phc-discussions - battcrypt: minor bug in reference implementation

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [day] [month] [year] [list]

Date: Sun, 22 Mar 2015 22:48:21 +0100
From: Axel von dem Bruch <axel-vdb@...eup.net>
To: discussions@...sword-hashing.net
Subject: battcrypt: minor bug in reference implementation

I'm working on a Java implementation of battcrypt. I want to use it
later as an optional key derivation function for a password encryption
program, which is currently under construction.
For comparison I called the C++ reference implementation with several
parameters.

I found a minor bug in the function battcryptKdf().
The bug occurs only if the requested key size is larger than 64 byte and
has the result that the function fails.
The function battcryptKdf() is an additional function to the PHC
standard function.
This bug does not affect the function PHC().

More detailed description:
When I use the key derivation mode with output size > 64, the function
battcryptKdf(...) returns 1.
This is because of line 165 or 171:
if (m_cost > 50 || (t_cost & 0xffff) > 62 || (t_cost >> 16) > 63 ||
outlen > HASH_LENGTH)
For the function battcryptKdf the last "outlen > HASH_LENGTH" should
be deleted, otherwise the key streching will not work.

Steve has suggested to report the bug here on the list.

I had also noted (a negligible topic) that the function others than the
function PHC uses the password again in the last loop and therefore the
password must remain in RAM.

Regards
Axel