Study Notes: Instruction Sets

Chapter 13: Instruction Sets: Characteristics and Functions

Machine Instruction Characteristics

• The operation of the processor is determined by the instructions it executes, referred to as machine instructions or computer instructions.

• The collection of different instructions that the processor can execute is referred to as the processor’s instruction set.

• Each instruction must contain the information required by the processor for execution.

Instruction Cycle State Diagram

• Instruction Fetch: The process starts with fetching the instruction.

• Operand Fetch: The required operands are retrieved for the operation.

• Multiple Operands Handling: If applicable, multiple operands are fetched.

• Operand Store: Results are stored back into memory or registers.

• Instruction Address Calculation / Instruction Operation Decoding: The address of the next instruction is calculated and the current instruction is decoded.

• Completion: Instructions are completed, and the next instruction or vector data is fetched.

Operand Locations

Source and result operands can exist in one of four areas:

• Main or Virtual Memory: The instruction must specify the address in memory.

• I/O Device: This includes specifying the I/O module and device for the operation.

• Processor Register: Refers to the internal storage locations within the processor.

• Immediate: The operand's value is given directly within the instruction.

Instruction Format

• Each instruction is represented by a sequence of bits divided into fields, corresponding to constituent elements of the instruction.
  
  

  

Instruction Representation

• Opcodes: Represented by abbreviations known as mnemonics (e.g., ADD, SUB, MUL, DIV, LOAD, STORE).

• Operands: Symbolically represented; the programmer indicates their locations.

• Each symbolic opcode has a fixed binary representation.

Instruction Types

• I/O Instructions: Transfer programs and data in/out of memory.

• Test Instructions: Check the value of data or computation status.

• Branch Instructions: Control flow based on conditions.

• Data Movement: Move data between registers, memory.

• Arithmetic Instructions: Perform computations on numeric data.

• Logic Instructions: Operate on bit-level representations.

Examples of Instruction Operations

Table 13.1: Utilization of Instruction Addresses (Nonbranching Instructions)

Instruction Set Design

Fundamental Design Issues:

• Operation Repertoire: Determine which operations to include and their complexity.

• Data Types: Identify types of data operations.

• Instruction Format: Define lengths, field numbers.

• Registers: Determine the number and use cases for registers.

• Addressing: Address operand specification modes.

Types of Operands

• Addresses: Refer to memory locations.

• Numbers: Include fixed-point and floating-point data types.

• Characters: Textual representation data.

• Logical Data: Boolean values represented as bits.

Numeric Data Types

• Options include:

  • Binary Integer: Fixed-point representation.

  • Binary Floating Point: Floats represented in scientific notation.

  • Decimal: Use binary-coded decimal (BCD).

  • Packed Decimal: Each decimal digit in a 4-bit representation.

Character Encoding

• Commonly used character encoding includes:

  • ASCII: American Standard Code for Information Interchange.

  • EBCDIC: Extended Binary Coded Decimal Interchange Code used by IBM.

Logical Data Representation

• Logical Data: Represented in n bits, with each bit having a value of 0 or 1.

• Advantages:

  • Efficient memory use for Boolean arrays.

  • Manipulation capabilities: Shifting bits in operations.

x86 Data Types

Table 13.2:

Numeric Data Formats

SIMD Data Types

• Introduced to the x86 architecture to optimize multimedia applications.

• Extensions: Include MMX and SSE.

• Data Types: Packed variants of bytes, words, doublewords, and floating-point numbers.

ARM Data Types

• ARM processors support data types:

  • 8, 16, 32 bits in length.

• Alignment Checking: Ensures proper data alignment. Unaligned access requires special handling.

Common x86 Instruction Set Operations

Table 13.3: Operations (1 of 3)

Table continued…

Processor Actions for Various Operations

Table 13.4: Actions

Data Transfer Instructions

• Most fundamental types of instructions.

• Must indicate source, destination, length, and addressing mode.

IBM EAS/390 Example Data Transfer Operations

Table 13.5

Arithmetic Instructions

• Basic operations typically include add, subtract, multiply, divide.

• Also provided for floating-point and packed decimal values.

• Additional operations: Absolute value, Negate, Increment, Decrement.

Logical Operations

Table 13.6: Example Operations

Shift and Rotate Operations

Table 13.7 Examples

Input/Output Approaches

• Varieties include isolated programmed I/O, memory-mapped I/O, and DMA.

• Many systems provide a few I/O instructions with specific actions specified by parameters/codes.

System Control Instructions

• Executed in privileged states; often reserved for OS use.

• Examples include modifying control registers and accessing process control blocks.

Transfer of Control Instructions

• Include Branch, Skip, and Procedure call instructions.

Procedure Call Instructions

• Self-contained programs that can be invoked multiple times.

• Involves a CALL instruction to invoke and a RETURN instruction to exit.

x86 Procedure Operations

Table 13.8: Status Flags

Single-Instruction, Multiple-Data (SIMD) Instructions

• Intel introduced MMX for multimedia tasks in 1996.

Table 13.10: MMX Instruction Set

Summary of Chapter 13

• Includes machine instruction characteristics, instruction representation, types, number of addresses, operand types, and operation types for Intel x86 and ARM.

References

• All rights reserved. Text derived from Pearson Education materials.