1
2
Fork 0
mirror of https://github.com/carlospolop/hacktricks.git synced 2023-12-14 19:12:55 +01:00
hacktricks/cryptography/hash-length-extension-attack.md

36 lines
1.7 KiB
Markdown
Raw Normal View History

# Hash Length Extension Attack
## Summary of the attack
Imagine a server which is **signing** some **data** by **appending** a **secret** to some known clear text data and then hashing that data. If you know:
* **The length of the secret** \(this can be also bruteforced from a given length range\)
* **The clear text data**
* **The algorithm \(and it's vulnerable to this attack\)**
* **The padding is known**
* Usually a default one is used, so if the other 3 requirements are met, this also is
* The padding vary depending on the length of the secret+data, that's why the length of the secret is needed
Then, it's possible for an **attacker** to **append** **data** and **generate** a valid **signature** for the **previos data + appended data**.
### How?
Basically the vulnerable algorithms generate the hashes by firstly **hashing a block of data**, and then, **from** the **previously** created **hash** \(state\), they **add the next block of data** and **hash it**.
Then, imagine that the secret is "secret" and the data is "data", the MD5 of "secretdata" is 6036708eba0d11f6ef52ad44e8b74d5b.
If an attacker wants to append the string "append" he can:
* Generate a MD5 of 64 "A"s
* Change the state of the previously initialized hash to 6036708eba0d11f6ef52ad44e8b74d5b
* Append the string "append"
* Finish the hash and the resulting hash will be a **valid one for "secret" + "data" + "padding" + "append"**
### **Tool**
{% embed url="https://github.com/iagox86/hash\_extender" %}
## References
You can find this attack good explained in [https://blog.skullsecurity.org/2012/everything-you-need-to-know-about-hash-length-extension-attacks](https://blog.skullsecurity.org/2012/everything-you-need-to-know-about-hash-length-extension-attacks)