Usages of the special purpose registers

Short answer to "usages of special registers"

The special purpose registers are used exclusively for:

Implementation (= make it happen) of the function of the CPU
I.e.: the special purpose registers are part of the machinary that enable the CPU to perform:

First: an analogy to help you understand the operation of a computer

I trust that you are familiar with a mechanical piano where a sheet of paper with hole are fed into the mechinical piano:
Each row of holes in the paper will cause the mechanical piano to hit a specific set of keys (= notes)

I like to compare a computer with this imaginary (made it up myself) mechanical calculator:

In the mechanized calculator analogy:

A row of hole = an instruction for the calculator
The mechanized calculator will get the next instruction on its own and perform the fetched instructions one at a time.

Explaining the operation of a computer using the mechanized calculator analogy:

A computer is an electrical device, and I only use the mechanized calculator analogy to make the operation of a computer easier to understand because:

The computer has a memory that stores the instructions and the data (that's the sheet of paper with holes)
The computer has a CPU that performs the computations encoded in the instructions (that's the calculator on top of the sheet of paper)

Operation of a CPU:

Slideshow:

(Prelude: program flow - computer program and computer instruction)

(Normal program flow: next computer instruction is the one after the current one)

(Out-of-order program flow: program jumps to another location in memory)

(Instruction Register (IR): contains the current computer instruction that is being executed)

(Program Counter (PC): contains the location (= address) of the next computer instruction )

❮ ❯

Giving you some context to understand the CPU operation: Instructions, program, computer memory and the current instruction

Maybe some of you have not had this information presented (taught) in CS170/CS171, so I will teach (= write) it here to give you the context to understand the operation of the CPU:

A computer program consists of computer instructions and is stored in memory:

When the computer program is run (= executed), then:

The CPU fetches the (computer) instructions in the program -- one instruction at a time, starting at address 0 - and execute the (fetched) instruction
When the CPU finishes executing an instruction, the CPU will fetch the next instruction and execute it.

The current instruction:
The next instruction:
The program instructions is executed in a specific order
There are 2 types of program flows:
The conditional statements (if, if-else statements) and loop statements (while, for statements) will cause out-of-order program flow in a program execution

The in-order or normal program flow is when:

After executing the current instruction in the program, the CPU fetches and executes the instruction that physically follows the current instruction in the program

The normal program flow depicted in a diagram:

An out-of-order program flow is when:

After executing an instruction in the program, the CPU fetches and executes a instruction that is not the physically next instruction in the program

The out-of-order program flow depicted in a diagram:

Comment:

The next executed instruction can also be located before th currently executing instrucion !!!
When an our-of-order program flow occurs, we say that:

The special purpose registers allows the CPU to perform its designed function which is:

Fetch the next instruction according to the program control flow from memory into the CPU
Execute the (fetched) instruction

We will now discuss the usage of the 3 special purpose registers

Usage of the Instruction Register (IR)

Instruction Register:

The Instruction Register contains the instruction that is currently being executed by the CPU:
The IR register contains the binary code of the current instruction that the CPU is executing
The circuitry inside the CPU will emit control signals to various parts of the CPU to instruct the components (such as registers, ALU, etc) to cooperate to achieve the result (= action) encoded by the current instruction

If you want to know what circuitries are used and how the different components coorporate to execute a computing instruction, you will have to take CS355.

Usage of the Program Counter (PC)

Program Counter:

The Program Counter (register) contains the address (= location) of the next instruction that the processor will fetch and execute (after the processor finish executing the instruction in the Instruction Register)

Usage of the Processor Status Register (PSR)

Processor Status Register:

The Processor Status Register (as the name says) contains the status of the processor:

The PSR register contains a (large) number bits (typically 32)

The value of each bit represents a certain execution condition

Example:

The PSR register in the ARM processor (called the "Current Program Status Register") contain the follown fields:
(The diagram above is taken from this website: https://azeria-labs.com/arm-data-types-and-registers-part-2/ )

For CS255 (where we use assembler programming to understand constructs in High Level Programming Languages), we only use the following computation status "flags" (= bits) in the PSR register:

Explanation:

N = 1 when the execution of the previous instruction resulted in a negative value (i.e., the left-most bit = 1
N = 0 otherwise.
Z = 1 when the execution of the previous instruction resulted in the value zero (= 0) (i.e., all bits are equal to 0
Z = 0 otherwise.
C = 1 when the execution of the previous instruction produced a carry in the 33^th bit position
C = 0 otherwise.
V = 1 when the execution of the previous instruction produced a value that cannot be represented with 32 bits (i.e.: value is too large, or overflow)
V = 0 otherwise.

How are the N,Z,V,C flags used in assembler programming

These N,Z,V,C flags are called:
because together, their values can be used to check for the comparison outcome of 2 values.

When you want to compare the value x and y, you can subtract the values (x - y) and check the condition flags to find out which value is large.

Example:

Compare: x = 4 and y = 5 4 - 5 -------- -1

The outcome of the subtraction contains information about whether:

x == y (result = 0 !) x != y (result != 0 ) x < y (result is a negative number !) x > y (result is a positive number !) x <= y (result = 0 or a negative number) x >= y (result = 0 or a psotive number)

You can check the setting of the flags to find out which condition is true.

Here is a table of what flag setting to look for:

Source: https://azeria-labs.com/arm-data-types-and-registers-part-2/

Comment:

You do not need to remember the flag settings
I will show you how to test for a given condition using easy to remember assembler instructions later