c programming reference

8/10/2019 C Programming Reference

http://slidepdf.com/reader/full/c-programming-reference 1/18

C Programming Reference >> C Basic Programming Tutorials >>

Pointersquisite Knowledege

21/07/07 - 12746 Views - Ratings : 4.29 of 5 / 87 Votes

Level - Beginner

Pointer is the feature which has grabbed most attention in C Programming combined all other

features . Pointers are most powerfull feature but can be also most dangerous if abused. The

pointers in C are used in my task which cannot be achieved by any other mean, such as dynamicmemmory allocation, changing value using function, dynamic data structures etc. . They are also

much much efficient in use of system resources than thier counter part and can be used to

enhance your program speed and reduce its size. The success of a c programmer depends

crucially on his understanding of C Pointers and that us reason why we are going to discuss themhere in painstaking detail.

Contents -

1. Pointer Fundamenta2. Pointer Variable3. Pointer Assignment 4. Pointer Conversion5. Pointer Arithemetic 6. Pointer Comparison 7. Arrays And Pointers 8. Multiple Indirection 9. Pointer Intialization 10. Pointer To A Function11. Dynamic Memmory Allocation Using Pointers 12. Restrict Pointers13. Pointer Drawbacks

1. Pointers Fundamentals

In the simplest term pointer is a nearly integer variable which stores a memmory address of acomputer which may contain other variable or even another pointer.

If a variable contains address of another variable than it is said that first variable points tosecond. Pointer can also be represented as a refrence to another variable but there is very subtle

diffrence in the two statements which is mostly dependent upon situation and enviorment.

Pointer is generally of size of integer on the machine but it may be of diffrent type which

indicates the type of variable the pointer is pointing to decides pointers properties and behaviour.

Pointer of one type cannot be implicitly converted from one type to another but can be explicitlyconverted using type casting. In such a conversion a pointer always assumes that it is point to a

object of its type but reality may differ and if used incorrectly may lead to disasters including

permanent machine damage.

http://www.cprogrammingreference.com/index.html

http://www.cprogrammingreference.com/basictutorials_home.html


http://www.cprogrammingreference.com/Tutorials/Basic_Tutorials/Pointers.php#1



























http://www.cprogrammingreference.com/index.html



Also it is important to note that all operation perform on pointers are done through two operators

'*' (Star) and '&' (Ampercent). '&' is a unary operator that returns a memmory address of a

variable. '*' is complement of '&' and return value stored at a memmory location stored in a pointer. '*' can interpreted as statement "at address" while '&' can be interpreted as statement

"address of".

Back To Top

2. Pointer Variable

Declaring a pointer variable is quite similar to declaring an normal variable all you have to do is

to insert a star '*' operator before it.

General form of pointer declaration is -

type* name;

where type represent the type to which pointer thinks it is pointing to.

Pointers to machine defined as well as user-defined types can be made.

Multiple pointers of similar type can be declared in one statement but make sure you use *

before every one otherwise they will become a variable of that type.

Back To Top

3. Pointer Assignment

This is a type of expression used to assign value of one pointer to another using assignment

operator '=' . In this value of right hand side points to memmory address of variable stored in lefthand side pointer. As a result both pointers point to same memmory location after this

expression.

Pointer of similar type can be used in expression easily as shown below but for diffrent type

pointers you need to type cast them as shown in next section.

#include <stdio.h>

int main (){

char ch = a;char* p1, *p2;

p1 = &ch;

p2 = p1; // Pointer Assignement Taking Place

printf (" *p1 = %c And *p2 = %c", *p1,*p2); // Prints 'a' twice

http://www.cprogrammingreference.com/Tutorials/Basic_Tutorials/Pointers.php






return 0;

}

Back To Top

4. Pointer Conversion

Pointer Conversion is a very powerfull yet very dangerous feature. Before concept of pointer

conversion you must understand the concept of a void pointer. Void pointer technically is a pointer which is pointing to the unknown. Void pointer has special property that it can be type

casted into anyother pointer without any type casting though every other conversion needs an

type casting. Also note that in C++ even void pointer needs an type casting so to maintain the

compatibility you may want to type cast void pointer anyway. Also in dynamic memmoryallocation function such as malloc ( ) and alloc ( ) returns void pointer which can be easily

converted to other types.

Also there is a pointer called null pointer which seems like void pointer but is entirely diffrent.

Null pointer is a pointer which points to nothing. Infact it points to the base address of you CPU

register and since register is not addressable usage of a null pointer will lead to crash or atminimum a segmentation fault.

Also be careful while typecasting one pointer to another because even after type casting your

pointer can point to anything but it will still think it is pointing to something of it declared type

and have properties of the orignal type.

Type conversion is a powerfull feature but yet it may lead difficult to remove bugs and crashes

and should be used with uttermost vigilance. It may also lead to unexpected and unreliable

results but program would compile succesfully.

Code below shows a type casting of one pointer into another -

#include <stdio.h>

int main (){

int i = 10;char* p1int *p2;

p2 = &i;

p1 = (char *) p2; // Type Casting and Pointer Conversion

printf (" *p1 = %c And *p2 = %d", *p1,*p2); // Output maybe unexpected depending

return 0;}




Back To Top

5. Pointer Arithmetic

Pointer arithemetic is quite diffrent from normal arithemetic unless and until you are work on

char type pointers, reason being they are 1 byte (special thnks to Emmanuel Deloget for pointingmistake) long under all enviorments.

Not all artihemetic operations are defined in pointers. You can increment them, decrement them,add and subtract integer values from them. You even can subtract two pointers.But you cannot

add two pointers, mulitply, divide,modulus them. You can not also add or subtract values other

than integer.

Now pointer artihemetic may look a little wierd but it has a deep sense attach to it.

Now consider a pointer X , its current value that address it is pointing to is 1000 (just

assuming).We make another assumption about the size of the data types. Size of data type ismachine dependent, for example int can be 16,32, or 64 bit long depending upon your machine.

Now if this X pointer is char type(assumed 1 Byte or 8Bit long) than X++ will have value 1001

and X-- will have value 999. Now if this X pointer is integer type (assumed 2 byte or 32 bit long)than X++ will have value 1002 and X-- will have value 998. Again if this X pointer is float type

(assumed 4 Byte or 32Bit long) than X++ will have value 1004 and X-- will have value 9996.

Also if this X pointer is double type(assumed 8 Byte or 64 Bit long) than X++ will have value

1008 and X-- will have value 992. Do you see the pattern here.

Reason is when you increment a pointer of certain base type it increase it value in such a way

that it points to next element of its base type. If you decrement a pointer its value decrease insuch a way that it points to previous value of its base type. So increment as well as decrement in

fixed quanta of size of the base type.

You can add or subtract any integer value, in such case value of pointer get increase and decrease

by the productv of the value to be added or subtract and size of the base type.

Pointer of user defined types such as structures and union also increase by the quanta of thier bit

values which can be determined using sizeof operator.

Pointer arithemtic in C may look a bit strange but it is extensively used in programming and

provides unmatched efficiency in performance of task such as accessing an array.

Back To Top

6. Pointer Comparison







Two pointers can be compared no matter where they point. Comparison can be done using <, >,

=, <= and >= operators. Though it is not forcibly implied but comparison of two pointers become

sensible only when they are related such as when they are poiniting to element of samearrays.Comparison of two unrelated pointers is unpredictable and your code should not rely upon

it. All comparison are generally done on basis of memmory organization in the host machine.

Following C source code shows a potential use of pointer comparison in C -

#include <stdio.h>

int main (){

int data[100];int* p1;int *p2;

for (int i = 0; i <100;i++){

data[i] = i;}

p1 = &data [1];p2 = &data [2];

if (p1 > p2){

printf ("\n\n p1 is greater than p2");}

else{

printf ("\n\n p2 is greater than p1");

}}

Back To Top

7. Array and Pointers

Array and Pointers in c are very closely related. Infact they are so similar to each othere in nature

that they can be used interchangebly in each other positions most of the time.

Important link joining them is that array name without the brackets is the pointer name and other

end a pointer can be indexed as if its an array. Consider following C Source Code for betterunderstanding in which cycle through the elements of array using a pointer as a array name andthrough pointer arithemetic.

#include <stdio.h>

int main (){

int data[100];




int* p1;

for (int i = 0; i <100;i++){

data[i] = i;}

p1 = data; //Assigning base address of an array to pointer

for (int i = 0; i <100;i++) //Accessing Array using index{

printf ("\n%d",p1[i]);}

for (int i = 0; i <100;i++) //Access Array using Pointer Arithemetic{

printf ("\n%d",*(p1 +i));}

return 0;}

Pointers like any other data type can be arrayed. This array is called array of pointers and is quite

usefull programming method. Array of pointers are declared as shown below -

data_type *variable_name [array_size];

For eg to declare an array of 5 int pointers we will use declaration -

int *array [5];

The C Source code below will show you a simple use of array of pointers in c programming.

#include <stdio.h>

int main (){

int data[5];int *array[5];

for (int i = 0; i <5;i++){

data[i] = i;}

//Assigning address of elements of array data to array of pointers.

for (int i = 0; i <5;i++)



{array[i] = &data[i];

}

for (int i = 0; i <5;i++) //Accessing Array value using index{

printf ("\n%d",data[i]);}

for (int i = 0; i <5;i++) //Access Array value using array of pointers{

printf ("\n%d",*array[i] );}

return 0;}

Back To Top

8. Multiple Indirection

Well, in C it is permitted for a pointer to point to another pointer. As a result many layers of

pointer can be formed and this called multiple indirection. As such there is no limitation on the

level of indirection but including great depth of indirection is difficult to follow and may lead toerrors. Normally more than 2 level depth indirection is not required.

A pointer to a pointer has declaration is similar to that of a normal pointer but have more asterixsign before them indicating the depth of the pointer.

int ** pointer

The above statement creates a pointer which points to pointer to a variable with int value.

Following c source code will help you clear any doubts on multiple indirection.

#include <stdio.h>

int main (){

int i = 10;int **p1int *p2;

p2 = &i;p1 = &p2; // Multiple indirection

printf (" **p1 = %d And *p2 = %d", **p1,*p2); //Statement will show 10 twice.

return 0;}




Back To Top

9. Pointer Intialization

When it comes to pointer intialization, c programming standard goes nut. Standard does not

forces any intialization to a non static local pointer but it does state that global pointers and static pointers get automatically initalize to null. So unintialized non static local pointer contains anunknown value. If you try to use this pointer without giving it proper value you may crash your

program and even operating system in extreme cases and permanent damage to computer in term

of software and hardware can occur. So pointers are very nasty if used incorrectly.

Most important term in pointer intialization would be "null". It is a standard practice of top

qualilty c programmers that any pointer which is not pointing to valid memmory location should be intialized to null showing that it is pointing to nothing and should not be used. In fact a null

pointer is pointing to something, it is pointing to the base address of your CPU register. Now

since register of CPU cannot be addressed the use of such pointer leads to abormal termination of

the program with a probable segmentation fault at runtime but no serious damage will be done toyour machine.A MACRO name NULL is also defined in many standard headers such as<stdio.h>.

You can initialize a pointer to null using 2 methods as shown below -

variable_type *pointer_name = 0;

or

variable_type *pointer_name = NULL;

Even though null pointer should not be used directly, it does have some indirect uses such as,they can be used in a comparison with other pointer. They mark an end of something such as

thier extensive use in data structures.

Also char pointers which are capable of saving strings or char array can be intialized in more

spectacular way as shown below -

char *pointer_name = "string value here";

In such a statement you may wonder since p is not an array and no memmory is allocated to it so

where does the string is actually stored. Answer to this question can be found in the technique

used by C Compilers to handle string constants. To handle such constants the compiler creates astring table in which dynamic memmory is allocated to pointer from heap and string is stored in

this string table.

Back To Top

10. Pointer To A Function







One of the rarest used yet powerfull feature of C programming language is the pointer to a

function. As most of the people dont know that every function defined in c programming

language have a base address attached to it. This base address acts as an entering point into thatfunction. As you can now guess this address can be stored in a specially designed pointer known

as function pointer. Address of the function can be obtained by using pointer name without

parenthesis.

Once stored this pointer can be of tremendous use such that it can be used to calll that function

but most important feast it let you attain is that it allows you to pass function as arguments toother functions as if it was a variable.

The most tricky thing about pointer to a function is declaration of the pointer that holds the baseaddress of the function, rest is piece of cake.The general form of declaration of a pointer to a

function is -

return_type (* pointer_name) ( variable1_type variable1_name ,variable2_type variable2_name,....);

The declaration above tells compiler that pointer of name pointer_name will store a address offunction which return a value of return_type type. Also it will take variables of name

variable1_name, variable2_name of type variable1_type, variable2_type respectively and so on.

The simple c source code below will further clear your concept of pointer to a function in c programming.

#include <stdio.h>

int add (int a, int b, int c)

{return a+b+c;}int main ()

{int (*function_pointer) (int,int,int); //Declaration of function pointer for storafunction_pointer = add;printf ("%d", function_pointer (1,2,3)); //Print 6 as outputreturn 0;

}

Back To Top

11. Dynamic MemmoryAllocation Using Pointers

Dynamic memmory allocation is one of the most powerfull feature of the c programming

language because it is the tool which lets you design dynamic softwares using c. Dynamic

software is something which changes itself on runtime according to situations created during

executions but static program in all circumstances follows the same path of execution. Through




DMA (Dynamic Memmory Allocation) you can obtain memmory at runtime. As you will be

knowing global variables are allocated storage at compile time from heap while local variables

are also allocated memmory from stack by the compiler, but in sum situations especailly true fordynamic programs that memmory requirement cannot be known at compile time but depends

upon the input user gives interaction or some other dynamic values which keeps changing. In

such cases memmory requirement of the program may expand or shrink at run time and in thisDMA comes handy. DMA is also extensively used in data structures such as link list, trees etc because their innate dynamic nature.

Memmory allocated by DMA is from heap. Heap is a free memmory available to your program

and can be safely worked on since it is not used by any other program, O.S. etc. Heap is

generally very large but can be exhausted as it is finite. Heap can also so be a severe restraining

factor in special situations such as embedded systems and microprocessors, keeping in mind thatthey have very limited memmory resources. So it is surely a good practice to return the

memmory back to heap when it has no longer used so that it can be reentered in the

circualtionfor reuse. Also size of the heap can be determined by some compiler dependant

functions but no function for this purpose is defined in C Standard.

Their are four functions defined in c standard for dynamic memmory allocation - calloc, free,malloc and realloc. But in the heart of DMA there are only 2 of them malloc and free. Malloc

stands for memmory allocations and is used to allocate memmory from the heap while free is

used to return allocated memmory from malloc back to heap. Both these functions uses a

standard library header <stdlib.h>

Warning !!! - free ( ) function should be used to free memmory only allocated previously from

malloc, realloc or calloc. Freeing a random or undefined or compiler allocated memmory canlead to severe damage to the O.S., Compiler and Computer Hardware Itself, in form of nasty

system crashes.

The prototype of malloc ( ) function is -

void *malloc (size_t number_of_bytes)

Important thing to nore is malloc return a void pointer which can be converted to any pointer

type as explained in previous points. Also size_t is a special type of unsigned integer defined in<stdlib.h> capable of storing largest memmory size that can be allocated using DMA,

number_of_bytes is a value of type size_t generally a integer indicating the amount of memmory

to be allocated. Function malloc ( ) will be returning a null pointer if memmory allocation fails

and will return a pointer to first region of memmory allocated when succsefull. It is alsorecommended you check the pointer returned for failure in allocation before using the returned

memmory for increasing stability of your program, generally programmers provide some error

handling code in case of failures. Also this returned pointer never needs a typecast in C since it isa void pointer, it is a good practice to do one since it is required by C++ and will produce a error

if you used C++ compiler for compilation.



Another commonly used operator used with malloc is sizeof operator which is used to calculate

the value of number_of_bytes by determing the size of the compiler as well as user defined types

and variables.

The prototype of free ( ) function is -

void free (void *p)

Function free ( ) is opposite of malloc and is used to return memmory previously allocated byother DMA functions. Also only memmory allocated using DMA should be free using free ()

otherwise you may corrupt your memmory allocation system at minimum.

C Source code shown below shows simple method of using dynamic memmory allocation

elegantly -

#include <stdio.h>#include <stdlib.h>

int main ()

{

int *p;

p = (int *) malloc ( sizeof (int) ); //Dynamic Memmory Allocation

if (p == NULL) //Incase of memmory allocation failure execute the error handling cod{

printf ("\nOut of Memmory");exit (1);}

*p = 100;

printf ("\n p = %d", *p); //Display 100 ofcourse.

return 0;}

Also the normal arrays can be increased in power and flexibilty using Dynamic memmoryallocation to be converted into dynamic allocated arrays. These dynamic allocated arrays though

have a little bit of complication involved with them in usage, so read carefully the explanationgiven below. Also thier declaration varies entirely.

For 1 - D array it is quite simple. Allocate ample memmory using DMA and then access arrayeither using pointer arithemetic or indexing as shown in C Source Code below -

#define SIZE 10 //Size of 1D Array




int main ()

{

int *p;

p = (int *) malloc ( SIZE * sizeof (int) ); //Dynamic Memmory Allocation of 1D Array

if (p == NULL) //Incase of memmory allocation failure execute the error handling cod {

printf ("\nOut of Memmory");exit (1);

}

for (i = 0; i<SIZE; i++)

{p [i] = i; // Loading the Array

}

for (i = 0; i<SIZE; i++){

printf ("\n%d", *(p + i)); // Displaying the Array}

return 0;}

In case of 2-D array allocation similar to that above can be used but in that case accessing the

array is not possible using indexing the pointer to which memmory being allocted reason beingthat the left index is not defined for the pointer that is it is not possible for compiler to know how

many rows and columns in the 2D array. In such cases you have to design code to access valuesusing complicated pointer arithemetic.

On other hand you can declare a pointer in a special way that compiler is informed of structure ofthe 2D array so that access to that array can be provide using indexing or simple pointer

arithemetic as shown in C Source code below -

#define SIZE1 10 //Number of Rows in 2D Array#define SIZE2 10 //Number of Columns in 2D Array


int main ()

{int (*p) [SIZE1]; //Observe the special pointer declaration to tell compiler 2D arra



p = (int (*) [SIZE1]) malloc ( SIZE1 * SIZE2 * sizeof (int) ); //Dynamic Memmory All

if (p == NULL) //Incase of memmory allocation failure execute the error handling cod {

printf ("\nOut of Memmory");exit (1);

}

for (j = 0; j<SIZE1; j++){for (i = 0; i<SIZE2; i++){

p [j] [i] = j*SIZE1 +i; // Loading the Array}

}

for (j = 0; j<SIZE1; j++){

for (i = 0; i<SIZE2; i++){

printf ("\n%d",p [j] [i]); // Displaying the Array}

}return 0;}

The concept above can be extended to 3D array and other complex multidimensional arrays.

Back To Top

12. Restrict Pointers

This a very innovative new feature added by C99. This is a brand new type qualifier applicable

only to pointers. Restrict qualifiers tells the compiler that pointer is the only mean by which the

object can be accessed. Another pointer can access it only if its based on the first one. Thisenables complier to produce highly efficient code by assuming that if two pointers are declared

as restrict then these two pointers doesnot overlap. The primary use in pointers passed in

function parameter and standard library makes an extensive use of this powerpacked feature of

C. One important thing to remember is under any circumstances restrict pointer doesnot changethe semantics of the program. Restrict pointers may sometime allow the complier to optimize

code.

Also as mentioned by Emmanuel Deloget below in the comments, in c standard the behaviour is

marked undefined when two restrict pointers overlap. The standard lef t it to compilers to

implement them. Some may produce compile time error, some may produce run time error andmost will give atleast a warning. In a rare case compiler can even ignore it leading too extremely

dangerous bugs and errors. You should read your C compilers documentation for futher

information about its strategy on restrict pointers.




even a warning is only a assumption and may not occur. The reason being compiler except this

because every point in memmory is represented in hexa decimal number which this decimal

number is implicitly converted. It is a rare but usefull thing in special circumstances to usehardcode memmory values especially in case of hardware programming and system

programming where values of memmory address such as port address always remain constant for

the machine for which the program is being used. But this concept allows accedential error likeabove which may prove dangerous. So stay cautious of such error while programming.

Correct and Safe Code would be -

#include <stdio.h>#include <alloc.h>

int main (){int a, *p = NULL;p = (int *) malloc (sizeof (int));a = 1;

p = &a;prinf ("%d",*p);return 0;}

>> Incorrect Pointer Comparison

Carefully look at the c source code below -


int main (){int *p = NULL, *q = NULL;p = (int *) malloc (sizeof (int));q = (int *) malloc (sizeof (int));if ( p < q )prinf ("\np is less than q"); //This will print garbage value

elseprintf ("\np is greater than q");

return 0;}

The comparison of pointer as done above is absolutely incorrect reason being that both p and q

pointers in code above are entirely unrelated and the ouput of the above code cannot be predicted. Comparison done above is entirely dependent upon compiler, operating system and to

great extent luck. When pointer comparison is performed, the comparison is done between the

integer values of the memmory locations they are pointing to. Now this memmory location

assigned to them is not fixed and is entirely random so p can be either less or greater than q eachwith 50% probability in mathematics terms. So the comparison cannot be relied upon is the

bottom line. But this comparison if used correctly can be usefull in logically related memmory



allocation such as in arrays. Also in very very very rare conditions you may even want to use this

unrelated pointer comparison.

>> Failure to do pointer reintialization

Carefully look at the c source code below -


int main (){int i;int *p = (int *) malloc ( sizeof (int) * 10 );int *q;

q = p;

for ( i = 0; i < 10; i++){

*p = i;p++;

}

for ( i = 0; i < 10; i++){

printf (" %d ", *p); //Prints Garbage Value*p++;

}

return 0;}

The mistake above is very common and creates a bug very tough to found. The point is when in

above code p is allocated memmory and used to load int values using pointer arithemetic and

then when same pointer is used to display the loaded values output will be garbage value. The

reason for this that p is not reinitalized to memmory address it was given during dynamicmemmory allocation. So p keeps running ahead of its last left poisition that is from the end of its

allocated memmory and runs into unknown memmory locations next in serial and only god

knows what it will read and destroy when used for writting memmory location. This problem can be easily avoided by reinitializing p to its original value.

Correct and Safe Code would be -


int main (){int i;int *p = (int *) malloc ( sizeof (int) * 10 );int *q;



q = p;

for ( i = 0; i < 10; i++){

*p = i;p++;

}

p = q;

for ( i = 0; i < 10; i++){

printf (" %d ", *p); //Prints Value 1 to 10;*p++;

}

return 0;}

c programming reference

Documents