What is the use of bit field in structure declaration?

Answer 1

A bit field allows you to assign an arbitrary number of bits to a member.

As an example, suppose we wish to implement an R6000 PPN (physical page number). A PPN is a 32-bit value divided up as follows:

page frame number (22 bits)

unused (3 bits)

cache coherency algorithm (3 bits)

nonreachable (1 bit)

dirty (1 bit)

valid (1 bit)

global (1 bit)

While we could use bitwise logic to determine the state of each bit or group of bits within a 32-bit value, it's always good programming practice to express ideas directly in code. Therefore we want a structure that is exactly 32-bits in length with members that exactly match those of the PPN itself. To achieve this we need to use bit fields:

//

// All example code based upon Chapter 25.5.5 "Bitfields" of

// "Programming -- Principles and Practice Using C++" by Bjarne Stroustrup

//

struct PPN { // R6000 Physical Page Number

unsigned int PFN : 22; // Page Frame Number

int : 3; // unused

unsigned int CCA : 3; // Cache Coherency Algorithm

bool nonreachable : 1;

bool dirty : 1;

bool valid : 1;

bool global : 1;

};

Any consumer of this structure need be left in little doubt regarding its purpose. Moreover, the implementer's job is greatly simplified:

void part_of_VM_system (PPN& ppn)

{

if (ppn.dirty) { // contents changed

// copy to disc

ppn.dirty = 0 ;

}

}

We can also see how much simpler it is to extract information from a structure's bit fields compared to that of a 32-bit value:

int get_CCA (const PPN& ppn) {

return (int) ppn.CCA;

}

int get_CCA (unsigned int ppn) {

return (int) (ppn>>3)&0x7;

}