epass2003: SM checking (MAC and SW) #2772
Conversation
|
Just for completeness: AES-CBC-MAC was tested on a real epass2003 token. I don't have a real token using DES, so the code for DES was written according to the construct_mac_tlv() function. This code was verified only in the simulation - (the simulator could be used as part of OpenSC CI tests, functionality example: https://github.com/popovec/OpenSC/actions/runs/4891381567) |
src/libopensc/card-epass2003.c
Outdated
|
|
||
| /* copy response to buffer and append padding */ | ||
| memcpy(data, in, mac_len); | ||
| memset(data + mac_len, 0, blocksize); |
There was a problem hiding this comment.
just use calloc instead of malloc above and you don't need to memset here
There was a problem hiding this comment.
It would be best to use alloca (let's save free maintenance). However, I did not find `alloc' anywhere in the OpenSC code, I understand that there may be a problem with portability.
There was a problem hiding this comment.
alloca is not supported on Windows, afaik. Also, I don't like it personally because you can't do any error handling (a failure in memory allocation with alloca is just undefined behavior).
This patch introduces basic verification of data coming from the token in the secure messaging mode: 1. compare raw SW and SW encoded in SM 2. verify MAC MAC verification is working only for DES and AES-CBC-MAC. Code for checking MAC in FIPS mode is still missing. modified: src/libopensc/card-epass2003.c
popovec
force-pushed
the
epass2003_SM_mac_check
branch
from
cb81904 to
2d2f57f
Compare
|
Rebased. Access outside the input buffer fixed by checking the |
3 participants
This patch introduces basic verification of data coming from the token in the secure messaging mode:
MAC verification is working only for DES and AES-CBC-MAC. Code for checking MAC in FIPS mode is still missing.