Subprograms
Introduction
• Two fundamental abstraction facilities
– Process abstraction
• Emphasized from early days
– Data abstraction
• Emphasized in the1980s
Fundamentals of Subprograms
• Each subprogram has a single entry point
• The calling program is suspended during execution of the called subprogram (i.e. there is only one
subprogram in execution of any given time).
• Control always returns to the caller when the called subprogram’s execution terminates
Basic Definitions
• A subprogram definition describes the interface to and the actions of the subprogram abstraction
• A subprogram call is an explicit request that the subprogram be executed. Active subprogram,
program that starts execution but not yet completed.
• A subprogram header is the first part of the definition, including the name, the kind of subprogram
(procedure, function), and the formal parameters (optional).
• Examples:
In Fortran: Subroutine Adder (parameters)
Ada: Procedure Adder (parameters)
C-based: void adder (parameters)
• The parameter profile (aka signature) of a subprogram is the number, order, and types of its
parameters
• The protocol is a subprogram’s parameter profile and, if it is a function, its return type
• Function declarations in C and C++ are often called prototypes. While Java and C# do not need
declarations of their methods, because there is no requirement in those languages that methods be
defined before they are called.
• A subprogram declaration provides the protocol, but not the body, of the subprogram
• A formal parameter is a dummy variable listed in the subprogram header and used in the
subprogram
• An actual parameter represents a value or address used in the subprogram call statement
Actual/Formal Parameter Correspondence
• Positional
– The binding of actual parameters to formal parameters is by position: the first actual
parameter is bound to the first formal parameter and so forth
– Safe and effective method, as long as the parameter lists are relatively short.
• Keyword: in case of lists are long
– The name of the formal parameter to which an actual parameter is to be bound is specified
with the actual parameter
– Parameters can appear in any order
• Ada procedures can be called using keyword parameters, as in :
Sumer (LengthMy_Length, List My_Array, SumMy_Sum)
– Where the definition of Sumer has the formal parameters Length, List, and Sum.
• Disadvantages of using keyword, is the user of the subprogram must know the named of formal
parameters.
• Ada, Fortran95 allow a mixed of both (positional and keyword)
Ex: Sumer (My_Length, SumMy_Sum, ListMy_Array);
Restriction with this approach, is after a keyword parameter appears in the list, all remaining parameters
must be keyworded. The reason is that, because a position may no longer be well defined after a
keyword parameter has appeared.
Formal Parameter Default Values
Ch9-1
• In certain languages (e.g., C++, Ada, Fortran95, and PHP), formal parameters can have default
values (if not actual parameter is passed)
• Consider the following Ada function header
function Compute (Income: Float; Exemptions: Integer:=1; Tax_Rate:Float)return Float;
• The Exemptions formal parameter can be absent in a call to Compute;
Pay:=Compute (2000.0, Tax_Rate 0.15);
– In C++, which has no keyword, default parameters must appear last because parameters are
positionally associated
– A C++ function header:
float Compute (float income, float tax_rate, int exemptions=1)
call
pay=Compute(2000.0, 0.15
• C# methods can accept a variable number of parameters as long as they are of the same type
Procedures and Functions
• There are two categories of subprograms
– Procedures are collection of statements that define parameterized computations.
– Procedures can produce results in the calling program unit by two methods.
1. If there are variables that are not formal parameters but are still visible in both the
procedure and the calling program unit, the procedure can change them.
2. If the subprogram has formal parameters that allow the transfer of data to the caller, those
parameters can be changed.
– Functions structurally resemble procedures but are semantically modeled on mathematical
functions
• They are expected to produce no side effects (i.e. modify neither its parameter nor
any variables defined out-side the function)
• In practice, program functions have side effects
– Functions are called by appearances of their names in expressions, along with the required
actual parameters.
– The value produced by a function’s execution is returned to the calling code, replacing the
call itself.
• Functions define new user-defined operators. Ex: if language does not have an exponentiation
operator, a function can be written for it. In C++ could be
float power (float base, float exp)
which could be called with
result=3.4* power(10.0, x);
• Fortran and Ada, provide both function and procedures. C-based have only functions (and/ or
methods).
Design Issues for Subprograms
• What parameter passing methods are provided?
• Are parameter types checked? (Actual against formal)
• Are local variables static or dynamic?
• Can subprogram definitions appear in other subprogram definitions (nested)?
• If subprograms can be passed as parameters and subprograms can be nested, what is the referencing
environment of a passed subprogram?
• Can subprograms be overloaded? i.e. has the same name as another subprogram in the same
referencing environment.
• Can subprogram be generic? Is one whose computation can be done on data of different types in
different calls.
Local Referencing Environments
• Subprograms can define their own variables, thereby defining local referencing environments.
Ch9-2
• Local variables can be either static or stack-dynamic.
• Local variables can be stack-dynamic (bound to storage where the subprogram designs execution
and are unbound from storage when that execution terminates)
– Advantages
• Support for recursion
• Storage for locals in an active subprogram is shared among some inactive subprograms (good
when memory is small)
– Disadvantages
• Allocation/de-allocation, initialization time cost
• Indirect addressing (because the place in the stack where a particular local variable
will reside can be determined only during execution)
• Subprograms cannot be history sensitive
• Or Local variables can be static
* Advantages: - More efficient (no indirection)
- No run-time overhead, allow subprograms to be history-sensitive
* Disadvantages: - Cannot support recursion
- Storage cannot be stored with the local variables of other inactive subprograms
• In most contemporary languages, local variables in a subprogram are by default stack-dynamic.
• In C and C++ functions, locals are stack-dynamic unless specifically declared to be static.
• Example:
int adder (int list[], int listlen)
{
static int sum  0; //static variable
int count;
for (count  0; count  listlen; count   )
sum   list[count];
return sum;
}
• Ada subprograms and the methods of C++, Java, and C# have only stack-dynamic local variables.
Parameter Passing Methods
• Ways in which parameters are transmitted to and/or from called subprograms
• Semantics models of parameter passing
– Formal parameters are characterized by one of three distinct semantics models
1. They can receive data from the corresponding actual parameters (in mode).
2. They can transmit data to the actual parameters (out mode).
3. They can do both (inout mode).
• Implementation models of parameter passing
– A variety of models has been developed by language designers to guide the implementation
of the three basic parameter transmission modes.
– Pass-by-value
– Pass-by-result
– Pass-by-value-result
– Pass-by-reference
– Pass-by-name
Models of Parameter Passing
Ch9-3
Pass-by-Value (In Mode)
• The value of the actual parameter is used to initialize the corresponding formal parameter, which
then acts as a local variable in the subprogram.
– Normally implemented by copying
– Can be implemented by transmitting an access path but not recommended (require that the
value be in a write-protected call (one that can only be read) enforcing write protection is not
easy)
* Disadvantages:
– When copies are used, additional storage is required
– Storage and copy operations can be costly (such as array with many elements)
Pass-by-Result (Out Mode)
• When a parameter is passed by result, no value is transmitted to the subprogram; the corresponding
formal parameter acts as a local variable; its value is transmitted to caller’s actual parameter when
control is returned to the caller
– Require extra storage location and copy operation
• Potential problem: there can be actual parameter collision, such as sub(p1, p1);
whichever formal parameter is copied back (last) will represent the current value of p1
• Thus the order in which the actual parameters are copied determines their value.
• Another problem: The implementer may be able to choose between two different times to evaluate
the address of the actual parameters: at the time of the call or at the time of the return.
• Ex: suppose a subprogram is called with the parameter list[index]. If index is changed by the
subprogram, either through global access or as a formal parameter, then the address of list[index]
will change between the call and the return.
• The implementer must choose the time at which the address to which to return the value will be
determined, at the time of the call or at the time of the return. Makes programs unportable between
implementations.
Pass-by-Value-Result (inout Mode)
• A combination of pass-by-value and pass-by-result
• Sometimes called pass-by-copy (actual values are copied to formal parameters at subprogram entry
and copied back at termination)
• Formal parameters have local storage and the value of the actual parameter is used to initialize the
corresponding formal parameters.
• At subprogram termination, the value of the formal parameter is transmitted back to the actual
parameter.
• Disadvantages:
– Those of pass-by-result (problems associated with the order in which actual parameters are
assigned)
– Those of pass-by-value (requiring multiple storage for parameters and time for copying
values)
Pass-by-Reference (Inout Mode)
• Pass an access path to the called subprogram
Ch9-4
• Also called pass-by-sharing
• Advantage: Passing process is efficient in terms of time and space. (no copying and no duplicated
storage)
• Disadvantages
– Slower accesses (compared to pass-by-value) to formal parameters (indirect addressing).
– Potentials for un-wanted side effects in case of one-way communication to the called subprogram.
– Un-wanted aliases (access broadened to nonlocal variables)
• There are several ways alias can be created when parameters are passed by reference.
• Collisions can occur between actual parameters.
• Ex: void fun(int &first, int &second)
• If the call to fun happens to pass the same variable twice, as in fun(total, total) the first and second in
fun will be aliases.
• Fun(list[i], list[j])
• If these two parameters are passed by reference and i happens to be equal to j, then first and second
are aliases.
• Problem with aliasing is harmful to readability and reliability.
Pass-by-Name (Inout Mode)
• Actual parameter is textually substituted for the corresponding formal parameter in all its
occurrences in the subprogram.
• Formal parameters are bound to actual values or addresses at the time of the subprogram call.
• Formals are bound to an access method at the time of the subprogram call, but actual binding to a
value or address takes place at the time of a reference or assignment
• Allows flexibility in late binding
Parameter Passing Methods of Major Languages
• Fortran
– Always used the inout semantics model (but does not specify whether pass-by-reference or
pass-by-value-result should be used).
– Before Fortran 77: pass-by-reference
– Fortran 77 and later: scalar (simple) variables are often passed by value-result
• C
– Pass-by-value
– Pass-by-reference (input mode) is achieved by using pointers as parameters
• C++
– A special pointer type called reference type for pass-by-reference
– Void fun(const int &p1, int p2, int &p3) {…}
– Where p1 is pass-by-reference but cannot be changed in the function fun, parameter p2 is
pass-by-value, and p3 is passed-by-reference.
• Java
– All parameters are passed by value
– Object parameters are passed by reference (because objects can be passed only through
reference variables)
• Ada
– Three semantics modes of parameter transmission: in, out, in out; in is the default mode
– Formal parameters declared out can be assigned but not referenced; those declared in can be
referenced but not assigned; in out parameters can be referenced and assigned.
– Consider the following Ada subprogram header:
Procedure Adder (A:in out Integer; B: in Integer; C: out Float)
• Fortran95, similar to Ada, but we use Intent attribute.
Ch9-5
subroutine Adder(A, B, C)
Integer, Intent(Inout) :: A
Integer, Intent(In) :: B
Integer, Intent(Out) :: C
• C#
– Default method: pass-by-value
– Pass-by-reference is specified by preceding both a formal parameter and its actual parameter
with ref
– void sumer(ref int oldsum, int newone) {…}
– Call  sumer(ref sum, newvalue); //1st by reference, 2nd by value
Type Checking Parameters
• Considered very important for reliability
• FORTRAN 77 and original C: none (neither the number of parameters nor their types were checked)
• Pascal, FORTRAN 90, Java, and Ada: it is always required
• ANSI C and C++: choice is made by the user: either in original C, the names of the parameters are
listed in parentheses and the type declarations for them follow, as in:
double sin(x)
Using this method avoids type checking, allowing calls such as:
double value;
double x;
int count;
{....}
...
value  sin(count); Be legal, although they are never correct.
Alternative prototypes
- in which the formal parameter types are included in list, as in
double sin(doubl e x)
{...}
- if this version of sin is called with the same call, i.e.
value=sin(count);
- it is also legal. The type of the actual parameter (int) is checked against that of the formal parameter (double)
- Although they do not mach, int is coerced to double, and it will work.
- If conversion is not possible (ex actual parameter is an array), or the number of parameters is wrong. Then
a syntax error is detected.
- All functions in C++, C99, must have their formal parameters in prototype form.
- However, type checking can be avoided for some of the parameters by replacing the last part of the
parameter list with an ellipsis, as in:
int printf(const char *format_string, …);
• Relatively new languages Perl, JavaScript, and PHP do not require type checking
Multidimensional Arrays as Parameters
• If a multidimensional array is passed to a subprogram and the subprogram is separately compiled, the
compiler needs to know the declared size of that array to build the storage mapping function (index function)
Multidimensional Arrays as Parameters: C and C++
• Programmer is required to include the declared sizes of all but the first subscript in the actual parameter
because it uses row-major order, so for z-dimensional array x[l1 : u2 , l2 : u2 ] , the index function
loc( x(i, j ))  loc( x(l1l2 )  (i  l1 )u2  ( j  l2 )
so we need the number of columns ( u2 ), but not ( u1 )
• Ex in C:
Ch9-6
void fun(int matrix[][10]){....}
void main()
{ int mat[5][10];
......
fun(mat);
.......
}
• The problem: Disallows writing flexible subprograms i.e. does not allow writing a function that accepts
matrixes with different numbers of columns. We need a new one for every matrix with different number of
columns.
• Solution: pass a pointer to the array and the sizes of the dimensions as other parameters; the user must
include the storage mapping function in terms of the size parameters. Consider the following prototype
void fun(float *mat_ptr, int num_rows, int num_cols);
Multidimensional Arrays as Parameters: Pascal and Ada
• Pascal
– Not a problem; declared size is part of the array’s type
Multidimensional Arrays as Parameters: Fortran
• Formal parameter that are arrays have a declaration after the header
– For single-dimension arrays, the subscript is irrelevant
– For multi-dimensional arrays, the subscripts allow the compiler to build the storage-mapping function
Multidimensional Arrays as Parameters: Java and C#
• Arrays are objects; they are all single-dimensioned, but the elements can be arrays
• Each array inherits a named constant (length in Java, Length in C#) that is set to the length of the
array when the array object is created
Design Considerations for Parameter Passing
• Two important considerations
– Efficiency
– One-way or two-way data transfer
• But the above considerations are in conflict
– Good programming suggests limited access to variables, which means one-way whenever
possible. Example, when a large array needs to be passed to a subprogram that does not
modify it.
– However, pass-by-value would require that the entire array be moved to a local storage area
of the subprogram, costly in time and space.
– But pass-by-reference is more efficient to pass structures of significant size (large arrays)
Parameters that are Subprogram Names
• It is sometimes convenient to pass subprogram names as parameters to other subprograms.
• Issues:
1. Are parameter types checked?
2. What is the correct referencing environment for a subprogram that was sent as a parameter?
Parameters that are Subprogram Names: Parameter Type Checking
• C and C++: functions cannot be passed as parameters but pointers to functions can be passed;
parameters can be type checked
• FORTRAN 95 type checks mechanism
• Later versions of Pascal and
• Ada does not allow subprograms to be passed as parameters.
Examples of parameter passing
• Consider the C function
Ch9-7
void swap1(int a, int b)
{
int temp  a;
a  b;
b  temp;
Suppose this function is called with
swap1(c, d);
}
• C used pass-by-value
• The action of swap1 can be described by the following pseudocode
a=b
b=d
temp=a
a=b
b=temp
• Although a ends up with d’s value, and b ends up with c’s value, the value of c and d are unchanged
because nothing is transmitted back to the caller.
• We can modify the c swap function to achieve the effect of pass-by-reference
void swap2(int * a, int * b)
{ int temp  *a;
swap2 can be called with
swap2(&c, &d);
* a  *b;
* b  temp;
}
• The actions of swap2 can be described with
a=&c -- move 1st parameter address in
b=&d -- move 2nd parameter address in
temp=*a
*a=*b
*b=temp
The swap operation is successful.
• swap2 can be written in C++ using reference parameters as follows
void swap2(int & a, int & b)
{
int temp  a;
a  b;
b  temp;
}
• This simple swap operation is not possible in Java, because it has neither pointers nor C++’s kind of
references. In Java, a reference variable can only point to an object, not a scalar value.
• Semantics of pass-by-value-result for input-mode scalar parameters.
• consider function swap3, which we assume uses pass-by-value-result
procedure swap3(a : in out Integer, b : in out Integer) is
temp : Integer;
begin
temp : a;
Suppose swap3 is called with
swap3(c, d);
a : b;
b : temp;
end swap3;
Ch9-8
• The actions of swap3 with this call are
addr_c=&c
addr_d=&d
a=*addr_c
b=*addr_d
temp=a
a=b
b=temp
*addr_c=a
*addr_d=b
-- move 1st parameter address in
-- move 2nd parameter address in
-- move 1st parameter value in
-- move 2nd parameter value in
-- move 1st parameter value out
-- move 2nd parameter value out
• Explore what happens when aliasing is involved with pass-by-value-result and pass-by-reference.
Consider the following C-like
int i  3 / * global variable * /
void fun(int a, int b)
{ i  b; }
void main()
{ int list[10];
list[i]  5;
fun(i, list[i]);
}
• In fun, if pass-by-reference is used, I and a are aliases.
• If call-by-value-result is used, I and a are not aliases.
• The action of fun, assuming pass-by-value-result are
addr_i=&i
-- move parameter address in
addr_list i=&list[i]
a=*addr_i
b=*addr_list i
i=b
-- sets i to 5
*adde_i=a
-- move 1st parameter value out
*addr_list i=b
• In this case, the assignment to the global I in fun changes its value from 3 to 5, but the copy back of
the first formal parameter sets it back to 3.
• Observe that, if pass-by-reference is used, the result is that the copy back is not part of the semantics,
and i remains 5.
• Also note that because the address of the 2nd parameter is computed at the beginning of fun, any
change to the global i has no effect on the address used at the end to return the value of list[i].
Parameters that are Subprogram Names: Referencing Environment
• In languages that allow nested subprograms, such as Javascript, there is question of what referencing
environment for executing the passed subprogram should be used. There are 3 choices:
• Shallow binding: The environment of the call statement that enacts the passed subprogram
• Deep binding: The environment of the definition of the passed subprogram
• Ad hoc binding: The environment of the call statement that passed the subprogram as an actual
parameters.
• The following example written in Javascript, illustrates these choices
Ch9-9
function sub1()
{ var x;
function sub2()
{ alert(x); //creates a dialog box with the value of x
}
function sub3()
{ var x;
x3
sub4(sub2);
};
function sub4(subx )
{
var x;
x  4;
subx();
};
x  1;
sub3();
Consider the execution of sub2 when it is called in sub4
- Shallow binding: the reference environment of that execution is that
of sub4, so that reference to x in sub2 is bound to local x in sub4,
and the output is 4.
- Deep binding: reference environment of sub2’s execution is that of
sub1, so the reference to x in sub2 is bound to local x in sub1,
output=1.
- Adhoc binding: the binding is the local x in sub3, output=3.
pr int
sub1 call
 sub3() 
 sub 4( sub 2) 
 sub 2 
 x
shallow  statement of pass parameter  x  4
deep  definition of pass parameter  x  1
Adhoc  passed actual parameter  x  3
};
Overloaded Subprograms
• An overloaded subprogram is one that has the same name as another subprogram in the same
referencing environment
– Every version of an overloaded subprogram has a unique protocol. i.e. it must be different from
the others in the number, or types of its parameters, or in its return type if it is a function.
• C++, Java, C#, and Ada include predefined overloaded subprograms
• In Ada, the return type of an overloaded function can be used to disambiguate calls (thus two
overloaded functions can have the same parameters profile but differ only in their return types). This
works because Ada does not allow mixed-mode expressions, so the context of the function call can
specify the type that is returned from the function.
• Because C++, Java and C# allow mixed-mode expressions, the return type is irrelevant to
disambiguation of overloaded functions (methods). Example, if a program has two functions named
fun and both take an int parameter but one returns an int and one returns a float, the program would
not compile, because the compiler could not determine which revision of fun should be used.
• Ada, Java, C++, and C# allow users to write multiple versions of subprograms with the same name
Generic Subprograms
• A generic or polymorphic subprogram takes parameters of different types on different activations
• Overloaded subprograms provide a particular kind of polymorphism called ad hoc polymorphism
• A subprogram that takes a generic parameter that is used in a type expression that describes the type
of the parameters of the subprogram provides parametric polymorphism
Examples of parametric polymorphism: C++ (generic functions)
template <class Type>
Type max(Type first, Type second) {
return first > second ? first : second;
}
Where type is the parameter that specifies the type of data on which the function will operate.
• The above template can be instantiated for any type for which operator > is defined
int max (int first, int second) {
return first > second? first : second;
}
Ch9-10
• Although this process could be defined as a macro, a macro would have the disadvantage of not
operating correctly if the parameters were expressions with side effects. Example, suppose the macro
were defined as
#define max(a, b) ((a)>(b)) ? (a):(b)
• This definition is generic, i.e., it works for any numeric type. However, it does not always work
correctly if called with a parameter that has a side effect, such as
Max(x++, y)
Which produces:
((x++)>(y)? (x++):(y))
• Whenever the value of x is greater than that of y, x will be incremented twice.
Design Issues for Functions
• Are side effects allowed?
– Parameters should always be in-mode to reduce side effect (like Ada)
• What types of return values are allowed?
– Most imperative languages restrict the return types of their functions
– C allows any type except arrays and functions
– C++ is like C but also allows user-defined types or classes to be returned from its functions.
– Ada allows any type
– Java and C# do not have functions but methods can have any type
User-Defined Overloaded Operators
• Operators can be overloaded in Ada and C++
• An Ada example that overloads the multiplication operator (*) to compute the dor product of two
arrays.
• The dot product of two arrays of equal length is the sum of the products of each of the corresponding
pairs of elements of the two arrays.
• Suppose Vector-type has been defined as an array type with integer elements.
Function “*”(A,B: in Vec_Type): return Integer is
Sum: Integer := 0;
begin
for Index in A’range loop
Sum := Sum + A(Index) * B(Index)
end loop
return sum;
end “*”;
…
c = a * b; -- a, b, and c are of type Vec_Type
Ch9-11