Brian Murray
b0c3c43dec
CMAC support for cipher with 64bit blocks (DES/3DES)
2016-10-05 14:19:15 +01:00
Manuel Pégourié-Gonnard
7b555f2928
cmac: more cosmetic changes
2016-10-05 14:19:15 +01:00
Manuel Pégourié-Gonnard
ab9c5fd9b3
cmac: avoid useless wrapping of function
...
probably a leftover from an earlier stage
2016-10-05 14:19:15 +01:00
Manuel Pégourié-Gonnard
d18c70708e
cmac: reduce visibility of macros
...
The #undef is usefull for people who want to to amalgamated releases
2016-10-05 14:19:15 +01:00
Manuel Pégourié-Gonnard
2c06306364
cmac: some more padding-related tune-ups
...
- use one less temporary buffer
- pedantic: in_len + 15 was a potential overflow
- use a more explicit name instead of 'flag'
- Mn was a bit misleading
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
d2c3d3eddb
cmac: clean up padding function and comments
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
475f06f609
cmac: make subkey gen more constant-time
...
The previous version had secret-dependent memory accesses. While it was
probably not an issue in practice cause the two bytes of the array are
probably on the same cache line anyway, as a matter of principle this should
be avoided.
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
d6cf75474b
cmac: zeroize sensitive intermediate values
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
a610b4c04b
cmac: factor multiply by u to its own function
...
We're doing exactly the same operation for K1 and K2.
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
3da5402a89
cmac: fix whitespace/codingstyle issues
2016-10-05 14:19:14 +01:00
Manuel Pégourié-Gonnard
690083c21d
cmac: improve argument order and naming
...
- always use the pattern "buffer, length"
- avoid using unqualified "length" as a name when there are more than one
2016-10-05 14:19:14 +01:00
Robert Cragie
3d23b1d5ed
Line endings wrong
2016-10-05 14:19:14 +01:00
Robert Cragie
dc5c7b98ac
Add support for AES-128-CMAC and AES-CMAC-PRF-128
2016-10-05 14:19:13 +01:00
