CWE-301: Reflection Attack in an Authentication Protocol
Reflection Attack in an Authentication Protocol
Weakness ID: 301 (Weakness Variant)
Status: Draft
Description
Description Summary
Simple authentication protocols are subject to reflection attacks if a malicious user can use the target machine to impersonate a trusted user.
Extended Description
A mutual authentication protocol requires each party to respond to a random challenge by the other party by encrypting it with a pre-shared key. Often, however, such protocols employ the same pre-shared key for communication with a number of different entities. A malicious user or an attacker can easily compromise this protocol without possessing the correct key by employing a reflection attack on the protocol.
Time of Introduction
Architecture and Design
Applicable Platforms
Languages
All
Common Consequences
Scope
Effect
Access Control
Technical Impact: Gain privileges / assume
identity
The primary result of reflection attacks is successful authentication
with a target machine -- as an impersonated user.
Likelihood of Exploit
Medium
Demonstrative Examples
Example 1
(Bad Code)
Example Languages: C and C++
unsigned char *simple_digest(char *alg,char *buf,unsigned int len,
int *olen) {
const EVP_MD *m;
EVP_MD_CTX ctx;
unsigned char *ret;
OpenSSL_add_all_digests();
if (!(m = EVP_get_digestbyname(alg))) return NULL;
if (!(ret = (unsigned char*)malloc(EVP_MAX_MD_SIZE))) return
NULL;
product authentication succeeds if user-provided
MD5 hash matches the hash in its database; this can be subjected to replay
attacks.
Potential Mitigations
Phase: Architecture and Design
Use different keys for the initiator and responder or of a different
type of challenge for the initiator and responder.
Phase: Architecture and Design
Let the initiator prove its identity before proceeding.
Other Notes
Reflection attacks capitalize on mutual authentication schemes in order to
trick the target into revealing the secret shared between it and another
valid user. In a basic mutual-authentication scheme, a secret is known to
both the valid user and the server; this allows them to authenticate. In
order that they may verify this shared secret without sending it plainly
over the wire, they utilize a Diffie-Hellman-style scheme in which they each
pick a value, then request the hash of that value as keyed by the shared
secret. In a reflection attack, the attacker claims to be a valid user and
requests the hash of a random value from the server. When the server returns
this value and requests its own value to be hashed, the attacker opens
another connection to the server. This time, the hash requested by the
attacker is the value which the server requested in the first connection.
When the server returns this hashed value, it is used in the first
connection, authenticating the attacker successfully as the impersonated
valid user.
[REF-7] Mark Dowd, John McDonald
and Justin Schuh. "The Art of Software Security Assessment". Chapter 2, "Insufficient Validation", Page
38.. 1st Edition. Addison Wesley. 2006.
Maintenance Notes
The term "reflection" is used in multiple ways within CWE and the
community, so its usage should be reviewed.
Description
