Explaining Endianness in C

I am on a journey to finish C Programming and Memory Management — Full Course. To follow, copy the code and paste it into an Online C compiler on pythontutor.com. The name of the journey is CTJ. These are my public notes. This is the step CTJ-53; The aim is to introduce the concepts of little and big endians.

1. Endianness is the order in which bytes are stored in memory

• coined by Danny Cohen in an Internet Experiment Note, 1980 [1]
• borrowed from Gulliver’s Travels (Jonathan Swift, 1726)
• in Lilliput, two factions war over which end of a boiled egg to crack
• Big-Endians crack the big end, Little-Endians the small end
• Cohen used it to mock the arbitrariness of the byte-order choice

2. C knows how to handle this automatically based on what architecture you’re on

• you can check with python

[embed]

3. There are two most common endian formats: big and little endian

• there is also middle-endian (mixed-endian) where bytes in neither big nor little order
• famous example: PDP-11 — stored 32-bit values as two 16-bit words big-endian, but bytes within each word little-endian
• bi-endian — architectures that can switch at runtime (ARM v3+, PowerPC, SPARC v9, RISC-V)

4. Big endian stores the most significant byte first at the lowest address

• the concept of {most,least}_siginifcant is straight from positional notation in math[2]
• most significant = highest weight: removing it changes the value the most
• e.g. in 1234: 1 is most significant digit (removing it loses 1000), 4 is least significant digit (removing it loses 4)
• same for bytes: in 0x12345678, 0x12 is the most significant byte
• e.g. given 0x12345678 value in hex
• 1 hex digit = 4 bits (a nibble); 2 hex digits = 1 byte
• why hex? because 4 bits → ²⁴ = 16 values → maps perfectly to 16 hex symbols (0–9, A-F) as hex is base16 and its values then aligns with bytes even though not perfectly (it’s 2 hex for 1 byte)
• if you wanted 1 hex to 1 byte, 8 bits → ²⁸ = 256 values → would need 256 symbols, unworkable for humans
• so we split the byte into two nibbles: two hex digits per byte
• 0x12345678 = 8 hex digits = 4 bytes — byte boundaries are immediately visible
• decimal 305419896 is the same number — byte boundaries invisible
• anyways, in Big endian you’d store it as you defined it

[embed]

• the most siginifcant byte (0x12) is stored at the lowest memory address
• TJ: there is a whole culture war in the endiann sphere about this

5. Little endian stores the least significant byte first at the lowest address

• 1978 — Intel 8086 released and it was little-endian
• 1981 — IBM PC standardized on x86, sealed the win
• not technical superiority — pure market dominance
• the least signifcant byte is stored first, see the opposite direction

[embed]

6. Let’s write some C

• let’s make a 4-byte integer variable (in hex so 2 values are always 1 byte)

[embed]

• then, let’s make a single-byte character pointer by creating an an array through reference

[embed]

• running p returns xV4\x12 because the numerical values are decoded into ASCII[3]

[embed]

• to print the lowest address in decimal, i.e. 120

[embed]

6. links

• [1] https://en.wikipedia.org/wiki/Endianness#Origin
• [2] https://en.wikipedia.org/wiki/Significant_figures
• [3] https://en.wikipedia.org/wiki/ASCII#/media/File:ASCIITable(suitable_for_printing).svg