## LSU EE 4720 -- Spring 2008 -- Computer Architecture
#
## MIPS  Overview

# Time-stamp: <4 March 2008, 15:53:06 CST, koppel@buz>

## Note
#
# This set based on EE 3755 Spring 2002 sets 9, 10, and 12.

## Contents
#
# Machine Language Basics
# Machine Languages Covered in ECE Courses
# Instructions, Registers, Immediates, and Memory
# Basic Three-Register Instructions
# Other Basic Instructions
# Instruction Coding
# Miscellaneous Integer Arithmetic Instructions
# Pseudo Instructions
#
# Types of Control-Transfer Instructions
# A Simple Jump Instruction
# A Branch Instruction
# More Branch Instructions
# More Jump Instructions
# Procedure Call Example
# Typical CTI Uses
#
# Load Byte (Unsigned)
# Store Byte
# Load Byte
# Load Word, Store Word
# Load and Store Half
# Array Access Examples
# Histogram Program


## References
#
# :PH:  Patterson & Hennessy, "Computer Organization & Design"
# :Mv1: MIPS Technologies, "MIPS32 Architecture for Programmers Vol I: Intro"
# :Mv2: MIPS Technologies, "MIPS32 Architecture for Programmers Vol II: Instr"


## Objectives
#
# Understand MIPS Register and Memory Organization
# Understand MIPS Instructions
#   Read programs using instructions, write program using instructions.
# Understand MIPS Instruction Coding and Formats
#   Determine binary coding from assembly, and vice versa.
# Ability to Learn New Instructions from MIPS Documentation
# Write Simple Assembly Language Programs


################################################################################
## Machine Language Review

# :PH: 3

# :Def: Machine Language
#
# The language understood by the computer processor itself.
# An ISA describes the machine language, plus other details.
#

# :Def: Machine Code
#
# All or part of a machine language program.


# :Def: High-Level Language
#
# A computer programming language for humans.
#
# Examples: C, Basic, Java, Fortran.
#
# A program in a high-level language may be converted into machine
# language and then run.
#
# A high-level language can be converted into the machine language of
# (compiled for) many machines.


# :Def: Assembly Language
#
# A language for humans used to prepare machine code.
#
# Assembly language is human readable, machine language is not.
# (At least without great effort or rare talent.)
v#
# For each machine language there is usually one assembly language.
# (In principle there could be many assembly languages for each
#  machine language.)
#
# The machine language and corresponding assembly language are sometimes
# known by the same name, for example MIPS.
#
# Almost all examples will be in assembly language.


 ## A Short MIPS Assembly Language Program
#
#
        add $s1, $s2, $s3       # s1 = s2 + s3
        sub $s1, $s4, $s1       # s1 = s4 - s1
#
# Comments start with a #.
#
# $s1-$s4 are storage locations called /registers/.
#
# Each register holds 32 bits.

# The program again, showing initial, intermediate, and final register values.
#
#                         Initial:  s1 = 10,   s2 = 20,   s3 = 30,   s4 = 40
        add $s1, $s2, $s3
#                                   s1 = 50,   s2 = 20,   s3 = 30,   s4 = 40
        sub $s1, $s4, $s1
#                         Final:    s1 =-10,   s2 = 20,   s3 = 30,   s4 = 40

# The program again, in three forms:
#
#  High-Level (C)       Assembly (MIPS)        Machine (MIPS)
#  a = b + c;           add $s1, $s2, $s3      0x02538820
#  a = d - a;           sub $s1, $s4, $s1      0x02918822
#
# Note: the relationship between high-level code and assembly code
#       is rarely one machine instruction per high-level statement.



################################################################################
## Instructions, Registers, Immediates, and Memory

# :PH: 3.3
# :Mv1: 2.8.4, 2.8.5  MIPS register descriptions


 ## Where MIPS Instructions Get Data
#
# Immediates:  Data stored in instruction.
# Registers:   A small number (66) of high-speed 32-bit storage locations.
# Memory:      A large number (2^32) of low-speed 8-bit storage locations.

        addi $a0, $a0, 1  # Gets data from register ($a0) and immed. (1)
        lbu $t1, 0($a0)   # Gets data from memory (address in $a0)

 ## What MIPS Instructions can Modify
#
# Registers
# Memory

        addi $a0, $a0, 1  # Modifies register, $a0.
        sb $t1,-1($a0)    # Modifies memory, address is $a0-1


 ## Registers
#
# Used in this Set  (MIPS has additional registers.)
#
# $0-$31:   32 bits, General Purpose Registers (GPRs). See GPR names below.
# PC:       32 bits (30 bits), Program Counter
# $lo, $hi: 32 bits, Registers used for multiplication and division.
# $f0-$f31: 32 bits, Floating Point (coprocessor 1) Registers


 ## General-Purpose Register (GPR) Names
#
# There are 32 (arguably 31) general purpose registers (GPRs).
#
# Each register holds 32 bits. (MIPS-32)
#
# Each has a name and a number.
#
#   Numbers range from 0 to 31.
#
#   Names indicate suggested use.
#
# Either names or numbers can be used in assembly language.
# In machine language only numbers are used.
#
# Ways to Reference GPR Number 1 (varies by assembler, textbook author, etc.)
#
#   By number:  1, $1, r1, R1
#
#   By name:    $at


# List of General-Purpose Registers. (Will be explained later.)
# See :PH: A-23
#
# The value of register 0 ($zero) is always zero.
#
# The value of the other registers is the last value written.
#
#
 ##      Names     Numbers   Suggested Usage
#      $zero:    0       The constant zero.
#      $at:      1       Reserved for assembler.
#      $v0-$v1:  2-3     Return value
#      $a0-$a3:  4-7     Argument
#      $t0-$t7:  8-15    Temporary (Not preserved by callee.)
#      $s0-$s7: 16-23    Saved by callee.
#      $t8-$t9: 24-25    Temporary (Not preserved by callee.)
#      $k0-$k1: 26-27    Reserved for kernel (operating system).
#      $gp      28       Global Pointer
#      $sp      29       Stack Pointer
#      $fp      30       Frame Pointer
#      $ra:     31       Return address.

# :Example:
#
# The two instructions below are identical, the first uses register
# names, the second uses register numbers.
        add $s1, $t2, $sp
        add $17, $10, $29
####


 ## Memory
#
# Memory consists of 2^32 locations. [MIPS-64 has 2^64 locations.]
#
# Each location stores 8 bits, called a character or a byte.
#
# Each location has an address, an integer ranging from 0 to 4294967295
#
# Memory access is usually slower than register access.

# :Example:
#
# Examples of instructions accessing memory.  The first loads
# something from memory into a register, the second stores something
# from a register to memory.  These will be covered in detail later.
	lw	$4, 0($19)
	sw	$31, 72($sp)
####

 ## Immediates
#
# An immediate is a constant value stored in the instruction itself.
#
# In MIPS immediate are usually 16 bits and can be interpreted as
# signed or unsigned numbers (depending on the instruction).

# :Example:
#
# An instruction using an immediate, in this case 100.
        addi $1, $2, 100


################################################################################
## Basic Three-Register Instructions

# :PH:  Throughout Chapter 3.
# :Mv2: A complete list of instructions.

# The three-register instructions each read two source registers and
# write one destination register.

 ## The Plain Old Add Instruction
#
# :Syntax: ADD rd, rs, rt
#          rd <- rs + rt
#
# How to read the syntax shown above:
#
#   The mnemonic (name) of the instruction is ADD.  A lower-case
#   version of the mnemonic (add) should be used in assembly language.
#
#   The instruction has three operands, rd, rs, rt.
#   These symbols refer to general-purpose register operands.
#   In assembly language they would be replaced by a register name or number.
#   (Machine language coding will be covered later.)
#
#   The line rd <- rs + rt indicates what the instructions does.
#
# ADD instruction assembly language examples:

        add $1, $2, $3
        add $s1, $t2, $v1

# The instruction below is NOT an add instruction because $rt is not a GPR.
# A "strict" assembler would NOT be able to convert that in to machine
# code.  (A "liberal" assembler would correct the mistake by replacing
# "add" with "addi".)

        add $1, $2, 100

# There is an instruction that can take the constant 100, "addi."


 ## Other Basic Three-Operand Instructions
#
# :Syntax: SUB rd, rs, rt
#          rd <- rs - rt
#
# :Syntax: AND rd, rs, rt
#          rd <- rs & rt
#
# :Syntax: OR rd, rs, rt
#          rd <- rs | rt
#
# :Syntax: NOR rd, rs, rt
#          rd <- ~( rs | rt )
#
# :Syntax: XOR rd, rs, rt
#          rd <- rs ^ rt
#
# :Syntax: SLT rd, rs, rt           # Set less than.
#          rd <- rs < rt ? 1 : 0;
#          Comparison as signed integers.
#          There's no SGT, which makes sense if you think about it.
#
# :Syntax: SLLV rd, rs, rt          # Shift Left Logical Variable
#          rd <- rs << rt[4:0]
#          Note: Only five low-order bits of rt are used (other bits ignored).
#
# :Syntax: SRLV rd, rs, rt          # Shift Right Logical Variable
#          rd <- rs >> rt[4:0]
#          Note: "Vacated" bit positions are zero.
#          Note: Only five low-order bits of rt are used (other bits ignored).
#
# :Syntax: SRAV rd, rs, rt          # Shift Right Arithmetic Variable
#          rd <- {rs[31],rs[31],..., rs >> rt[4:0] }
#          Note: Sign is extended: bit (rs[31]) placed in vacated positions.
#          Note: Only five low-order bits of rt are used, the rest are ignored.

# :Example:
#
# Simplified assembler for the following line of C:
#
#  x = a + b - ( c + d );
#
#  Asm regs: x, $8;  a, $9;  b, $10;  c, $11;  d, $12

        add $8, $9, $10    # a + b
        add $13, $11, $12  # c + d
        sub $8, $8, $13
####


# :Example:
#
# Use of register zero.
#
# Simplified assembler for the following line of C:
#
#  x = -a;
#  y = -1;
#
#  Asm regs: x, $8;  a, $9;  y, $10
#  Reminder: register $0 is always zero.

        sub $8, $0, $9
        nor $10, $0, $0
####


# :Example:
#
# Simplified assembler for the following line of C:
#
#  x = ( ir >> off ) & mask;
#
# Asm regs: x, $8;  ir, $9;  off, $10;  mask, $11.

        srlv $8, $9, $10
        and $8, $8, $11
####


# :Example:
#
# Simplified assembler for the following line of C:
#
#  x = a + b > ( c & mask | d );
#
# Asm regs: x, $8;  a, $9;  b, $10;  mask, $11;  c, $12;  d, $13

        add $20, $9, $10    # a + b
        and $21, $12, $11   # c & mask
        or $21, $21, $13    # c & mask | d
        slt $8, $21, $20    # >


################################################################################
## Other Basic Instructions

# :PH:  Throughout Chapter 3.
# :Mv2: A complete list of instructions.

# :Def: Immediate
#
# A constant stored in the instruction itself.
#
# In MIPS most immediates are 16 bits, including the ones in this section.


 ## Assembler Metasyntactic Symbols for Immediates
#
# Metasyntactic symbols are used in syntax descriptions.
#
# immed:  16-bit immediate, for general use.
# sa:     5-bit immediate (shift amount), used in shift instructions.


 ## The Plain-Old Addi Instruction
#
# :Syntax: ADDI rt, rs, immed            # Add Immediate
#          rt <- rs + sign_extend(immed)
#
# Notes on the syntax above.
#
#   The destination register is now called rt.
#
#   Symbol immed refers to a 16 bit value which is stored in the instruction.
#
#   For the ADDI instruction immed is interpreted as a signed integer.


 ## Some More Instructions
#
# The following are NOT MIPS32 instructions:
#   SUBI, NORI
#
# :Syntax: ANDI rt, rs, immed           # And Immediate
#          rt <- rs & immed
#
# :Syntax: ORI rt, rs, immed
#          rt <- rs | immed
#
# :Syntax: XORI rt, rs, immed
#          rt <- rs ^ immed
#
# :Syntax: SLTI rt, rs, immed           # Set Less Than Immediate
#          rt <- rs < sign_extend(immed) ? 1 : 0;
#          Comparison as signed integers, immed is sign-extended.
#          There's no SGT, which makes sense if you think about it.
#
# :Syntax: SLL rd, rt, sa               # Shift Left Logical
#          rd <- rt << sa
#
# :Syntax: SRL rd, rt, sa               # Shift Right Logical
#          rd <- rt >> sa
#          Note: "Vacated" bit positions are zero.
#
# :Syntax: SRA rd, rt, sa               # Shift Right Arithmetic
#          rd <- {rt[31],rt[31],..., rs >> sa }
#          Note: Sign is extended: bit (rt[31]) placed in vacated positions.
#
# :Syntax: LUI rt, immed                # Load Upper Immediate
#          rt <- {immed,16'b0}


# :Example:
#
# Simple examples to illustrate what some instructions do.
#
        # Assuming $9 = 23
        addi $8, $9, 1
        # $8 <- 24

        # Assuming $9 = 23
        addi $8, $9, -1
        # $8 <- 22

        # Assuming $9 = 23
        addi $8, $9, 0x12345  # ASSEMBLER SHOULD REJECT THIS INSTRUCTION.
        # No result because no such instruction exists.

        # Assuming $9 = 0x12345678
        andi $8, $9, 0xff     # Called a "mask" operation.
        # $8 <- 0x78

        # Assuming $9 = 23
        slti $8, $9, 20       # if( $9 < 20 ) $8 = 1; else $8 = 0;
        # $8 <- 0
        slti $8, $9, 30
        # $8 <- 1

        # Assuming $9 = 1
        sll $8, $9, 1
        # $8 <- 2
        sll $8, $9, 3
        # $8 <- 8

        # Assuming $9 = 16
        srl $8, $9, 1
        # $8 <- 8
        srl $8, $9, 3
        # $8 <- 2
        sra $8, $9, 1
        # $8 <- 8
        sra $8, $9, 3
        # $8 <- 2

        # Assuming $9 = -1 = 0xffffffff
        srl $8, $9, 1
        # $8 <- 2147483647 = 0x7fffffff
        srl $8, $9, 3
        # $8 <- 536870911 = 0x1fffffff
        sra $8, $9, 1
        # $8 <- -1 = 0xffffffff
        sra $8, $9, 3
        # $8 <- -1 = 0xffffffff

        # Assuming $9 = -7 = 0xfffffff9
        srl $8, $9, 1
        # $8 <- 2147483644 = 0x7ffffffc
        srl $8, $9, 3
        # $8 <- 536870911 = 0x1fffffff
        sra $8, $9, 1
        # $8 <- -4 = 0xfffffffc

        lui $8, 0x1234
        # $8 <- 0x12340000


 ## Uses for Immediates
#
# Register Initialization
# Constants


# :Example:
#
# Register initialization examples.
#
# Simplified assembler for the following line of C:
#
#  a = 23;
#  b = 105;
#  c = 0x1234a678;  // Sixteen bits not enough.
#  d = 0;
#
# Asm regs: a, $8;  b, $9;  c, $10;  d, $11

        addi $8, $0, 23         # a = 23
        addi $9, $0, 105        # b = 105
        lui  $10, 0x1234        # c = 0x12340000 (so far)
        ori  $10, $10, 0xa678   # c = 0x1234a678
        add  $11, $0, $0        # d = 0 (lots of ways to do this)
#
####


# :Example:
#
# Simplified assembler for the following line of C:
#
#  rs = ( ir >> 21 ) & 0x1f;
#
# Asm regs: rs, $8;  ir, $9.

        srl $8, $9, 21
        andi $8, $8, 0x1f
####


################################################################################
## Instruction Coding

# :PH: 3.4
# :Mv1: 4.2


 ## The Three MIPS Instruction Formats
#
# R Format:  Typically used for three-register instructions.
# I Format:  Typically used for instructions requiring immediates.
# J Format:  Used for jump instructions (not covered yet).
#
# Every integer MIPS instruction is in one of these formats.

 ## R Format
# _________________________________________________________________
# | opcode    | rs      | rt      | rd      | sa      | function  |
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
# 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
#
# Bits    Field Name    Unabbreviated Name     Typical Use
#
# 31:26:  opcode                               First part of opcode.
# 25:21:  rs            (Register Source)      Source register one.
# 20:16:  rt            (Register Target)      Source register two.
# 15:11:  rd            (Register Destination) Destination register.
# 10:6:   sa            (Shift Amount)         Five-bit immediate.
#  5:0    function                             Second part of opcode.
#

 ## I Format
# _________________________________________________________________
# | opcode    | rs      | rt      | immed                         |
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
# 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
#
# Bits    Field Name    Unabbreviated Name     Typical Use
#
# 31:26:  opcode                               Entire opcode (for I and J).
# 25:21:  rs            (Register Source)      Source register one.
# 20:16:  rt            (Register Target)      Source register two.
# 15:0:   immed         (Immediate)            Immediate value.

 ## J Format
# _________________________________________________________________
# | opcode    | ii                                                |
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
# 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
#
# Bits    Field Name    Unabbreviated Name     Typical Use
#
# 31:26:  opcode                               Entire opcode (for I and J).
# 25:0:   ii            (Instruction Index)    Part of jump target.


 ## Field Values
#
# Values for fields can be found in the following places:
#    :PH:  Page A-55 on (2nd edition) Listed by instruction.
#    :Mv2: Listed by instruction. "Official" MIPS documentation.
#
# opcode, function
#
#  Determined by the instruction.
#  For R-format instructions opcode is limited to a few values: e.g., 0, 1.
#  The function field takes on a larger range of values.
#
#  Values for opcode and function can be found in the following places:
#    :PH:  Figure A.19 (page A-54)  Compact, but perhaps intimidating at first.
#    :PH:  Page A-55 on (2nd edition) Listed by instruction.
#    :Mv2: Listed by instruction. "Official" MIPS documentation.
#
#
# rs, rt, rd
#
#  Usually register numbers represented using 5-bit unsigned integers.
#  Occasionally used as part of opcode or to specify something other
#    than a register.
#
#
# sa (Shift Amount)
#
#  A 5-bit constant usually used by shift instructions.
#
#
# immed
#
#  A 16-bit quantity.
#
#
# ii (Instruction Index)
#
#  A 26-bit quantity used to form a jump target address.


 ## R Format Examples
# _________________________________________________________________
# | opcode    | rs      | rt      | rd      | sa      | function  |
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
# 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
#
# :Syntax: ADD rd, rs, rt
        add $2, $3, $4
#
# Based on the instruction above:
#   rd -> 2,  rs -> 3,  rt -> 4
#
# Using something in :PH: Appendix A or looking up the instruction in :Mv2:
#   opcode -> 0,  function -> 0x20,  sa -> 0
#
# Machine code for the instruction above:
#     0 3 4 2 0 0x20                      (By field, each field in hexadecimal)
#   = 000000 00011 00100 00010 00000 100000    (By field, binary)
#   = 0000 0000 0110 0100 0001 0000 0010 0000  (Groups of four bits)
#   = 0x00641020                               (Hexadecimal)
#
#
# :Syntax: SLL rd, rt, sa               # Shift Left Logical
        sll $2, $3, 4
# Based on the instruction above:
#   rd -> 2,  rt -> 3,  sa -> 4
#
# Using something in :PH: Appendix A or looking up the instruction in :Mv2:
#   opcode -> 0,  function -> 0,  rs -> 0
#
# Machine code for the instruction above:
#     0 0 3 2 4 0                      (By field, each field in hexadecimal)
#   = 000000 00000 00011 00010 00100 000000    (By field, binary)
#   = 0000 0000 0000 0011 0001 0001 0000 0000  (Groups of four bits)
#   = 0x00031100                               (Hexadecimal)
#

 ## I Format Examples
# _________________________________________________________________
# | opcode    | rs      | rt      | immed                         |
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
# 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
# 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
#
# :Syntax: ORI rt, rs, immed
        ori $2, $3, 4
#
# Based on the instruction above:
#   rt -> 2,  rs -> 3,  immed -> 4
#
# Using something in :PH: Appendix A or looking up the instruction in :Mv2:
#   opcode -> 13
#
# Machine code for the instruction above:
#     0xd 3 2 4                      (By field, each field in hexadecimal)
#   = 001101 00011 00010 0000000000000100     (By field, binary)
#   = 0011 0100 0110 0010 0000 0000 0000 0100 (Groups of four bits.)
#   = 0x34620004                              (Hexadecimal)
#



################################################################################
## Pseudo Instructions

# :PH: 3.9 (Briefly under assembler heading.)

# A.K.A. Synthetic Instructions

# :Def: Pseudo Instruction
#
# An assembly language instruction that does not (necessarily)
# correspond to a machine language instruction.  The assembler will
# substitute one or two real machine instructions. Pseudo instructions
# are intended to make assembler code more readable and to support
# future ISA extensions.  (Some pseudo instructions may become real
# instructions.)
#
# When these notes are viewed in HTML or with the class Emacs package,
# pseudo instructions (those recognized by the SPIM simulator) will be
# italicized.  See examples.


 ## Some Pseudo Instructions
#
# This is not a complete list of pseudo instructions.
#
# (Note: the examples show both pseudo and real instructions.  In
#  actual assembly code you would use one or the other.)
#
# :Syntax: NOP                          # No Operation
#
#  A special NOP instruction is not needed because many instructions
#  will do nothing if register zero is used as the destination.
#  Instruction sll is the "official" NOP because its coding is all
#  zeros.
#
        nop               # Pseudo
        sll $0, $0, 0     # Real.

#
# :Syntax: MOVE rd, rs                  # Move
#          rd <- rs
#
        move $1, $2       # Pseudo
        addu $1, $0, $2   # Real

#
#
# :Syntax: LA rd, imm32                 # Load Address
# :Syntax: LI rd, imm32                 # Load Integer
#          rd <- imm32
#          Note: SPIM resolves la and li into the same instructions.
#          (That is, they do the same thing.  Use the one that better
#           conveys intent to the human assembly code reader.)
#
#
        li $2, 0x12345678   # Pseudo
        lui $1, 0x1234      # Real (First of two.)
        ori $2, $1, 0x5678  # Real (Second of two.)

        li $2, 0x12         # Pseudo
        ori $2, $0, 0x12    # Real (Assembler realizes lui not needed.)


################################################################################
## Types of Control-Transfer Instructions

# :PH:  Material covered in a different order.
# :Mv1: 4.1.3 (Brief overview.)

 ## Types of Control Transfers
#
# Branch
#   Used for loops, ifs, etc.  Most commonly used CTI.
#   Instruction specifies a condition and an address, the target.
#   If condition is true continue execution at target.
#
# Branch-and-Link
#   Used for procedure calls.
#   Instruction specifies a condition and an address, the target.
#   If condition is true save PC in $31 and continue execution at target.
#
# Jump
#   Used for returns and other purposes.
#   Instruction specifies a target
#   Continue execution at target.
#
# Jump-and-Link
#   Used for procedure calls and returns and other purposes.
#   Instruction specifies a target
#   Save PC in $31 and continue execution at target.
#
# Trap (Not covered in detail in this set.)
#   If condition true continue at specified entry in OS-prepared trap table.
#
#

 ## Conditions and Targets
#
# The methods used to specify conditions and targets vary by instruction.
# See descriptions in following sections

 ## Delayed CTI    <- PAY ATTENTION TO THIS.
#
# All of the CTIs covered here are delayed (except traps).
#
# The instruction after an ordinary delayed CTI IS ALWAYS EXECUTED.
#
        add $8, $0, $0
        j FORJOY
        addi $8, $8, 1
        addi $8, $8, 2

FORJOY:
        addi $8, $8, 4
        addi $8, $8, 8


################################################################################
## A Simple Jump Instruction

# :PH: 3.5
# :Mv2:

# Simple jump (j) covered here, others (jr,jal,jalr) covered in a
# later selection.

# Assembler :Syntax: J target
# Coding    :Syntax: J ii
#          After next instruction go to { PC[31:28], ii, 2'b0 },
#          where ii is bits 27:2 of target.
#
# In assembly language "target" is the line label for the target.
# In machine language "target", the ii field, is bits 27:2 of the
#   target address.
# Bits 31:28 of the target address must match bits 31:28 of the jump
#   instruction's address. (If not, use the jr instruction.)

# :Example:
#
# Code used to illustrate how target is determined.

FIRSTLINE:  # Address of add: 0x400000
        add $8, $0, $0
        j FORJOY
        addi $8, $8, 1
        addi $8, $8, 2

FORJOY:     # Address of addi: 0x400010
        addi $8, $8, 4
        addi $8, $8, 8

# In the code above:
#
#   FORJOY is a line label, part of the assembly language.
#   Suppose the assembler [really the linker] sets FORJOY = 0x400010


################################################################################
## A Branch Instruction

# :PH: 3.5
# :Mv2:

# One branch instruction covered here (bne), others (beq, bltz, etc.)
# covered in a later section.

# Assembler :Syntax: BNE rs, rt, target    # Branch Not Equal
# Coding    :Syntax: BNE rs, rt, offset
#           if rs != rt, after next instruction
#                        go to PC + 4 + 4 * sign_extend(offset)
#
# "target" is the line label to continue at.
# "offset" is the number of instructions to skip, starting at
#   the instruction after the branch.

# :Example:
#
# Code to illustrate the offset.

FIRSTLINE:
        add $9, $0, $0
        addi $10, $9, 1
        bne $9, $10, LINE
        addi $8, $8, 1      # Offset computed from this instruction.
        addi $8, $8, 2
LINE:
        addi $8, $8, 4      # The branch target, offset = 2 instructions.
        addi $8, $8, 8


# :Example:
#
# Use of a branch in a loop.

        add $8, $0, $0   # $8 -> 0
        addi $9, $0, 10  # $9 -> 10
        add $10, $0, $0  # $10 -> 0 (i)
LOOP:
        add $8, $8, $10  # $8 = $8 + i
        addi $9, $9, -1
        bne $9, $0, LOOP
        addi $10, $10, 1 # i = i + 1


################################################################################
## More Branch Instructions

# :PH: 3.5
# :Mv2:

# Assembler :Syntax: BEQ rs, rt, target    # Branch Equal
# Coding    :Syntax: BEQ rs, rt, offset
#           if rs == rt after next instruction
#                       go to PC + 4 + 4 * sign_extend(offset)
#
# Assembler :Syntax: BGEZ rs, target    # Branch >= Zero
# Coding    :Syntax: BGEZ rs, offset
#           if rs >= 0 after next instruction
#                      go to PC + 4 + 4 * sign_extend(offset)
#
# Assembler :Syntax: BGTZ rs, target    # Branch > Zero
# Coding    :Syntax: BGTZ rs, offset
#           if rs > 0 after next instruction
#                     go to PC + 4 + 4 * sign_extend(offset)
#
# Assembler :Syntax: BLEZ rs, target    # Branch <= Zero
# Coding    :Syntax: BLEZ rs, offset
#           if rs <= 0 after next instruction
#                      go to PC + 4 + 4 * sign_extend(offset)
#
# Assembler :Syntax: BLTZ rs, target    # Branch < Zero
# Coding    :Syntax: BLTZ rs, offset
#           if rs < 0 after next instruction
#                     go to PC + 4 + 4 * sign_extend(offset)


################################################################################
## More Jump Instructions

# :PH: 3.5
# :Mv2:

# Assembler :Syntax: JAL target   # Jump and link.
# Coding    :Syntax: JAL ii
#          $31 <- PC + 8
#          After next instruction go to { PC[31:28], ii, 2'b0 },
#          where ii is bits 27:2 of target.
#          Note: $31 a.k.a. $ra
#
# :Syntax: JR rs                  # Jump register.
#          After next instruction go to address in rs.
#
# :Syntax: JALR rs                # Jump and link register.
#          $31 <- PC + 8
#          After next instruction go to address in rs
#          Note: $31 a.k.a. $ra
#


# :Example:
#
# Use of j instruction.

        # Jump to LINEX
        j LINEX
        nop
        nop

LINEX:
        addi $t2, $t2, 1


# :Example:
#
# Use of jal instruction.

LINEA:  # LINEA = 0x12345678
        # Instruction below is stored at (PC=) 0x12345678 
        # Set $ra -> PC + 8 = 0x12345680 and jump to LINEX
        jal LINEX
        nop
        nop

LINEX:
        addi $t2, $t2, 1


# :Example:
#
# Use of jr instruction.

        # Load the address of a line into $to.
        la $t0, LINEX

        lui $t0, hi(LINEX)
        ori $t0, $t0, lo(LINEX)


        # Jump to the line.
        jr $t0
        nop
        nop

LINEX:
        addi $t2, $t2, 1


# :Example:
#
# Use of jalr instruction.

LINEA:  # LINEA = 0x12345678
        la $t0, LINEX   # Note: this resolves to two instructions.
        # Instruction below is stored at (PC=) 0x1234