C bit fields pack individual bits into struct members, saving memory when many boolean or small-range fields are needed. Declare with unsigned type : nbits to allocate exactly N bits for the field.

Use bit fields for hardware registers, flags, or tight memory constraints.

// compile: gcc -o bitfields bitfields.c
// run: ./bitfields
// description: packing bits into struct
 
#include <stdio.h>
 
struct Flags {
    unsigned enabled : 1;
    unsigned mode : 2;
    unsigned priority : 3;
    unsigned reserved : 2;
};
 
struct Register {
    unsigned int addr : 16;
    unsigned int flags : 8;
    unsigned int status : 8;
};
 
int main() {
    struct Flags f;
    f.enabled = 1;
    f.mode = 2;
    f.priority = 5;
 
    printf("Flags size: %zu bytes\n", sizeof(f));
    printf("Register size: %zu bytes\n", sizeof(struct Register));
 
    return 0;
}

Bit field declarations:

• unsigned type: nbits: N-bit field
• Total bits in struct packed into minimal bytes
• Padding added as needed

Portability issues:

• Bit ordering, signedness, padding are implementation-defined
• Avoid bit fields for cross-platform data
• Use for local optimization only

Limitations:

• Can't take address of bit field (no &)
• Can't use bit fields in arrays
• Alignment and packing unpredictable

Best uses:

• Hardware register mapping
• Local flags struct
• Memory-constrained embedded systems