device is <data_block_size>.115 116fec_start <offset>117 This is the offset, in <data_block_size> blocks, from the start of the118 FEC device to the beginning of the encoding data.119 120check_at_most_once121 Verify data blocks only the first time they are read from the data device,122 rather than every time. This reduces the overhead of dm-verity so that it123 can be used on systems that are memory and/or CPU constrained. However, it124 provides a reduced level of security because only offline tampering of the125 data device's content will be detected, not online tampering.126 127 Hash blocks are still verified each time they are read from the hash device,128 since verification of hash blocks is less performance critical than data129 blocks, and a hash block will not be verified any more after all the data130 blocks it covers have been verified anyway.131 132root_hash_sig_key_desc <key_description>133 This is the description of the USER_KEY that the kernel will lookup to get134 the pkcs7 signature of the roothash. The pkcs7 signature is used to validate135 the root hash during the creation of the device mapper block device.136 Verification of roothash depends on the config DM_VERITY_VERIFY_ROOTHASH_SIG137 being set in the kernel. The signatures are checked against the builtin138 trusted keyring by default, or the secondary trusted keyring if139 DM_VERITY_VERIFY_ROOTHASH_SIG_SECONDARY_KEYRING is set. The secondary140 trusted keyring includes by default the builtin trusted keyring, and it can141 also gain new certificates at run time if they are signed by a certificate142 already in the secondary trusted keyring.143 144try_verify_in_tasklet145 If verity hashes are in cache, verify data blocks in kernel tasklet instead146 of workqueue. This option can reduce IO latency.147 148Theory of operation149===================150 151dm-verity is meant to be set up as part of a verified boot path. This152may be anything ranging from a boot using tboot or trustedgrub to just153booting from a known-good device (like a USB drive or CD).154 155When a dm-verity device is configured, it is expected that the caller156has been authenticated in some way (cryptographic signatures, etc).157After instantiation, all hashes will be verified on-demand during158disk access. If they cannot be verified up to the root node of the159tree, the root hash, then the I/O will fail. This should detect160tampering with any data on the device and the hash data.161 162Cryptographic hashes are used to assert the integrity of the device on a163per-block basis. This allows for a lightweight hash computation on first read164into the page cache. Block hashes are stored linearly, aligned to the nearest165block size.166 167If forward error correction (FEC) support is enabled any recovery of168corrupted data will be verified using the cryptographic hash of the169corresponding data. This is why combining error correction with170integrity checking is essential.171 172Hash Tree173---------174 175Each node in the tree is a cryptographic hash. If it is a leaf node, the hash176of some data block on disk is calculated. If it is an intermediary node,177the hash of a number of child nodes is calculated.178 179Each entry in the tree is a collection of neighboring nodes that fit in one180block. The number is determined based on block_size and the size of the181selected cryptographic digest algorithm. The hashes are linearly-ordered in182this entry and any unaligned trailing space is ignored but included when183calculating the parent node.184 185The tree looks something like:186 187 alg = sha256, num_blocks = 32768, block_size = 4096188 189::190 191 [ root ]192 / . . . \193 [entry_0] [entry_1]194 / . . . \ . . . \195 [entry_0_0] . . . [entry_0_127] . . . . [entry_1_127]196 / ... \ / . . . \ / \197 blk_0 ... blk_127 blk_16256 blk_16383 blk_32640 . . . blk_32767198 199 200On-disk format201==============202 203The verity kernel code does not read the verity metadata on-disk header.204It only reads the hash blocks which directly follow the header.205It is expected that a user-space tool will verify the integrity of the206verity header.207 208Alternatively, the header can be omitted and the dmsetup parameters can209be passed via the kernel command-line in a rooted chain of trust where210the command-line is verified.211 212Directly following the header (and with sector number padded to the next hash213block boundary) are the hash blocks which are stored a depth at a time214(starting from the root), sorted in order of increasing index.215 216The full specification of kernel parameters and on-disk metadata format217is available at the cryptsetup project's wiki page218 219 https://gitlab.com/cryptsetup/cryptsetup/wikis/DMVerity220 221Status222======223V (for Valid) is returned if every check performed so far was valid.224If any check failed, C (for Corruption) is returned.225 226Example227=======228Set up a device::229 230 # dmsetup create vroot --readonly --table \231 "0 2097152 verity 1 /dev/sda1 /dev/sda2 4096 4096 262144 1 sha256 "\232 "4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076 "\233 "1234000000000000000000000000000000000000000000000000000000000000"234 235A command line tool veritysetup is available to compute or verify236the hash tree or activate the kernel device. This is available from237the cryptsetup upstream repository https://gitlab.com/cryptsetup/cryptsetup/238(as a libcryptsetup extension).239 240Create hash on the device::241 242 # veritysetup format /dev/sda1 /dev/sda2243 ...244 Root hash: 4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076245 246Activate the device::247 248 # veritysetup create vroot /dev/sda1 /dev/sda2 \249 4392712ba01368efdf14b05c76f9e4df0d53664630b5d48632ed17a137f39076250