Assembly Language

1
Assembly Language

• Chapter 5
• CIS-2261
• Foundations in Computer Science

2
Homework Chapter 5
Exercises 1, 2, 3, 4, 6, 8, 9, 11, 12, 14, 17,
18, 24, 25
3
OVERVIEW

• Assemblers
• Decimal I/O and Immediate Addressing
• Symbols
• Assignment Statements

4
Introduction

• The Assembly level was invented to relieve the
  programmer of the tedium of programming in binary
• We just covered the operation of a computer at
  level 3
• This chapter focuses on how level 5 abstracts the
  details at level 3

5
Assemblers

• Two types of bit patterns at level 3
• Bit pattern for instructions
• Bit pattern for data
• The types are a direct consequence of the von
  Neumann design
• Data and program share same memory with a binary
  representation for each

6
Assemblers

• Assembly language contains two types of
  statements corresponding to each bit pattern
• Mnemonic statements correspond to the
  instruction bit patterns
• Pseudo-operations correspond to the data bit
  patterns

7
Instruction Mnemonics
Level 3 0000 1000 0000 0000 1001 1010
Opcode
Reg Spec
Address mode
Level 5 LOADA h009A, i
Instruction Meaning Load accumulator with the
contents of the Operand Specifier (h009A)
8
Instruction Mnemonics
Level 3 0000 1001 0000 0000 1001 1010
Opcode
Reg Spec
Address mode
Level 5 LOADA h009A, d
Instruction Meaning Load accumulator with the
contents from memory pointed to by the Operand
Specifier (A lt memh009A)
9
Instruction Mnemonics
Level 3 0000 1011
Opcode
Reg Spec
Address mode
Level 5 LOADA ,X
Instruction Meaning Load accumulator with the
contents from memory pointed to by the Base
Index Register (memBX)
10
Examples of LOADR
Machine Language
Assembly Language
0000 1000 0000 0000 1001 1010 LOADA h009A,
i 0000 1001 0000 0000 1001 1010 LOADA h009A,
d 0000 1010 0000 0000 1001 1010 LOADA h009A,
s 0000 1011 LOADA , x 0000 1100 0000 0000
1001 1010 LOADX h009A, i 0000 1101 0000 0000
1001 1010 LOADX h009A, d 0000 1110 0000 0000
1001 1010 LOADX h009A, s 0000 1111 LOADX ,
i
11
Pep/6 Instruction Set
12
Pep/6 Instruction Set
13
Pep/6 Instruction Set
14
Pep/6 Instruction Set
15
Pep/6 Instruction Set

• DECI, DECO, HEXO are new instructions available
  only at level 5
• Instructions are implemented by the operating
  system

16
Pseudo-Operations

• Pseudo-ops are assembly language statements
• Do not have opcodes
• Do not correspond to any of the 32 instructions
• Pep/6 has 8 pseudo-ops

17
Pseudo-Operations
18
Program 5.1
Assembler Input CHARO h0007, d Output
H located at mem0007 CHARO h0008, d
Output i located at mem0008 STOP .ASCII
/Hi/ Generate contiguous bytes
of ASCII .END Assembler Output E1 00 07 E1 00 08
00 48 69 zz Program Output Hi
19
Assemblers

• An assembler is a program that translates
  assembly level (level 5) code to machine level
  code (level 3)
• Developed to relieve the early programmer of the
  need to hand assemble his/her program
• Assembler input is an assembly program (source
  code) and output is that program translated to
  machine level (object code)
• Assembler does not execute the program just
  gets it ready for the loader to load

Input
Output
Processing
CHARO h0007, d CHARO h0008, d STOP
.ASCII /Hi/ .END
E1 00 07 E1 00 08 00 48 69 zz
Assembler
Object code
Source code
20
Program 5.2
Assembler Input CHARI h000D, d
Input first character CHARI h000E, d
Input second character CHARO h000E, d
Output second character CHARO h000D, d
Output first character STOP .BLOCK
d1 Storage for first character .BLOCK
d1 Storage for second character .END Assembler
Output D9 00 0D D9 00 0E E1 00 0E E1 00 0D 00 00
00 zz Program Input Hi Program Output Hi
.BLOCK - generate a block of bytes
d denotes decimal character expected
21
Program 5.3
.WORD - assembler generates one word instead of
d bytes AND the programmer can specify the
content of the word
22
Assembler

• .WORD and .BYTE pseudo-op similar to .BLOCK
  except
• .WORD - assembler generates one word instead of
  d bytes AND the programmer can specify the
  content of the word
• .BYTE - assembler generates one byte instead of
  d bytes AND the programmer can specify the
  content of the byte

23
Using the Pep/6 Assembler

• First the assembler is loaded into main memory
  and the applications program is taken as the
  input file
• The output of this run is the machine language
  version of the applications program
• This machine language is then loaded for running

24
Using the Pep/6 Assembler
Input
Output
Processing
CHARI h000D, d CHARI h000E, d CHARO
h000E, d CHARO h000D, d STOP .BLOCK
d1 .BLOCK d1 .END
Assembler (machine language)
D9 00 0D D9 00 0E E1 00 0E E1 00 0D 00 00 00 zz
FIRST RUN
D9 00 0D D9 00 0E E1 00 0E E1 00 0D 00 00 00
zz
up
pu
SECOND RUN
25
General Rules for Pep/6 Assembler

• Place at least one space after the mnemonic or
  dot command
• Use any combination of upper and lower case
  letters
• The assembler will protest if you made any errors
  and will not generate any object code
• Can generate an assembler listing if desired

26
Program 5.2 - Source Code
27
Program 5.2 - Assembler Listing
28
Cross Assemblers

• An assembler that produces the object code
  program for a different machine than the one
  running the assembler
• Brand X producing the object code for download to
  Brand Y
• Brand X is the Host machine
• Brand Y is the Target machine
• Often used if programming small special purpose
  machines

29
OVERVIEW

• Assemblers
• Decimal I/O and Immediate Addressing
• Symbols
• Assignment Statements

30
Immediate Addressing

• Direct addressing uses the Operand Specifier as
  the address in main memory to find the operand
• Oprnd MemOprndSpec
• Immediate addressing uses the Operand Specifier
  to actually be the Operand
• Oprnd OprndSpec

31
Program 5.1 - Direct Addressing
32
Program 5.1 Assembler Listing
33
Program 5.4 - Immediate Addressing
34
Program 5.4 - Assembler Listing
c stands for a character constant
Assembler translates CHARO c/H/, i as E00048
1110 0000 0000 0000 0100 1000
35
Decimal I/O

• Still need to find a better way of doing decimal
  I/O
• Right now pretty hard to do
• One character at a time
• Need to convert
• New instructions for handling decimals!
• DECI
• DECO

36
Decimal I/O - Program 5.5
DECI and DECO are really handled by the operating
system
37
Decimal I/O

• Previous program placed all data at the top of
  the program
• First instruction is an unconditional branch
  around the data
• Can now change code without having to refigure
  everything!
• Notice the operation of the PC the von Neumann
  execution cycle (fetch, increment, execute,
  repeat)

38
Decimal I/O

• Operating System provides the DECI and DECO
  instructions
• DECI converts a sequence of ASCII characters to a
  single Word in twos complement
• DECO does just the opposite - twos complement
  word to sequence of ASCII
• RANGE??

39
Hexadecimal Output

• Third instruction implemented by the operating
  system
• No Hex input available
• HEXO outputs one Word of information as four
  hexadecimal digits

40
HEXO - Program 5.6
41
Disassemblers

• Assemblers use one-to-one mapping to transform an
  assembly program into a machine program
• Translation process is unique and not invertable
• Disassembler tries to recover the source program
  from the object program
• not 100 successful
• Prone to ambiguity errors

42
OVERVIEW

• Assemblers
• Decimal I/O and Immediate Addressing
• Symbols
• Assignment Statements

43
Symbols

• Talked about branching around data at the
  beginning of a program
• Still difficult process because you have to count
  in hexadecimal
• If you modify your data section, you have to go
  back in and modify all of the instructions that
  reference the changed data section
• ughhh!

44
Symbols

• We need to develop a way for the assembler to
  associate a memory address with data
• USE SYMBOLS
• Use symbols within the assembly code to reference
  the data instead of using the actual address
• Assembler now needs a look-up table to reference
  a symbol to a specific location in memory
• Build the symbol table during the assembly process

45
Symbols

• Use syntax rules for symbols that are similar to
  C
• First character is a letter
• Following characters must be letters or digits
• Maximum length is 8
• Terminate a symbol by using a colon
• Place the symbol at the beginning of the line it
  references
• No spaces

Defines the symbol num and allocates two bytes
for it
num .BLOCK d2
46
Symbols

• When the assembler detects a symbol
• Store the symbol and its value in a symbol table
• The value is actually the memory address of the
  symbol
• When referring to the symbol in the code, do not
  include the colon

47
Program 5.7
48
Symbols

• Remember - the value of a symbol is an address,
  not the content of the cell at that address
• The value of the cell will not normally be known
  until runtime unless the symbol is a constant
• The symbol is reserving space for the runtime
  variable
• Also - symbols are easier to read and understand
  than straight hex code

49
Program 5.8
.EQUATE acts like a constant definition
50
OVERVIEW

• Assemblers
• Decimal I/O and Immediate Addressing
• Symbols
• Assignment Statements

51
Compilers

• Compiler translates high-order language programs
  into lower level programs
• Some compilers translate directly to machine
  language
• other compilers translate to assembly and require
  an assembler to translate to machine
• A compiler is a program like any other program
• Must execute as machine level code
• Input to compiler is called the source program
• Output of the compiler is called the object
  program

52
Compilers

• Mappings from high-order to assembly is not
  typically one-to-one
• Normally the mapping will be one-to-many
• Not all high-order elements are assembled
• include - used to tell the compiler how to make
  the correct interface to the operating system

53
Compilers

• Its time to start looking at how a compiler would
  translate high-order language into lower level
  languages!!
• Lets start simple - how to translate an output
  statement

54
Program 5.10
High -Order Language (Source Code) include
ltiostream.hgt main() cout ltlt Love
This is essentially a single statement program.
55
Program 5.10 Object Code
Translation of a simple single C statement This
is not the only possible translation!!
56
Program 5.10 Machine Listing
57
Compilers
Input
Output
Processing
include ltiostream.hgt main() cout ltlt
Love
E0 00 4C E0 00 6F E0 00 76 E0 00 65 00 zz
Compiler
Level 6 - Level 3
Machine code
Source code
CHARO c/L/, i CHARO c/o/, i CHARO
c/v/, i CHARO c/e/, i STOP
.END
include ltiostream.hgt main() cout ltlt
Love
Compiler
Level 6 - Level 5
Source code
Assembly code
58
Variables and Types

• Every C variable has three attributes
• name
• type
• value
• Compiler reserves memory cells in the machine
  language program for each and C variable
• A variable in C is simply a memory cell in
  machine/assembly code

59
Variables and Types

• Level 6 programs refer to variables by names
  which are C identifiers
• Level 3 programs refer to them by addresses
• Value of the variable is the value in the memory
  cell at the address associated with the C
  identifier
• Compiler must remember the which address
  corresponds to which C variable name
• Use a symbol table very similar to that used by
  an assembler

60
Variables and Types

• Compiler symbol table more complicated than table
  used by an assembler
• More than 8 characters
• Must store variable type as well as associated
  address

61
Program 5.11 - Assignment Statements
High -Order Language (Source Code) include
ltiostream.hgt char ch int i main() cin gtgt ch
gtgt i i 5 ch
cout ltlt ch ltlt endl ltlt i
62
Program 5.11 Assembly List
char ch
int i
Endl h000A implied
63
Program 5.11 - Compiler Symbol Table
64
Code Generation in Program 5.11
Compiler Translates To
Cin gtgt ch gtgt I
Chari h0003, d Deci h0004, d
LoadA I, d AddA d5, i StoreA I, d
I 5
LdbytA ch, d AddA d1, i StbytA ch, d
Ch
65
Type Compatibility

• Compilers also need to keep track of different
  variable types and not allow certain operations
  to be performed on them if the type is incorrect
• example should never be able to take modulus of
  a floating point number
• Assembly programming would allow this to happen
  even though the result would be garbage

66
Program 5.12 - Source Code
High -Order Language (Source Code) include
ltiostream.hgt const int bonus 5 int
exam1 int exam2 int exam3 main() cin gtgt
exam1gtgt exam2 score (exam1
exam2)/2 bonus cout ltlt score
ltlt score
67
Program 5.12 - Assembly Listing
68
Program 5.12 - Symbol Table
Note that there is no assembly code for bonus.
You dont need to reserve memory space for a
constant. Instead put it in the table and store
its value directly at the using location. That is
how .EQUATE is handled by the compiler.
69
END Chapter 5