A cryptographic hash function is an algorithm that takes data of arbitrary size as its input and returns a fixed size value, called message digest or checksum, as its output. For example, sha256sum calculates the SHA256 (Secure Hash Algorithm 256) message digest. SHA256, as the name indicates, returns a checksum of size 256 bits (32 bytes). This checksum is usually written using hexadecimal digits. Knowing that a hexadecimal digit represents 4 bits, the 256 bits checksum can be represented as 64 hexadecimal digits.
In the terminal output below, we calculate the SHA256 hash values for three files of varying sizes: 4 bytes, 275 MB, and 5.2 GB. Using sha256sum to calculate the message digest for each of the three files, we get three completely different values that appear random. It is worth stressing that the length of the resulting message digest or checksum is the same, no matter how small or big the file is. In particular, the four-byte file abc.txt and the 5.2 GB file resulted in message digests of equal length independent of the file size.
Terminal
`user@TryHackMe$ ls -lh total 5.5G -rw-r--r--. 1 strategos strategos 4 7月 21 12:46 abc.txt -rw-r--r--. 1 strategos strategos 275M 2月 12 19:08 debian-hurd.img.tar.xz -rw-r--r--. 1 strategos strategos 5.2G 4月 26 16:55 Win11_English_x64v1.iso $ sha256sum * c38bb113c89d8fec6475a9936411007c45563ecb7ce8acd5db7fb58c0872bda0 abc.txt 0317ff0150e0d64b70284b28c97bb788310585ea7ac46cc8139d5a3c850dea55 debian-hurd.img.tar.xz 4bc6c7e7c61af4b5d1b086c5d279947357cff45c2f82021bb58628c2503eb64e Win11_English_x64v1.iso`
But why would we need such a function? There are many uses, in particular:
- Storing passwords: Instead of storing passwords in plaintext, a hash of the password is stored instead. Consequently, if a data breach occurs, the attacker will get a list of password hashes instead of the original passwords. (In practice, passwords are also “salted”, as discussed in a later task.)
- Detecting modifications: Any minor modification to the original file would lead to a drastic change in hash value, i.e. checksum.
In the following terminal output, we have two files, text1.txt and text2.txt, which are almost identical except for (literally) one bit being different; the letters T and t are different in one bit in their ASCII representation. Even though we have flipped only a single bit, it is evident that the SHA256 checksums are entirely different. Consequently, if we use a secure hash function algorithm, we can easily confirm whether any modifications have taken place. This can help protect against both intentional tampering and file transfer errors.
Terminal
`user@TryHackMe$ hexdump text1.txt -C 00000000 54 72 79 48 61 63 6b 4d 65 0a |TryHackMe.| 0000000a $ hexdump text2.txt -C 00000000 74 72 79 48 61 63 6b 4d 65 0a |tryHackMe.| 0000000a $ sha256sum text1.txt f4616fd825a10ded9af58fbaee09f3e31751d15591f9323ea68b03a0e8ac3783 text1.txt $ sha256sum text2.txt 9ffa3533ee33998aeb1df76026f8031c8af6ccabd8393eca002d5b7471a0b536 text2.txt`
Some of the hashing algorithms in use and still considered secure are:
- SHA224, SHA256, SHA384, SHA512
- RIPEMD160
Some older hash functions, such as MD5 (Message Digest 5) and SHA-1, are cryptographically broken. By broken, we mean that it is possible to generate a different file with the same checksum as a given file. This means that we can create a hash collision. In other words, an attacker can create a new message with a given checksum, and detecting file or message tampering won’t be possible.
HMAC
Hash-based message authentication code (HMAC) is a message authentication code (MAC) that uses a cryptographic key in addition to a hash function.
According to RFC2104, HMAC needs:
- secret key
- inner pad (ipad) a constant string. (RFC2104 uses the byte
0x36repeated B times. The value of B depends on the chosen hash function.) - outer pad (opad) a constant string. (RFC2104 uses the byte
0x5Crepeated B times.)
Calculating the HMAC follows the following steps as shown in the figure:
- Append zeroes to the key to make it of length B, i.e., to make its length match that of the ipad.
- Using bitwise exclusive-OR (XOR), represented by ⊕, calculate k__e__y ⊕ i__p__a__d.
- Append the message to the XOR output from step 2.
- Apply the hash function to the resulting stream of bytes (in step 3).
- Using XOR, calculate k__e__y ⊕ o__p__a__d.
- Append the hash function output from step 4 to the XOR output from step 5.
- Apply the hash function to the resulting stream of bytes (in step 6) to get the HMAC.
The figure above represents the steps expressed in the following formula: H(K⊕o__p__a__d,H(K⊕i__p__a__d,t__e__x__t)).
To calculate the HMAC on a Linux system, you can use any of the available tools such as hmac256 (or sha224hmac, sha256hmac, sha384hmac, and sha512hmac, where the secret key is added after the option --key). Below we show an example of calculating the HMAC using hmac256 and sha256hmac with two different keys.
Terminal
`user@TryHackMe$ hmac256 s!Kr37 message.txt 3ec65b7e80c5bf2e623e52e0528f1c6a74f605b10616621ba1c22a89fb244e65 message.txt user@TryHackMe$ hmac256 1234 message.txt 4b6a2783631180fca6128592e3d17fb5bff6b0e563ad8f1c6afc1050869e440f message.txt user@TryHackMe$ sha256hmac message.txt --key s!Kr37 3ec65b7e80c5bf2e623e52e0528f1c6a74f605b10616621ba1c22a89fb244e65 message.txt user@TryHackMe$ sha256hmac message.txt --key 1234 4b6a2783631180fca6128592e3d17fb5bff6b0e563ad8f1c6afc1050869e440f message.txt`