Assembler Language Tutorial for Mainframe Programmers

Assembler/Session 1
Assembler LanguageAssembler Language
forfor
Mainframe ProgrammersMainframe Programmers

Objectives
• Familiarize with IBM 370 Assembly Language
ObjectivesObjectivesObjectivesObjectives
Assembler/Session 1

SESSION 1SESSION 1
Day 1Day 1
Introduction
SESSION 2SESSION 2
Day 1Day 1
Addressing
SESSION 3SESSION 3
Day 2Day 2 Machine Instructions
Assembler/Session 1
COURSE SCHEDULECOURSE SCHEDULE

Writing a complete program
SESSION 4SESSION 4
Day 3Day 3
Program Sectioning
SESSION 5SESSION 5
Day 3Day 3
Assembler Directives
SESSION 6SESSION 6
Day 3Day 3
Assemble and link programSESSION 7SESSION 7
Day 4Day 4
COURSE SCHEDULE
Assembler/Session 1

Macro LanguageSESSION 8SESSION 8
Day 4Day 4
Other TopicsSESSION 9SESSION 9
Day 5Day 5
Assembler/Session 1
COURSE SCHEDULE

SESSION 1
Assembler/Session 1

Objectives
• An assembler language is a symbolic form of
machine language
• Assembler translates assembler language
program to machine language
• An assembler program consists of many
statements
• In general, one assembler language statement
corresponds to one machine language
instruction
INTRODUCTIONINTRODUCTION
Assembler/Session 1

Objectives
STATEMENT FORMATSTATEMENT FORMAT
1 10 16 30
label operation operands comments
e.g..
INIT1 LA R5,4 ;INITIALISE REGISTER 5
Rules for choosing labels:Rules for choosing labels:
• maximum 8 characters
• Alphabets, digits, @, #, $
• First character should not be a digit
• label should begin in column 1
Assembler/Session 1

Objectives
Col1 Col10 Col.16
L 2,A
A 2,B
ST 2,ANS
…..
…..
A DC F’15’
B DC F’20’
ANS DS F
Sample programSample program
Assembler/Session 1

Objectives
Operation
• One of the 200 M/C instruction mnemonics (eg. MVC)
Operand
• can be a register or memory location
Continuing a statement
• Place any character in column 72 of the line to be continued
• Continue the statement from column 16 of next line
• Maximum 2 continuation lines for a statement
Assembler/Session 1

Objectives
Comment Statement
• * in column 1
• Any text in columns 2 - 71
Note : Fields separated by one or more
blanks
Assembler/Session 1

Objectives
TYPES OF INSTRUCTIONSTYPES OF INSTRUCTIONS
1. Machine Instructions
2. Assembler Instructions (Directives)
3. Macro Instructions
Assembler/Session 1

Objectives
REGISTERSREGISTERS
Registers are storage areas inside the processor
Advantages:
- No need to retrieve data from main storage
(saves time)
- Shared resource that allows inter
communication between programs
Assembler/Session 1

Objectives
REGISTERSREGISTERS
General purpose registers:
* 16 registers available
* Numbered 0 - 15
* Holds 32 bits (4 bytes) of data (1 Full word)
Floating point registers:
* 4 registers available
* Numbered 0,2,4,6
* Holds 64 bits (8 bytes) of data
Note : The registers 0, 1, 13, 14 and 15 are reserved for special purpose
By IBM convention these registers are used for calling
subprograms
Assembler/Session 1

Objectives
DATA REPRESENTATIONDATA REPRESENTATION
Binary fields
- Always fixed in length, either 2 or 4 bytes
(Full word or Half word)
- Negative numbers stored in 2’s complement form
Examples:
A DC H’295’ 01 27
B DC H’-75’ FF 35
Assembler/Session 1

Objectives
2’s complement form2’s complement form
Assembler/Session 1
How to identify a negative number?How to identify a negative number?
-- Leading bit contains a 1 (In Hex 8 to F)Leading bit contains a 1 (In Hex 8 to F)
How to convert to a negative numberHow to convert to a negative number??
-First switch the bits (1 to 0 , 0 to 1)First switch the bits (1 to 0 , 0 to 1)
-Finally add 1Finally add 1

Objectives
Boundary requirementsBoundary requirements
Assembler/Session 1
Full word – Should begin in a full word boundary
(Achieved by aligning with 0F)
Half word – Should begin in a half word boundary
(Achieved by aligning with 0H)
How to find:
The starting address of Full word should end with
0, 4, 8 or C and Half words should end with 0, 2, 4,
6, 8, A, C or E

Objectives
Characters
- One byte (EBCDIC form)
- Character representation of decimal digits is called
Zoned Decimal (first nibble is zone and next is digit)
Zone digit Zone Code
0 - 9 + C, A,E,F
- D, B
+, - , blank Blank F
Assembler/Session 1

Objectives
Floating Point Numbers
- Always fixed in length, 4, 8 or 16 bytes
(Full word, double word, double double word)
- Left most bit represents sign
(0 - positive; 1 - negative)
- Next 7 bits represent exponent
- Remaining bytes represent the fraction
Assembler/Session 1

Objectives
Decimal numbers ( Packed Decimal representation)
- Each byte but the rightmost has 2 decimal digits (0-9)
- The right most byte contains a digit in the left half and
a sign indicator in the right
Sign indicator: C- Positive
D - Negative
Example: 753 - 7 5 3 C
Assembler/Session 1

Objectives
Addressing Operands
• Register addressing
• Base, displacement addressing
• Base, index and displacement addressing
Assembler/Session 1

Objectives
INSTRUCTION FORMATS
RR opcode R1 R2
SI opcode I2 B1 D1
SS opcode L B1 D1 B2 D2
SS opcode L1 L2 B1 D1 B2 D2
RX opcode R1 X2 B2 D2
RS opcode R1 R3 B2 D2
Assembler/Session 1

Objectives
Addressing RX Operands:
Implicit format:
L 3,VAR
Explicit format:
L 3,100(0,12)
Register Displacement Index reg Base reg
Assembler/Session 1

SESSION 2SESSION 2
Addressing
Assembler/Session 2

ObjectivesSTORAGE DEFINITIONSSTORAGE DEFINITIONS
Two ways to define fields :
1. Define a field and initialize the data in it using
the DC assembler directive
2. Define a field without initializing using the DS
assembler directive
Assembler/Session 2

Objectives
STORAGE DEFINITIONSSTORAGE DEFINITIONS
Format:
label {DS/DC} dtLn’value’
where :
label : Label used to name the field (optional)
d : Duplication factor (optional)
t : Type of data ( required)
Ln : The letter ‘L’ followed by the length of the field in
bytes (optional)
value : Represents the value enclosed in apostrophes
Assembler/Session 2

Objectives
Examples:
ALPHA DC C’ABC EF’
FLDS DS 3CL2
H1 DC H’29’
F2 DC F’-10’
F1 DC X’03’
F3 DC PL4’-72’
Note : for character constants truncation or padding is to
the right and for almost all others it is to the left.
Assembler/Session 2

Objectives
DC TYPES
Type Implied Alignment Data Representation
Length
C - None Character
X - None Hex digits
B - None Binary digits
F 4 Full word Binary
H 2 Half word Binary
E 4 Full word Floating point
D 8 Double word Floating point
L 16 Double word Floating point
P - None Packed decimal
Assembler/Session 2

Objectives
Data Representation in other languages:
Assembler FORTRAN COBOL PASCAL BASIC
Language
DC Type
C Character Display String String
F, H Integer COMP Integer Integer
E Real COMP-1 Real Single
precision
D Double COMP-2 Real Double
Precision Precision
X, B Logical N/A Boolean Hex
P N/A COMP-3 N/A N/A
Assembler/Session 2

Objectives
STORAGE DEFINITIONS
Literals
• A literal is a constant preceded by an equals sign ‘=‘.
• Can be used as a main-storage operand but not as a
destination field of an instruction
• Causes assembler to define a field that is initialized with
the data specified
• All constants defined by literals are put by the assembler
in a literal pool, usually at the very end of the program
(Unless changed by LTORG instruction)
L R4,=F’1’
Assembler/Session 2

Objectives
Exercise 1 Q 1 and Q2.
2.What will happen in the following cases
DC CL5’123’
DC CL5’123456’
DC X’A1245’
DC XL2’A1245’
DC XL5’A1245’
DC F’19’
DC FL1’513’
Assembler/Session 2

Objectives
EQU (Assembler directive)
• The EQU statement is used to associate a
fixed value with a symbol
R4 EQU 4
DRBACK EQU OUT+25
Assembler/Session 2

Objectives
ESTABLISHING ADDRESSABILITYESTABLISHING ADDRESSABILITY
• By establishing the addressability of a
coding section, you can refer to the
symbolic addresses defined in it in the
operands of machine instruction
• Assembler will convert the implicit
addresses into explicit addresses
(base - displacement form)
Assembler/Session 2

Objectives
To establish the address of a coding section :
• Specify a base address from which the
assembler can compute displacements
• Assign a base register to contain this base
address
• Write the instruction that loads the base
register with the base address
Note: The base address should remain in the base
register throughout the execution of the program
Assembler/Session 2

Objectives
Establishing Base Register
The USING and DROP assembler instructions
enable one to use expressions representing
implicit addresses as operands of machine
instruction statements, leaving the assignment of
base registers and the calculation of
displacements to the assembler
USING - Use Base Address Register
- allows one to specify a base address and assign
one or more base registers
Assembler/Session 2

Objectives
To use the USING instruction correctly, one should know :
• which locations in a coding section are made addressable
by the USING statement
• where in a source module you can use these established
addresses as implicit addresses in instruction operands
Format:
symbol USING base address,basereg1| basereg2|,..
e.g. USING BASE,9,10,11
USING *,12
Assembler/Session 2

Objectives
Range of a USING instruction:
• The range of a USING instruction is the 4096
bytes beginning at the base address specified in
the USING instruction
Domain of a USING instruction
• The domain of a USING instruction begins
where the USING instruction appears in a source
module to the end of the source module
Assembler/Session 2

Objectives
The assembler converts implicit address references into
their explicit form:
• if the address reference appears in the domain of a
USING instruction
• if the addresses referred to lie within the range of the
same USING instruction
Guideline:
• Specify all USING instructions at the beginning of the
source module
• Specify a base address in each USING instruction that lies
at the beginning of each control section
Assembler/Session 2

Objectives
RELATIVE ADDRESSINGRELATIVE ADDRESSING
• Relative addressing is the technique of addressing
instructions and data areas by designating their location
in relation to the location counter or to some symbolic
location
ALPHA LR 3,4
CR 4,6 ALPHA+2 or BETA-4
BCR 1,14
BETA AR 2,3
Note : Always avoid using relative addressing
Assembler/Session 2

SESSION 3 & 4
Machine Instructions
Assembler/Session 3 & 4

Objectives
HANDLING CHARACTER DATAHANDLING CHARACTER DATA
Move Character Instruction (MVC)
• Copy data from one place in memory to another
Format : MVC operand1,operand2
S1(L), S2 - implicit
D1(L,B1),D2(B2) - explicit
e.g...
MVC INPUT(5),OUTPUT

Objectives
Move Immediate Instruction (MVI)
• Can move only one byte of constant data to a field
Format : MVI operand1,operand2
S1,I2 - implicit
D1(B1),I2 - explicit
e.g..
MVI CTL,C’B’

Objectives
Advanced Techniques
1. Explicit lengths and relative addressing
MVC PAD+6(4),=CL4’ ‘
PAD DS CL10
2. Overlapping fields and the MVC instruction
MVC FLDB,FLDA
FLDA DC C’A’
FLDB DS CL3
Limitation of MVC : Can only move 256 bytes

Objectives
Moving more than 256 characters: MVCL instruction
Uses 2 pairs of even-odd pair of registers
Format : MVCL R1,R2 (Both are even registers)
Reg R1 – Address of destination R1+1 – Length
Reg R2 - Source R2+1 – Padding character (1st
8 bits) and Length
Eg: LA 2,Q
LA 3,2000
LA 4,P
LA 5,1500
MVCL 2,4

Objectives
Comparison Instructions
• Compares 2 values - the values are found in fields, in
registers or in immediate data
CLC - Compare logical character
e.g. CLC FLDA,FLDB
CLI - Compare logical immediate
e.g. CLI FLDA,C’K’

Objectives
Exercise 2 Q1 and Q2
2. What will be the effect of the following instructions :
MVI OUTAREA,C’ ‘
MVC OUTAREA+1(132),OUTAREA
OUTAREA DS 133C

Objectives
BINARY INSTRUCTIONSBINARY INSTRUCTIONS
Three types of binary instructions
•Full word
•Half word
•Register
The Binary Move Instructions
L, LH, LR ,ST, STH
Type : R,X Register and indexed storage
e.g... L 5,FULL LR 5,7
STH 7,HALF

Objectives
Note : Do not mix up the instruction types and field types
e.g.
LH 5,FULL - right half of Reg 5 gets 1st 2 bytes at FULL
L 6,HALF - Reg 6 gets 4 bytes starting from HALF
ST 3,RES - 4 bytes of reg 3 are stored starting from RES
RES DS H
HALF DC H’15’
FULL DC F’8’

Objectives
Binary Addition (A, AH and AR)
• Fixed-point overflow occurs when the sum will not
fit in the receiving register
• Type R-X
e.g.
A 5,FULL
AH 6,HALF
AR 7,3

Objectives
Binary Subtraction (S, SH and SR)
• Type R-X
e.g.
S 5,FULL
SH 6,HALF
SR 7,3

Objectives
Binary comparisons (C, CH and CR)
e.g.
C 5,FULL
CH 6,HALF
CR 7,3
Condition code set as HIGH, LOW or EQUAL

ObjectivesBinary Multiplication (M, MR, MH)
Format : M op1,op2
op1 : An even numbered register; refers to an even-odd
pair of registers
(any register in case of half word format)
op2 : storage area (full word/half word/register)

Binary Multiplication (M, MR, MH) ...
Function : The value in OP2 is multiplied by the
value in the odd register of the even-odd pair and the result
placed in even-odd registers
(For half word format : The half word specified in OP2 is
multiplied by the value in OP1 and result stored in OP1.)

Objectives
Binary Division (D, DR)
Format: D op1,op2
Type : R-X / R-R
Op1 : An even numbered register. It refers to an even-odd pair
of registers. The pair holds the double word to be
divided. The even register receives the remainder; the
odd register receives the quotient.
e.g. D 4,FULL

Objectives
BC and BCR Instructions
• instructions that do or do not branch depending on
the value of the condition code
Format : BC M1,S2
BCR M1,R2
e.g. BC B’1001’,BRPTA
will cause a branch to the instruction named
BRPTA, if at the time the instruction is executed,
the condition code is 0 or 3.

Objectives
BRANCHINGBRANCHING
A branch causes execution to continue at some
other instruction in the program
• Branch conditions : Arithmatic B, BZ,BP,BM,
BNZ,BNP,BNM,BO,BNO
• Comparison BH, BL, BE, BNH, BNL,BNE
e.g : CLI FLDA,C’K’
BNL GOOD

Objectives
CONDITION CODE PROCESSINGCONDITION CODE PROCESSING
• condition code occupies 2 bits of PSW
• condition code is set by each of a number of instructions
• condition code is an extremely important intermediary
between arithmetic instructions and conditional branch
instructions
• very important in implementing control structures
CC Arithmetic Comparison
0 Zero First operand = Second operand
1 < Zero First operand < Second operand
2 >Zero First operand > second operand
3 Overflow Not set

Objectives
LPR, LNR and LCR Instructions
Format: LPR,LNR or LCR R1,R2
LPR - Load positive register (Loads into R1 the
absolute value of R2)
LNR Load Negative register (Loads into R1 the
negative of absolute value of R2)
LCR Load complement register (Loads opposite sign
of the value in R2)
Note: R1 and R2 can be the same

Objectives
BIT MANIPULATIONSBIT MANIPULATIONS
Operation S-I S-S R-R R-X
OR OI OC OR O
AND NI NC NR N
Exclusive OR XI XC XR X
e.g... OI FLDA,X’0F’
NR 5,7
X 9,FULL

Objectives
OR Second 0 1 AND Second 0 1
First 0 0 1 First 0 0 0
1 1 1 1 0 1
Exclusive OR
Second 0 1
First 0 0 1
1 1 0

Objectives
Testing individual bits - Test under mask (TM)
TM S1,I2
Function : The bits of S1 ( a single byte) are tested
under the control of the mask in I2 and condition
code is set as ‘all zeroes’, all ones’ or ‘mixed’
e.g. TM EMP,B’00000101’
BNM NEXT

Objectives
Bit Shifting Instructions
SLL, SLDL Left logical
SRL, SRDL Right logical
(No condition code set)
SLA, SLDA Left arithmetic
SRA, SRDA Right arithmetic
(Sign bit not affected and condition code set)
e.g. SLL 5,1
SRDA 4,5

Objectives
Bit Shifting Instructions
Condition code setting for arithmetic shift
instructions
0- Result is zero
1- Result is negative
2- Result is positive
3- Overflow generated
Overflow is generated when a bit other than the sign
bit is shifted out

Objectives
BIT MANIPULATIONS
Translations
• To translate from one bit combination to another
Format : TR S1(L),S2 or S1,S2
S1 : The field whose data is to be translated
S2 : A 256-byte translation table
Function : The value of the original byte is used as a
displacement into the translation table. The byte found there
replaces the original byte.
e.g. TR WORK,XTABLE
If the source byte is x’40’ (Space), then the displacement into
the table is 64. The value in the table at displacement 64 will
be replacing the source.

Objectives
BIT MANIPULATIONS
Translations
1 byte - 256 possible combinations
x’00’,x’01’, x’02’, x’03’,…………..x’0F’
x’10’,x’11’,x’12’,…………………..x’1F’
…………………………………………..
x’F1’,x’F2’,x’F3’,…………………x’FF’
The table should start with replacement byte for
x’00’ and end with replacement for x’FF’

Objectives
BIT MANIPULATIONS (TRT)
Translations - TRT (Translate and test register)
-Similar to TR but the source is not changed
-Table is searched similar to TR taking the displacement
into the table
-Usually employed for editing purposes
-The characters we need to search will have non zeros
(x’00’) but other characters will be x’00’.
-Source is searched one character at a time from left to
right
-The first nonzero match in the table halts the
instruction
-Condition code is set to 1 if match found before last
byte, 2 if found at the last and 0 if not found
-Loads address of source operand if found in last 24 bits
of register 1, value from the table into last bit of register
2. No bits are changed in both the registers

Objectives
BIT MANIPULATIONS (TRT continued)
Translations - TRT (Translate and test register)
This example searches for a period X’4B’
The period 4B is decimal 75. So the X’4B’ is placed at the
76th position in the table. (Any non zero character may
be placed in the table
Table should be declared as follows:
TABLE DC 75X’00’
DC X’4B’
DC 180X’00’

Objectives
Numeric ConversionsNumeric Conversions
1. Conversion to binary (CVB)
Format: CVB operand1,operand2
operand1 : Register
operand2 : a double word (containing
valid packed decimal number)
e.g. CVB 5,DOUBLE
Use : Character data -(PACK)->Packed decimal-(CVB)->
binary

Objectives
2. Conversion from binary (CVD)
Format: CVD operand1,operand2
operand1 : Register
operand2 : a double word
e.g. CVD 5,DOUBLE
Use : Binary-(CVD)->Packed decimal-(UNPK)->
Character data

Objectives
3. Conversion from Zoned decimal to packed
(PACK) (SS instruction)
Format: PACK operand1,operand2
operand1 : Packed decimal
operand2 : Zoned Decimal
e.g. PACK PACKED(3),ZONED(5)

Objectives
4 Packed decimal to Zoned decimal (UNPACK)
Format: UNPACK operand1,operand2
operand1 : Zoned decimal
operand2 : Packed decimal
e.g. UNPACK ZD(5),PACKED(2)

Objectives
Relation between CVD,CVB,PACK and UNPACKRelation between CVD,CVB,PACK and UNPACK
Binary inBinary in
RegisterRegister
PackedPacked
DecimalDecimal
ZonedZoned
DecimalDecimal
CVBCVB
PACKPACK InputInput
CVDCVD UNPKUNPK OutputOutput

Objectives
Example code for Different conversionsExample code for Different conversions
PACK PNUM(8),START(3)PACK PNUM(8),START(3)
CVB 7,PNUMCVB 7,PNUM
A 7,=F’1’A 7,=F’1’
CVD 7,PNUMCVD 7,PNUM
UNPK ANS(3),PNUM(8)UNPK ANS(3),PNUM(8)
……
……
START DC C’125’START DC C’125’
ANS DS CL3ANS DS CL3
PNUM DS DPNUM DS D

Objectives
Packed decimal operationsPacked decimal operations
SS format - OPCODE D1(L1,B1),D2(L2,B2)
AP - Add packed
SP - Subtract packed
ZAP - Zero and add packed
MP - Multiply packed
DP - Divide packed
CP - Compare packed
Note: All these operations ignore the decimal places. You have to track the
decimal places and edit it with ED and EDMK instructions

Objectives
Advanced instructions:
SRP - Shift and Round packed OPCODE D1(L,B1),D2(B2),I3
First operand - Memory location including length
Second operand - Direction and number of places to shift
Third operand - Whether to round or not
-------------------------------------------------------------------------
Second operand, <= 32, left shift is done and 33 to 64 right shift is done.
Number for right shift = ( 64 - number of digits to be shifted)
(No rounding is involved in left shift

Objectives
Advanced instructions: (SRP continued)
NUM is a 5 byte packed decimal number and contains 001234567C.
What is the value in number after each of these instructions?
1. SRP NUM(5),2,0
2. SRP NUM(5),62,0
3. SRP NUM(5),62,5
4. SRP NUM(5),60,5

Objectives
Advanced instructions:
MVZ - Move Zone (Moves the first half of each byte)
MVN - Move numeric (Moves the second half of each byte)
MVO - Move with offset
EG: Multiply A by 100 where value of A is 123
MVC TEMP(3),A
MVN TEMP+2(1),=X’00’
MVZ TEMP+3(1),=X’00’
MVN TEMP+3(1),A+2
A DC PL3’123’
TEMP DS PL4

Objectives
Editing the output for printingEditing the output for printing
ED and EDMK instructions ( D1(L,B1), D2(B2)) (Pattern and PD
number)
Patterns:
x’20’ - Digit selector
x’21’ - Significance selector
x’22’ - Field separator x’60’ - Sign indicator
Pattern and the packed decimal number processed from left 1 byte at a time
X 0 1 2 3 4 5 6 C (Instruction: ED P(12),X)
Fill Character
P 40 20 20 6B 20 21 20 4B 20 20 60 40 (Before execution)
P 40 40 F1 6B F2 F3 F4 4B F5 F6 40 40 (After execution)
1 , 2 3 4 . 5 6 (Last 2 bytes spaces since
number is positive)
……
……

Objectives
Values being
examined
Action taken
Pattern
byte
PDdigit Newpattern New state of
SI
Digit
selector
0
1-9
Fill character
digit in
EBCIDIC
Off
On
Significanc
e starter
0
1-9
Fill character
digit in
EBCIDIC
On
On
Field
seperator
None Fill character Off
When the
significant
indicator is off
Anyother
byte
Digit
selector
0-9 digit in
EBCIDIC
On
Significanc
e starter
0-9 digit in
EBCIDIC
On
Field
seperator
When the
significant
indicator is on
Anyother
byte
None Pattern byte
notchanged
On

Objectives
-ED and EDMK can detect the difference between significant and non signi
ficant digits ie between leading and non leading zeros
- Significance starter forces all subsequent digits to be considered significant
-When significance indicator is off and detection of a significant digit turns it
on, the address of that significant digit placed in 8-31 of register 1 by EDMK
-EDMK allows a floating currency and/or algebraic sign but ED does not allow

Objectives
TABLE PROCESSINGTABLE PROCESSING
A table is a named storage structure consisting of
subunits or entries
e.g. RATE DS 6F
L 4,RATE+8
Accessing table elements with indexed storage
operands:
e.g. LH 9,=F8’
L 5,RATE(9) (9 - index register)

Objectives
Multi-purpose branching instructions
Convenient when counted repetition structure (table processing) is
needed
• Branch on count (BCT and BCTR)
Format: BCT op1,op2 (R-X)
Function: First the op1 value is decremented by 1. Second the
branch is taken to the address specified in op2 only if the value in op1
is not 0.
e.g. LH 9,=H’12’
REPEAT EQU *
..
BCT 9,REPEAT

Objectives
• Branch on index high and branch on index low or equal (BXH
and BXLE)
Format: BXLE op1,op2,op3
BXH
op1 : A register known as the index register
op2 : A even-odd pair of registers
Even register - increment register
Odd register - Limit register
op3 : A storage operand. This is the branch address.

Objectives
Function : First, the value in the increment
register is added to the indexed register. Second,
the branch is taken only when the value in the
index register is ‘lower than or equal to’ / ‘higher
than’ the value in the limit register
Useful when the same register is to be used as the
count and index register

Objectives
BXLE - ‘DO UNTIL’ repetitions
BXH- ‘DO WHILE’ repetitions
e.g... LH 7,=H’0’ index
LH 2,=H’2’ increment amount
LH 3,=H’18 the limit
---
REPEAT ...
LH 6,TABLE(7)
...
BXLE 7,2,REPEAT

ObjectivesLoad instructions with additional features
• Load and Test (LTR)
e.g... LTR 15,15
BNZ ERROR
• Load Address (LA)
LA R1,D2(X2,B2)

Objectives
USING EQUATESUSING EQUATES
• To associate a fixed value with a symbol
• Useful for length and relative address calculation
e.g. TABLE DS 0H
DC C’01
DC C’02’
...
TBLEND EQU *
TBLSIZE EQU TBLEND-TABLE

Objectives
USING EQUATESUSING EQUATES
Can be used for the following purposes:
1. To assign single absolute values to symbols.
2. To assign the values of previously defined
symbols or expressions to new symbols, thus
allowing you to use different mnemonics for
different purposes.
3. To compute expressions whose values are
unknown at coding time or difficult to calculate.
The value of the expressions is then assigned to a
symbol.

SESSION 5
Program Sectioning
Assembler/Session 5

Objectives
Beginning and End of Source ModulesBeginning and End of Source Modules
•Code a CSECT segment before any
statement that affects the location
counter
•END statement is required as the last
statement in the assembly
Assembler/Session 5

Objectives
CONTROL SECTIONSCONTROL SECTIONS
•A source module can be divided into
one or more control sections
•A control section is the smallest
subdivision of a program that can be
relocated as a unit
Assembler/Session 5

• At coding time, establish the addressability
of each control section within the source
module, and provide any symbolic linkages
between control sections that lie in different
source modules.
• Initiated by using the START or CSECT
instruction

Objectives
•Any instruction that affects the location
counter, or uses its current value,
establishes the beginning of the first
control section.
Assembler/Session 5

Format of CSECT:
Name Operation Operand
Any symbol CSECT Not required
or blank
Note: The end of a control section or portion of a
control section is marked by (a) any instruction that
defines a new or continued control section, or (b) the
END instruction.

Objectives
DUMMY SECTIONSDUMMY SECTIONS
•A dummy control section is a reference
control section that allows you to describe
the layout of data in a storage area without
actually reserving any virtual storage.
Assembler/Session 5

• Use the DSECT instruction to initiate a
dummy control section or to indicate its
continuation.
Format of DSECT:
Name Operation Operand
Any symbol DSECT Not required
or blank

Objectives
To use a dummy section :
• Reserve a storage area for the
unformatted data
• Ensure that this data is loaded into the area
at execution time
Analogy: Cobol copybook
Assembler/Session 5

• Ensure that the locations of the symbols in
the dummy section actually correspond to
the locations of the data being described
• Establish the addressability of the dummy
section in combination with the storage area
You can then refer to the unformatted data
symbolically by using the symbols defined in the
dummy section.

Objectives
ASMBLY2 CSECT
BEGIN BALR 2,0
USING *,2
... Reg 3 points to data area
LA 3,INPUT
USING INAREA,3
CLI INCODE,C'A'
BE ATYPE
...
ATYPE MVC WORKA,INPUTA
MVC WORKB,INPUTB
. .
Assembler/Session 5

WORKA DS CL20
WORKB DS CL18
INPUT DS CL39
...
INAREA DSECT
INCODE DS CL1
INPUTA DS CL20
INPUTB DS CL18
...

Objectives
Assembler DirectivesAssembler Directives
TITLE : To provide headings for each page of
the assembly listing of the source modules.
EJECT : To stop the printing of the assembler
listing on the current page, and continue the
printing on the next page.
ORG : To reset the location counter
Assembler/Session 5

LTORG : A literal pool is created
immediately after a LTORG instruction or,
if no LTORG instruction is specified, at the
end of the first control section.
PRINT : To control the amount of detail to
be printed in the listing of programs.
PRINT NOGEN / GEN
Assembler DirectivesAssembler Directives

SESSION 6
Writing a complete program
Assembler/Session 6

Objectives
Program Entry and Exit LogicProgram Entry and Exit Logic
Program entry - Preserve register contents
Program Exit - Restore register contents
Register save area
Always calling program provides a save area
of 18 Full words long used for storage of
registers
Save area address passed through register 13 by
IBM convention
Assembler/Session 6

Objectives
A register save area (18 consecutive full words)
Word Address Contents
1 SAV
2 SAV+4 Address of calling program’s save area
3 SAV+8 Address of called program’s save area
4 SAV+12 Contents of Register 14
...
Assembler/Session 6

Objectives
Responsibilities of called program
Program entry conventions
1.Save contents of registers 0-12,14 & 15 in
calling program’s save area
2.Establish base register
3.Store calling program’s save area in the 2nd
word of its own save area
Assembler/Session 6

Objectives
Program entry conventions (contd..)
4. Store the address of its register save area in the
third word of the calling program’s register save
area
(The addresses in the 3d word of save area establish
a chain of register save areas. This will be useful in
reading the dump when program crashes).
Assembler/Session 6

Objectives
Responsibilities of called program (contd..)
Program Entry
STM R14,R12,12(R13)
BALR R12,0
USING *,R12
ST R13,SAVOWN+4 store calling programs save area
LR R14,R13
LA R13,SAVOWN Reg 13 contains current prog’s SA
...
ST R13,8(R14)
Assembler/Session 6

Objectives
Program Exit conventions
1. Restore registers 0-12 and 14
2. Place the address of the save area provided by the
calling program in Reg 13
3. Place a return code in the low order byte of
register 15 if one is required. Otherwise restore
register 15.
Assembler/Session 6

Objectives
Program Exit
L R13,4(R13)
LM R14,R12,12(R13)
BR R14
Assembler/Session 6

Objectives
Responsibilities of calling program
1. Register 13 must contain the address of a register
save area.
2. Register 15 should be set to the beginning address
of the subroutine
L R15,=V(SUBENTRY)
where SUBENTRY is the entry address (usually the CSECT
name) of the subroutine
Assembler/Session 6

Objectives
Responsibilities of calling program (contd...)
3. Register 14 should have the return address
4. Register 1 should have the address of the parameter
list
A BALR instruction stores the address of the next
instruction in the calling program into register 14 and
transfers control to the called subroutine
BALR R14,R15
Assembler/Session 6

Objectives
Passing parameters to a subroutine
• The standard interface requires that addresses of
parameters be placed in a block of storage, and the
address of the block be loaded into register 1 as the
subroutine is called
• Both input and output parameters are treated the same
way
e.g... ADDS DC A(T)
DC A(U)
DC A(V)
LA R1,ADDS
Assembler/Session 6

Objectives
R1 Main storage
Addr of parmlist Parmlist parm3
Addr of parm1
Addr of parm2 parm1
Addr of parm3 parm2
Assembler/Session 6

Objectives
Called subroutine B may get the second parameter
by
L R3,4(,R1)
L R8,0(,R3)
Assembler/Session 6

Objectives
Registers with special use
R0 : Contains single word output of a
subroutine
R1 : contains the address of an area of
main storage that contains addresses of
parameters
Assembler/Session 6

Objectives
Registers with special use (contd...)
R14 : Contains the return address, the address
in the calling routine to which a subroutine
should return control when finished
R15 : contains the address of the entry point in
the subroutine
R13 : contains the address of an area in which
register contents can be stored by a subroutine
Assembler/Session 6

Objectives
The subroutine RANDOM
RANDOM STM R14,R12,12(R13)
BALR R12,0
USING *,R12
L R7,RN
M R6,=F’65541’
ST R7,RN
LR R0,R7
LM R1,R12,24(R13)
BR R14
RN DC F’8193’
Assembler/Session 6

Objectives
Subroutine RDIGIT
RDIGIT STM R14,R12,12(R13)
BALR R12,0
USING *,R12
ST R13,SAV+4
LA R13,SAV
...
L R15,RANDAD
BALR R14,R15
...
L R13,SAV+4
LM R14,R15,12(R13)
LM R1,R12,24(R13)
BR R14
SAV DS 18F
RANDAD DC A(RANDOM)
Assembler/Session 6

Objectives
Linkage ConventionsLinkage Conventions
•Program divided into 2 or more source
modules
•Source module divided into 2 or more control
sections
•For link-editing, a complete object module or
any individual control section of the object
module can be specified
Assembler/Session 6

Objectives
Communicating between program parts
• To communicate between 2 or more source
modules, symbolically link them together
• To communicate between 2 or more control
sections within a source module, establish proper
addressability
Assembler/Session 6

Objectives
Establishing symbolic linkage
• Identify external symbols in the EXTRN or WXTRN
instruction or the V-type address constant
• provide A-type or V-type address constants to reserve
storage for addresses represented by external symbols
• In the external source modules, identify these symbols
with the ENTRY instruction
(name entry of a START or CSECT instruction is
automatically identified as an entry symbol)
External symbol dictionary
Assembler/Session 6

Objectives
Establishing symbolic linkage (contd...)
e.g. program A
EXTRN TABLEB
WXTRN TABLEB
TABADR DS V(TABLEB)
program B
ENTRY TABLEB
TABLEB DS ...
Assembler/Session 6

Objectives
Address Constants (A and V)
• An address constant is a main storage address contained
in a constant
• A V-type constant is the value of an external symbol - a
relocatable symbol that is external to the current control
section.
Used for branching to locations in other control sections
e.g L 5,ADCON
ADCON DC A(SOMWHERE)
GSUBAD DC V(READATA)
Assembler/Session 6

SESSION 7
Assemble and Link Program
Assembler/Session 7

Objectives
Processing of Instructions
Time/ M/C Assembler ENTRY Macro
Activity instructions. EXTRN Instr.
Code source m/c DC,DS
instruc.
Preassembly Refer to macro
instruc.
Assembly object code
LKED
Prog fetch
Execution data area form data
area in load mod
Processing of Instructions
Time/ M/C Assembler ENTRY Macro
Activity instructions. EXTRN Instr.
Code source m/c DC,DS
instruc.
Preassembly Refer to macro
instruc.
Assembly object code
LKED
Prog fetch
Execution data area form data
area in load mod
Assembler/Session 7

Objectives
JCL ‘ parm’ processing
EXEC PGM=pgmname,PARM=
When program gets control :
•Register 1 contains the address of a full word
on a full word boundary in program’s address
space
•the high order bit of this full word is set to 1
(this convention is to indicate the last word in
a variable length parameter list)
Assembler/Session 7

JCL ‘ parm’ processing ...
• Bits 1-31 of the full word contain the
address of a 2-byte length field on a half
word boundary
• The length field contains a binary count
of the no. of bytes in the PARM field
which immediately follows the length
field

Objectives
COBOL to Assembler
CALL asmpgm USING COMM-AREA
PL/I to Assembler
DCL ASMSUB ENTRY OPTIONS(ASSEMBLER)
CHARSTRING CHAR(25);
CALL ASMSUB(CHARSTRING);
Ref : PL/I Programming Guide, COBOL programming
Guide
Assembler/Session 7

SESSION 8
Macro Language
Assembler/Session 8

ObjectivesMacros
• Short source routines written and
stored in libraries
•Assembler inserts the source
statements in the program where
the macro appears
Assembler/Session 8

Macro Definition
Format :
•A header statement
•A prototype
•Model statements
•A trailer statement

Objectives
Header statement:
MACRO
Prototype:
&name MOVE &TO,&FROM,&LENGTH
Model statements:
A set of machine and assembler instructions
Trailer statement:
&name MEND
Assembler/Session 8

Objectives
Macro Instruction:
• A statement containing the name of a
macro
• when expanded, the symbolic parameters in
the model statements are replaced by
corresponding parameters from the macro
instructions
• symbolic parameters may be positional or
keyword
Assembler/Session 8

Macro Instruction ...
MACRO
&LABEL HALFSWAP &REG,&SV
&LABEL ST &REG,&SV
SLL &REG,8
IC &REG,&SV
SLL &REG,8
IC &REG,&SV+1
MEND

Objectives
SET Symbols (global or local)
3 types :
• arithmetic (SETA)
• binary (SETB)
• character (SETC)
• SET symbols are declared using,
LCLA LCLB LCLC
GCLA GCLB GCLC
Assembler/Session 8

Objectives
Format:
Label operation operands
symbol-name SETA An expression
SETB
SETC
e.g.
LCLA &A1
GCLA &A2
&A1 SETA 1
&A2 SETA &A1+3
Assembler/Session 8

Objectives
Attributes
There are 6 attributes of a symbol or
symbolic parameter :
type, length, scaling, integer, count and
number
System variable symbols
&SYSINDX, &SYSDATE, &SYSTIME, &SYSECT,
&SYSPARM, &SYSLOC
Assembler/Session 8

Objectives
Conditional Assembly
The assembler can be made to branch and loop
among assembler language statements using
sequence symbols and the assembler
instructions AIF and AGO
Sequence symbol : Period followed by 1 to 7
alphabets or digits of which the first is a letter
e.g. .Z23Ab
Assembler/Session 8

Objectives
Format:
Label Operation Operand
seq symbol AGO seq. symbol
or blank
-do- AIF A logical expression
enclosed in parenthesis,
followed by seq symbol
Assembler/Session 8

A logical expression is composed of one or
more relations or values of SETB symbols
connected by logical connects AND, OR, AND
NOT, OR NOT
A relation consists of 2 arithmetic expressions
or 2 character expressions connected by a
relational operator EQ, NE, LT, LE, GT, GE

Objectives
e.g.
MACRO
PSRCH &PARAMS,&STRING
GBLB &FOUND
LCLA &I
&FOUND SETB 0
.LP AIF ((&I GE N’&PARAMS) OR &FOUND) .E
&I SETA &I+1
&FOUND SETB (‘&PARAMS(&I)’ EQ ‘&STRING’)
AGO .LP
.E MEND
Assembler/Session 8

Objectives
Accessing QSAM files:
Keywords in DCB parameter:
DSORG PS Physical sequential
RECFM F,FA,FB,FBA,V,VBA
BLKSIZE Block length
LRECL Record Length
DDNAME Dataset name in JCL
MACRF Macro GM - Get Move GL - Get Locate
PM - Put Move PL - Put locate
Move parameter directly puts the record in the storage area
specified while Locate mode Loads the address of the record in
Register 1
Assembler/Session 8

Objectives
Accessing VSAM files: ACB macro
AM - VSAM (For documentation)
BUFND - No. of I/O buffers for data control intervals
BUFNI - No. of I/O buffers for index control intervals
BUFSP - Size of an area for data and Index I/O buffers
DDNAME - Filename used in the DD statement. If omitted
refers to the ACB macro name
EXLST - Address to the EXLST macro. Generates a list of
addresses for user routines
MACRF - Types of processing the file will do
Assembler/Session 8

Objectives
Accessing VSAM files: ACB macro (Continued)
EXLST options:
AM - VSAM
EODAD = (Address, A/N, L) (Load module)
EXCPAD = (Address, A/N, L) (Load module)
JRNAD = (Address, A/N, L) (Load module)
LERAD = (Address, A/N, L) (Load module)
SYNAD = (Address, A/N, L) (Load module)
Active/No, Stored in load module
Assembler/Session 8

Objectives
Accessing VSAM files: RPL macro (Request parameter list)
ACB - Address of the ACB macro
AREA - Address of the work area to be used
AREALEN - Length of the work area (Should be large enough
to hold largest record in Move mode and at least 4 bytes in the Locate mode)
RECLEN -Length of the records in the file (For VB you have
to put the length before writing using MODCB)
ARG - Label containing the key for the search (Key for
KSDS, RRN for RRDS and RBA for ESDS)
OPTCD - 5 sets of groups of parameters
Assembler/Session 8

Objectives
Accessing VSAM files: RPL macro (Continued)
Options for OPTCD:
KEY/CNV/ADR - Access by key,Control interval or
Relative byte address
SEQ/DIR/SKP - Sequential processing,Direct, Skip
sequential
FWD/BWD - Forward sequential processing,Backward
ARD/LRD -Start seq.processing with ARG specified/
Backward processing from the last record
NUP/NSP/UPD - No updating(Next rec not ready),No
updating Next rec ready(DA only), Record updating)
MVE/LOC - Move mode/ Locate mode
Assembler/Session 8

Objectives
Accessing VSAM files:
OPEN - Open the file
CLOSE - Close the file
GET - Read a record
PUT - Store a record
ERASE - Delete a record
POINT - Position for access
Advanced macros: SHOWCB, TESTCB, MODCB
Assembler/Session 8

SESSION 9
Other Topics
Assembler/Session 9

ObjectivesCharacteristics of good assembler program
• has simple, easy to understand logic
• uses mostly simple instructions
• has no relative addressing
• uses subroutines
Assembler/Session 8

Characteristics of good assembler program ...
• uses DSECTs
• has efficient code (LA R10, 4(0,R10 - A R10,=F’4)
• does not abnormally terminate due to user error
• requests and check feedback from macro instructions
• provides meaningful error messages

Objectives
Characteristics of good assembler program
(contd..)
• lets the assembler determine lengths
• has opcodes, operand and comments aligned
• contains meaningful comments
• uses meaningful labels
Assembler/Session 8

Objectives
Structured Programming
• To improve design and understandability of a
program
• made up of building blocks of subroutines
Conventions for general purpose registers
• Base registers
• Link registers
Assembler/Session 8

Objectives
The EXecute Instruction
• the EX instruction is a R-X type instruction that
directs the execution of an instruction called the
subject instruction, which is addressed by the second
operand
• the subject instruction is in effect a one-instruction
subroutine
Assembler/Session 9

•The subject instruction is modified before execution
(though not altered at its main storage location) :
bits 8-15 of the instruction ORed with bits 24-31 of
register R1 to form the second byte of the instruction
actually executed
e.g. Let reg 9 have the length of string to be moved
EX R9,VARMVC
VARMVC MVC A(0),B
The EXecute Instruction (contd...)

Objectives
DEBUGGINGDEBUGGING
Exceptions and Interrupts
Interrupts that result directly from attempts at invalid
program execution are called program-check
interrupts; identified by a code
Interruption code 1 : Operation
Interruption code 2 : Privileged operation
Interruption code 4 : Protection
Interruption code 5 :Addressing
Interruption code 6 :Specification
Assembler/Session 9

Objectives
DEBUGGING
Exceptions and Interrupts (contd..)
Interruption code 7 : Data
Interruption code 8 : Fixed-Point Overflow
Interruption code 9 : Fixed-Point Divide
Other Interruption codes ( 3, 10, 11, 12, 13,
14, 15)
Assembler/Session 9

Objectives
DEBUGGINGDEBUGGING
Reading dumps
• whenever a program abends an indicative
dump is generated
• The completion code is a code furnished by
the O/S to designate the reason for the
termination of the job step
• In case of program check interruption, the
first 2 digits of the completion code is 0C
Assembler/Session 9

• Locate the entry point of your program
Reading dumps ...
DEBUGGINGDEBUGGING

Objectives
DEBUGGINGDEBUGGING
Reading dumps (contd...)
• The register contents are the contents at the
point of interruption (the instruction that
caused the interrupt is usually the one just
before the interrupt address given)
• use address at interrupt and entry address to
locate the instruction that caused the program-
check interruption
Assembler/Session 9

Objectives
DEBUGGINGDEBUGGING
Full and Partial dumps
• //SYSUDUMP DD SYSOUT=A
• SNAP macro
Assembler/Session 9

Reading the dump
• SAVE AREA trace
• P/P Storage
• Examine register contents, PSW and listed entry
point to find the portion of program being executed
• Look at main storage dump to determine the data
being used
DEBUGGINGDEBUGGING

Objectives
SYSTEM MACROSSYSTEM MACROS
Data Management Macros
DCB - Construct a data control block
OPEN - Logically connect a dataset
CLOSE - Logically disconnect a dataset
GET - Obtain next logical record (queued access)
PUT - Write next logical record (queued
access)
READ - Read a block (basic access)
WRITE - Write a block (basic access)
Assembler/Session 9

Objectives
Supervisor Services Macros
ABEND - Abnormally terminate a task
CALL - Pass control to a control section
GETMAIN - Allocate virtual storage
FREEMAIN - Free virtual storage
LOAD - Bring a load module into virtual storage
RETURN - return control to the calling program
SAVE - Save register contents
Assembler/Session 9

Objectives
Supervisor Services Macros (contd)
SNAP - Dump virtual storage and continue
LINK - Pass control to a Program in
Another load module
WTO - Write to operator
Assembler/Session 9

Objectives
e.g. File I/O
OPEN (INFILE,INPUT)
GET INFILE,RECAREA
PUT OUTFILE,RECAREA
CLOSE (INFILE)
INFILE DCB
DSORG=PS,MACRF=GM,DDNAME=IFILE
OUTFILE DCB
DSORG=PS,MACRF=PM,DDNAME=OFILE
(RECFM=,LRECL=,BLKSIZE=,)
Assembler/Session 9

Objectives
Three forms :
Standard form : Results in instructions that store
into an inline parameter list and pass control to
the required program
List form : Provides as out-of-line parameter list
Execute form : Provides the executable instructions
required to modify the out-of-line parameter list
and pass control to the required program
Assembler/Session 9

Assembler Language Tutorial for Mainframe Programmers

More Related Content

What's hot (20)

Viewers also liked (17)

Similar to Assembler Language Tutorial for Mainframe Programmers (20)

More from Srinimf-Slides (20)

Recently uploaded (20)

Assembler Language Tutorial for Mainframe Programmers

Editor's Notes