[c] How to reverse a singly linked list using only two pointers?

I wonder if there exists some logic to reverse a singly-linked list using only two pointers.

The following is used to reverse the single linked list using three pointers namely p, q, r:

struct node {
    int data;
    struct node *link;
};

void reverse() {
    struct node *p = first,
                *q = NULL,
                *r;

    while (p != NULL) {
        r = q;
        q = p;
        p = p->link;
        q->link = r;
    }
    first = q;
}

Is there any other alternate to reverse the linked list? What would be the best logic to reverse a singly linked list, in terms of time complexity?

Just for fun (although tail recursion optimization should stop it eating all the stack):

Node* reverse (Node *root, Node *end) {

    Node *next = root->next;
    root->next = end;

    return (next ? reverse(next, root) : root);
}

root = reverse(root, NULL);

Yes there is a way using only two pointers. That is by creating new linked list where the first node is the first node of the given list and second node of the first list is added at the start of the new list and so on.

To swap two variables without the use of a temporary variable,

a = a xor b
b = a xor b
a = a xor b

fastest way is to write it in one line

a = a ^ b ^ (b=a)

Similarly,

using two swaps

swap(a,b)
swap(b,c)

solution using xor

a = a^b^c
b = a^b^c
c = a^b^c
a = a^b^c

solution in one line

c = a ^ b ^ c ^ (a=b) ^ (b=c)
b = a ^ b ^ c ^ (c=a) ^ (a=b)
a = a ^ b ^ c ^ (b=c) ^ (c=a)

The same logic is used to reverse a linked list.

typedef struct List
{
 int info;
 struct List *next;
}List;


List* reverseList(List *head)
{
 p=head;
 q=p->next;
 p->next=NULL;
 while(q)
 {
    q = (List*) ((int)p ^ (int)q ^ (int)q->next ^ (int)(q->next=p) ^ (int)(p=q));
 }
 head = p;
 return head;
}  

#include<stdio.h>
#include<conio.h>
#include<stdlib.h>
struct node
{
int data;
struct node *link;
};
struct node *first=NULL,*last=NULL,*next,*pre,*cur,*temp;
void create()
{
cur=(struct node*) malloc(sizeof(struct node));
printf("enter first data to insert");
scanf("%d",&cur->data);
first=last=cur;
first->link=NULL;
}
void insert()
{
int pos,c;
cur=(struct node*) malloc(sizeof(struct node));
printf("enter data to insert and also its position");
scanf("%d%d",&cur->data,&pos);
if(pos==1)
{
cur->link=first;
first=cur;
}
else
{
c=1;
    next=first;
    while(c<pos)
    {
        pre=next;
        next=next->link;
        c++;
    }
        if(pre==NULL)
        {
            printf("Invalid position");
        }
        else
        {
        cur->link=pre->link;
        pre->link=cur;
        }
}
}
void display()
{
cur=first;
while(cur!=NULL)
{
printf("data= %d\t address= %u\n",cur->data,cur);
cur=cur->link;
}
printf("\n");
}
void rev()
{
pre=NULL;
cur=first;
while(cur!=NULL)
{
next=cur->link;
cur->link=pre;
pre=cur;
cur=next;
}
first=pre;
}
void main()
{
int choice;
clrscr();
do
{
printf("Options are: -\n1:Create\n2:Insert\n3:Display\n4:Reverse\n0:Exit\n");
printf("Enter your choice: - ");
scanf("%d",&choice);
switch(choice)
{
case 1:
create();
break;
case 2:
insert();
break;
case 3:
display();
break;
case 4:
rev();
break;
case 0:
exit(0);
default:
printf("wrong choice");
}
}
while(1);
}

Here's a simpler version in python. It does use only two pointers slow & fast

def reverseList(head: ListNode) -> ListNode:

    slow = None
    fast = head
    while fast:  
        node_next = fast.next
        
        fast.next = slow
        slow = fast
            
        fast = node_next
    return slow

You can go for recursive approach:

Here is the pseudo code:

Node* reverse(Node* root)
{
    if(!root) return NULL;

    if(!(root->next)) temp = root;
    else
    {
        reverse(root->next);
        root->next->next = root;
        root->next = NULL;
    }

    return temp;
}

After the call is made to the function, it returns the new root[temp] of the linked list. As it is very clear that it makes use of only two pointers.

How about the more readable:

Node *pop (Node **root)
{
    Node *popped = *root;

    if (*root) {
        *root = (*root)->next;
    }

    return (popped);
}

void push (Node **root, Node *new_node)
{
    new_node->next = *root;
    *root = new_node;
}


Node *reverse (Node *root)
{
    Node *new_root = NULL;
    Node *next;

    while ((next = pop(&root))) {
        push (&new_root, next);
    }

    return (new_root);
}

class Node {
    Node next;
    int data;

    Node(int item) {
        data = item;
        next = null;
    }
}

public class LinkedList {

    static Node head;

    //Print LinkedList
    public static void printList(Node node){

        while(node!=null){
            System.out.print(node.data+" ");
            node = node.next;
        }
        System.out.println();
    }

    //Reverse the LinkedList Utility
    public static Node reverse(Node node){

        Node new_node = null;

        while(node!=null){

            Node next = node.next;
            node.next = new_node;
            new_node = node;
            node = next;

        }
        return new_node;
    }

    public static void main(String[] args) {

        //Creating LinkedList
        LinkedList.head = new Node(1);
        LinkedList.head.next = new Node(2);
        LinkedList.head.next.next = new Node(3);
        LinkedList.head.next.next.next = new Node(4);

        LinkedList.printList(LinkedList.head);

        Node node = LinkedList.reverse(LinkedList.head);

        LinkedList.printList(node);

    }


}

A simple algorithm if you use the linked list as a stack structure:

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

typedef struct list {
    int key;
    char value;
    struct list* next;
} list;
void print(list*);
void add(list**, int, char);
void reverse(list**);
void deleteList(list*);

int main(void) {
    list* head = NULL;
    int i=0;
    while ( i++ < 26 ) add(&head, i, i+'a');
    printf("Before reverse: \n");
    print(head);
    printf("After reverse: \n");
    reverse(&head);
    print(head);
    deleteList(head);

}
void deleteList(list* l) {

    list* t = l;    
    while ( t != NULL ) {
        list* tmp = t;
        t = t->next;
        free(tmp);
    }

}
void print(list* l) {
    list* t = l;
    while ( t != NULL) {
        printf("%d:%c\n", t->key, t->value);
        t = t->next;
    }
}

void reverse(list** head) {
    list* tmp = *head;
    list* reversed = NULL;
    while ( tmp != NULL ) {
        add(&reversed, tmp->key, tmp->value);
        tmp = tmp->next;
    }
    deleteList(*head);
    *head = reversed;
}

void add(list** head, int k, char v) {

    list* t = calloc(1, sizeof(list));
    t->key = k; t->value = v;
    t->next = *head;
    *head = t;

}

The performance may be affected since additional function call to the add and malloc so the algorithms of address swaps are better but that one actually creates new list so you can use additional options like sort or remove items if you add a callback function as parameter to the reverse.

As an alternative, you can use recursion-

struct node* reverseList(struct node *head)
{
    if(head == NULL) return NULL;
    if(head->next == NULL) return head;

    struct node* second = head->next;       
    head->next = NULL;

    struct node* remaining = reverseList(second);
    second->next = head;

    return remaining;
}

Here is my version:

void reverse(ListElem *&head)
{
    ListElem* temp;
    ListElem* elem = head->next();
    ListElem* prev = head;
    head->next(0);

    while(temp = elem->next())
    {
        elem->next(prev);
        prev = elem;
        elem = temp;
    }
    elem->next(prev);
    head = elem;
}

where

class ListElem{
public:
    ListElem(int val): _val(val){}
    ListElem *next() const { return _next; }
    void next(ListElem *elem) { _next = elem; }
    void val(int val){ _val = val; }
    int val() const { return _val;}
private:
    ListElem *_next;
    int _val;
};

You can simply reverse a Linked List using only one Extra pointer. And the key to do this is by using a Recursion.

Here is the program in Java.

public class Node {
   public int data;
   public Node next;
}

public Node reverseLinkedListRecursion(Node p) {
    if (p.next == null) {
        head = p;
        q = p;
        return q;
    } else {
        reverseLinkedListRecursion(p.next);
        p.next = null;
        q.next = p;
        q = p;
        return head;
    }
}

// call this function from your main method.
 reverseLinkedListRecursion(head);

As you can see this is a simple example of a head recursion. We have mainly two different kinds of Recursion.

1. Head Recursion:- When the Recursion is the first thing executed by a function.
2. Tail Recursion:- When the Recursion is the last thing executed by a function.

Here the program will keep calling itself Recursively until our Pointer "p" reaches to the last node and then before returning the stack frame we will point head to the last node and the extra Pointer "q" to build the linked list in the backward direction.

Here the Stack Frames will keep on returning until the stack is empty.

here is a little simple solution...

void reverse()
{
    node * pointer1 = head->next;
    if(pointer1 != NULL)
    {
        node *pointer2 = pointer1->next;
        pointer1->next = head;
        head->next = NULL;
        head = pointer1;

        if(pointer2 != NULL)
        {

            while(pointer2 != NULL)
            {
                pointer1 = pointer2;
                pointer2 = pointer2->next;
                pointer1->next = head;
                head = pointer1;
            }

            pointer1->next = head;
            head = pointer1;
        }       
   }
 }

I have a slightly different approach. I wanted to make use of the existing functions (like insert_at(index), delete_from(index)) to reverse the list (something like a right shift operation). The complexity is still O(n) but the advantage is more reused code. Have a look at another_reverse() method and let me know what you all think.

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

struct node {
    int data;
    struct node* next;
};

struct node* head = NULL;

void printList(char* msg) {
    struct node* current = head;

    printf("\n%s\n", msg);

    while (current != NULL) {
        printf("%d ", current->data);
        current = current->next;
    }
}

void insert_beginning(int data) {
    struct node* newNode = (struct node*) malloc(sizeof(struct node));

    newNode->data = data;
    newNode->next = NULL;

    if (head == NULL)
    {
        head = newNode;
    } else {
        newNode->next = head;
        head = newNode;
    }
}

void insert_at(int data, int location) {

    struct node* newNode = (struct node*) malloc(sizeof(struct node));

    newNode->data = data;
    newNode->next = NULL;

    if (head == NULL)
    {
        head = newNode;
    }

    else {
        struct node* currentNode = head;
        int index = 0;

        while (currentNode != NULL && index < (location - 1)) {
            currentNode = currentNode->next;
            index++;
        }

        if (currentNode != NULL)
        {
            if (location == 0) {
                newNode->next = currentNode;
                head = newNode;
            } else {
                newNode->next = currentNode->next;
                currentNode->next = newNode;
            }
        }
    }
}


int delete_from(int location) {

    int retValue = -1;

    if (location < 0 || head == NULL)
    {
        printf("\nList is empty or invalid index");
        return -1;
    } else {

        struct node* currentNode = head;
        int index = 0;

        while (currentNode != NULL && index < (location - 1)) {
            currentNode = currentNode->next;
            index++;
        }

        if (currentNode != NULL)
        {
            // we've reached the node just one prior to the one we want to delete

            if (location == 0) {

                if (currentNode->next == NULL)
                {
                    // this is the only node in the list
                    retValue = currentNode->data;
                    free(currentNode);
                    head = NULL;
                } else {

                    // the next node should take its place
                    struct node* nextNode = currentNode->next;
                    head = nextNode;
                    retValue = currentNode->data;
                    free(currentNode);
                }
            } // if (location == 0)
            else {
                // the next node should take its place
                struct node* nextNode = currentNode->next;
                currentNode->next = nextNode->next;

                if (nextNode != NULL
                ) {
                    retValue = nextNode->data;
                    free(nextNode);
                }
            }

        } else {
            printf("\nInvalid index");
            return -1;
        }
    }

    return retValue;
}

void another_reverse() {
    if (head == NULL)
    {
        printf("\nList is empty\n");
        return;
    } else {
        // get the tail pointer

        struct node* tailNode = head;
        int index = 0, counter = 0;

        while (tailNode->next != NULL) {
            tailNode = tailNode->next;
            index++;
        }

        // now tailNode points to the last node
        while (counter != index) {
            int data = delete_from(index);
            insert_at(data, counter);
            counter++;
        }
    }
}

int main(int argc, char** argv) {

    insert_beginning(4);
    insert_beginning(3);
    insert_beginning(2);
    insert_beginning(1);
    insert_beginning(0);

    /*  insert_at(5, 0);
     insert_at(4, 1);
     insert_at(3, 2);
     insert_at(1, 1);*/

    printList("Original List\0");

    //reverse_list();
    another_reverse();

    printList("Reversed List\0");

    /*  delete_from(2);
     delete_from(2);*/

    //printList();
    return 0;
}

Robert Sedgewick, "Algorithms in C", Addison-Wesley, 3rd Edition, 1997, [Section 3.4]

In case that is not a cyclic list ,hence NULL is the last link.

typedef struct node* link;

struct node{
int item;
link next;
};
/* you send the existing list to reverse() and returns the reversed one */
link reverse(link x){
link t, y = x, r = NULL;
while(y != NULL){
t = y->next;
y-> next = r;
r = y;
y = t;
}
return r;
}

using 2-pointers....bit large but simple and efficient

void reverse()

{

int n=0;

node *temp,*temp1;

temp=strptr;

while(temp->next!=NULL)

{

n++;      //counting no. of nodes

temp=temp->next;

}
// we will exchange ist by last.....2nd by 2nd last so.on....
int i=n/2;  

temp=strptr;

for(int j=1;j<=(n-i+1);j++)

temp=temp->next;
//  i started exchanging from in between ....so we do no have to traverse list so far //again and again for exchanging

while(i>0)

{

temp1=strptr;

for(int j=1;j<=i;j++)//this loop for traversing nodes before n/2

temp1=temp1->next;

int t;

t=temp1->info;

temp1->info=temp->info;

temp->info=t;

i--;

temp=temp->next; 

//at the end after exchanging say 2 and 4 in a 5 node list....temp will be at 5 and we will traverse temp1 to ist node and exchange ....

}

}

Here's the code to reverse a singly linked list in C.

And here it is pasted below:

// reverse.c

#include <stdio.h>
#include <assert.h>

typedef struct node Node;
struct node {
    int data;
    Node *next;
};

void spec_reverse();
Node *reverse(Node *head);

int main()
{
    spec_reverse();
    return 0;
}

void print(Node *head) {
    while (head) {
        printf("[%d]->", head->data);
        head = head->next;
    }
    printf("NULL\n");
}

void spec_reverse() {
    // Create a linked list.
    // [0]->[1]->[2]->NULL
    Node node2 = {2, NULL};
    Node node1 = {1, &node2};
    Node node0 = {0, &node1};
    Node *head = &node0;

    print(head);
    head = reverse(head);
    print(head);

    assert(head == &node2);
    assert(head->next == &node1);
    assert(head->next->next == &node0);

    printf("Passed!");
}

// Step 1:
//
// prev head  next
//   |    |    |
//   v    v    v
// NULL  [0]->[1]->[2]->NULL
//
// Step 2:
//
//      prev head  next
//        |    |    |
//        v    v    v
// NULL<-[0]  [1]->[2]->NULL
//
Node *reverse(Node *head)
{
    Node *prev = NULL;
    Node *next;

    while (head) {
        next = head->next;
        head->next = prev;
        prev = head;
        head = next;
    }

    return prev;
}

I don't understand why there is need to return head as we are passing it as argument. We are passing head of the link list then we can update also. Below is simple solution.

#include<stdio.h>
#include<conio.h>

struct NODE
{
    struct NODE *next;
    int value;
};

typedef struct NODE node;

void reverse(node **head);
void add_end(node **head,int val);
void alloc(node **p);
void print_all(node *head);

void main()
{
    node *head;
    clrscr();
    head = NULL;
    add_end( &head, 1 );
    add_end( &head, 2 );
    add_end( &head, 3 );
    print_all( head );
    reverse( &head );
    print_all( head );
    getch();
}
void alloc(node **p)
{
    node *temp;
    temp = (node *) malloc( sizeof(node *) );
    temp->next = NULL;
    *p = temp;
}
void add_end(node **head,int val)
{
    node *temp,*new_node;
    alloc(&new_node);
    new_node->value = val;
    if( *head == NULL )
    {
        *head = new_node;
        return;
    }
    for(temp = *head;temp->next!=NULL;temp=temp->next);
    temp->next = new_node;
}
void print_all(node *head)
{
    node *temp;
    int index=0;
    printf ("\n\n");
    if (head == NULL)
    {
        printf (" List is Empty \n");
        return;
    }
    for (temp=head; temp != NULL; temp=temp->next,index++)
        printf (" %d ==> %d \n",index,temp->value);
}
void reverse(node **head)
{
    node *next,*new_head;
    new_head=NULL;
    while(*head != NULL)
    {
        next = (*head)->next;
        (*head)->next = new_head;
        new_head = (*head);
        (*head) = next;
    }
    (*head)=new_head;
}

Following is one implementation using 2 pointers (head and r)

ListNode * reverse(ListNode* head) {

    ListNode *r = NULL;

    if(head) {
        r = head->next;
        head->next = NULL;
    }

    while(r) {
        head = reinterpret_cast<ListNode*>(size_t(head) ^ size_t(r->next));
        r->next = reinterpret_cast<ListNode*>(size_t(r->next) ^ size_t(head));
        head = reinterpret_cast<ListNode*>(size_t(head) ^ size_t(r->next));

        head = reinterpret_cast<ListNode*>(size_t(head) ^ size_t(r));
        r = reinterpret_cast<ListNode*>(size_t(r) ^ size_t(head));
        head = reinterpret_cast<ListNode*>(size_t(head) ^ size_t(r));
    }
    return head;
}

//with this no extra space and no extra scans but this code is reversing code but read
// in reverse direction no changes made in the linked list 
PrintInReverse(Node node)
{
   // given list is null
   if(node ==null)
       return null;
   // if list contains only one node
   if(node->next ==null)
   {
       print(node.value)
   }
   // call recursively 
   else
   {
       //while(node->next != null)// due to while loop it goes into infinite loop.use //if
       if(node->next!=NULL)
       {
           PrintInReverse(node->next)
           print(node.value)
       }
   }
}

Add comments...

I hate to be the bearer of bad news but I don't think your three-pointer solution actually works. When I used it in the following test harness, the list was reduced to one node, as per the following output:

==========
4
3
2
1
0
==========
4
==========

You won't get better time complexity than your solution since it's O(n) and you have to visit every node to change the pointers, but you can do a solution with only two extra pointers quite easily, as shown in the following code:

#include <stdio.h>

// The list element type and head.

struct node { 
    int data;
    struct node *link;
};
static struct node *first = NULL;

// A reverse function which uses only two extra pointers.

void reverse() {
    // curNode traverses the list, first is reset to empty list.
    struct node *curNode = first, *nxtNode;
    first = NULL;

    // Until no more in list, insert current before first and advance.
    while (curNode != NULL) {
        // Need to save next node since we're changing the current.
        nxtNode = curNode->link;

        // Insert at start of new list.
        curNode->link = first;
        first = curNode;

        // Advance to next.
        curNode = nxtNode;
    }
}

// Code to dump the current list.

static void dumpNodes() {
    struct node *curNode = first;
    printf ("==========\n");
    while (curNode != NULL) {
        printf ("%d\n", curNode->data);
        curNode = curNode->link;
    }
}

// Test harness main program.

int main (void) {
    int i;
    struct node *newnode;

    // Create list (using actually the same insert-before-first
    // that is used in reverse function.

    for (i = 0; i < 5; i++) {
        newnode = malloc (sizeof (struct node));
        newnode->data = i;
        newnode->link = first;
        first = newnode;
    }

    // Dump list, reverse it, then dump again.

    dumpNodes();
    reverse();
    dumpNodes();
    printf ("==========\n");

    return 0;
}

This code outputs:

==========
4
3
2
1
0
==========
0
1
2
3
4
==========

which I think is what you were after. It can actually do this since, once you've loaded up first into the pointer traversing the list, you can re-use first at will.

Solution using 1 variable (Only p):

typedef unsigned long AddressType;

#define A (*( AddressType* )&p )
#define B (*( AddressType* )&first->link->link )
#define C (*( AddressType* )&first->link )

/* Reversing linked list */
p = first;

while( first->link )
{
    A = A + B + C;
    B = A - B - C;
    A = A - B;
    C = A - C;
    A = A - C;
}

first = p;

#include <stddef.h>

typedef struct Node {
    struct Node *next;
    int data;
} Node;

Node * reverse(Node *cur) {
    Node *prev = NULL;
    while (cur) {
        Node *temp = cur;
        cur = cur->next; // advance cur
        temp->next = prev;
        prev = temp; // advance prev
    }
    return prev;
}

Here's a simpler version in Java. It does use only two pointers curr & prev

public void reverse(Node head) {
    Node curr = head, prev = null;

    while (head.next != null) {
        head = head.next; // move the head to next node
        curr.next = prev; //break the link to the next node and assign it to previous
        prev = curr;      // we are done with previous, move it to next node
        curr = head;      // current moves along with head
    }

    head.next = prev;     //for last node
}

curr = head;
prev = NULL;

while (curr != NULL) {
    next = curr->next; // store current's next, since it will be overwritten
    curr->next = prev;
    prev = curr;
    curr = next;
}

head = prev; // update head

Yes. I'm sure you can do this the same way you can swap two numbers without using a third. Simply cast the pointers to a int/long and perform the XOR operation a couple of times. This is one of those C tricks that makes for a fun question, but doesn't have any practical value.

Can you reduce the O(n) complexity? No, not really. Just use a doubly linked list if you think you are going to need the reverse order.

You can have solution of this problem with help of only one extra pointer, that has to be static for the reverse function. It's in O(n) complexity.

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

typedef struct List* List;
struct List {
   int val;
   List next;
};

List reverse(List list) { /* with recursion and one static variable*/
    static List tail;
    if(!list || !list->next) {
        tail = list;

        return tail;
    } else {
        reverse1(list->next);
        list->next->next = list;
        list->next = NULL;

        return tail;
    }
}

Here is a slightly different, but simple approach in C++11:

#include <iostream>

struct Node{
    Node(): next(NULL){}
    Node *next;
    std::string data;
};

void printlist(Node* l){
    while(l){
        std::cout<<l->data<<std::endl;
        l = l->next;
    }
    std::cout<<"----"<<std::endl;
}

void reverse(Node*& l)
{
    Node* prev = NULL;
    while(l){
        auto next = l->next;
        l->next = prev;
        prev=l;
        l=next;
    }
    l = prev;
}

int main() {
    Node s,t,u,v;
    s.data = "1";
    t.data = "2";
    u.data = "3";
    v.data = "4";
    s.next = &t;
    t.next = &u;
    u.next = &v;
    Node* ptr = &s;
    printlist(ptr);
    reverse(ptr);
    printlist(ptr);
    return 0;
}

Output here

You need a track pointer which will track the list.

You need two pointers :

first pointer to pick first node. second pointer to pick second node.

Processing :

Move Track Pointer

Point second node to first node

Move First pointer one step, by assigning second pointer to one

Move Second pointer one step, By assigning Track pointer to second

Node* reverselist( )
{
   Node *first = NULL;  // To keep first node
   Node *second = head; // To keep second node
   Node *track =  head; // Track the list

    while(track!=NULL)
    {
      track = track->next; // track point to next node;
      second->next = first; // second node point to first
      first = second; // move first node to next
      second = track; // move second node to next
    }

    track = first;

    return track;

}

No, nothing faster than the current O(n) can be done. You need to alter every node, so time will be proportional to the number of elements anyway and that's O(n) you already have.

Using two pointers while maintaining time complexity of O(n), the fastest achievable, might only be possible through number casting of pointers and swapping their values. Here is an implementation:

#include <stdio.h>

typedef struct node
{
    int num;
    struct node* next;
}node;

void reverse(node* head)
{
   node* ptr;
   if(!head || !head->next || !head->next->next) return;
   ptr = head->next->next;
   head->next->next = NULL;
   while(ptr)
   {
     /* Swap head->next and ptr. */
     head->next = (unsigned)(ptr =\
     (unsigned)ptr ^ (unsigned)(head->next =\
     (unsigned)head->next ^ (unsigned)ptr)) ^ (unsigned)head->next;

     /* Swap head->next->next and ptr. */
     head->next->next = (unsigned)(ptr =\
     (unsigned)ptr ^ (unsigned)(head->next->next =\
     (unsigned)head->next->next ^ (unsigned)ptr)) ^ (unsigned)head->next->next;
   }
}

void add_end(node* ptr, int n)
{
    while(ptr->next) ptr = ptr->next;
    ptr->next = malloc(sizeof(node));
    ptr->next->num = n;
    ptr->next->next = NULL;
}

void print(node* ptr)
{
    while(ptr = ptr->next) printf("%d ", ptr->num);
    putchar('\n');
}

void erase(node* ptr)
{
    node *end;
    while(ptr->next)
    {
        if(ptr->next->next) ptr = ptr->next;
        else
        {
            end = ptr->next;
            ptr->next = NULL;
            free(end);
        }
    }
}

void main()
{
    int i, n = 5;
    node* dummy_head;
    dummy_head->next = NULL;
    for(i = 1; i <= n ; ++i) add_end(dummy_head, i);
    print(dummy_head);
    reverse(dummy_head);
    print(dummy_head);
    erase(dummy_head);
}

Work out the time complexity of the algorithm you are using now and it should be obvious that it can not be improved.

#include <stdio.h>
#include <malloc.h>

tydef struct node
{
    int info;
    struct node *link;
} *start;

void main()
{
    rev();
}

void rev()
{
    struct node *p = start, *q = NULL, *r;
    while (p != NULL)
    {
        r = q;
        q = p;
        p = p->link;
        q->link = r;
    }

    start = q;
}

I am using java to implement this and approach is test driven development hence test cases are also attached.

The Node class that represent single node -

package com.adnan.linkedlist;

/**
 * User  : Adnan
 * Email : [email protected]
 * Date  : 9/21/13
 * Time  : 12:02 PM
 */
public class Node {

    public Node(int value, Node node){
        this.value = value;
        this.node = node;
    }
    private int value;
    private Node node;

    public int getValue() {
        return value;
    }

    public Node getNode() {
        return node;
    }

    public void setNode(Node node){
        this.node = node;
    }
}

Service class that takes start node as input and reserve it without using extra space.

package com.adnan.linkedlist;

/**
 * User  : Adnan
 * Email : [email protected]
 * Date  : 9/21/13
 * Time  : 11:54 AM
 */
public class SinglyLinkedListReversal {

    private static final SinglyLinkedListReversal service 
= new SinglyLinkedListReversal();
    public static SinglyLinkedListReversal getService(){
        return service;
    }



    public Node reverse(Node start){
        if (hasOnlyNodeInLinkedList(start)){
            return start;
        }
        Node firstNode, secondNode, thirdNode;
        firstNode = start;
        secondNode = firstNode.getNode();
        while (secondNode != null ){
            thirdNode = secondNode.getNode();
            secondNode.setNode(firstNode);
            firstNode = secondNode;
            secondNode = thirdNode;
        }
        start.setNode(null);
        return firstNode;
    }

    private boolean hasOnlyNodeInLinkedList(Node start) {
        return start.getNode() == null;
    }


}

And The test case that covers above scenario. Please note that you require junit jars. I am using testng.jar; you can use any whatever pleases you..

package com.adnan.linkedlist;

import org.testng.annotations.Test;

import static org.testng.AssertJUnit.assertTrue;

/**
 * User  : Adnan
 * Email : [email protected]
 * Date  : 9/21/13
 * Time  : 12:11 PM
 */
public class SinglyLinkedListReversalTest {

    private SinglyLinkedListReversal reversalService = 
SinglyLinkedListReversal.getService();

    @Test
    public void test_reverseSingleElement() throws Exception {
        Node node = new Node(1, null);
        reversalService.reverse(node);
        assertTrue(node.getNode() == null);
        assertTrue(node.getValue() == 1);
    }


    //original - Node1(1) -> Node2(2) -> Node3(3)
    //reverse - Node3(3) -> Node2(2) -> Node1(1)
    @Test
    public void test_reverseThreeElement() throws Exception {
        Node node3 = new Node(3, null);
        Node node2 = new Node(2, node3);
        Node start = new Node(1, node2);


        start = reversalService.reverse(start);
        Node test = start;
        for (int i = 3; i >=1 ; i -- ){
          assertTrue(test.getValue() == i);
            test = test.getNode();
        }


    }

    @Test
    public void test_reverseFourElement() throws Exception {
        Node node4 = new Node(4, null);
        Node node3 = new Node(3, node4);
        Node node2 = new Node(2, node3);
        Node start = new Node(1, node2);


        start = reversalService.reverse(start);
        Node test = start;
        for (int i = 4; i >=1 ; i -- ){
            assertTrue(test.getValue() == i);
            test = test.getNode();
        }
    }

        @Test
        public void test_reverse10Element() throws Exception {
            Node node10 = new Node(10, null);
            Node node9 = new Node(9, node10);
            Node node8 = new Node(8, node9);
            Node node7 = new Node(7, node8);
            Node node6 = new Node(6, node7);
            Node node5 = new Node(5, node6);
            Node node4 = new Node(4, node5);
            Node node3 = new Node(3, node4);
            Node node2 = new Node(2, node3);
            Node start = new Node(1, node2);


            start = reversalService.reverse(start);
            Node test = start;
            for (int i = 10; i >=1 ; i -- ){
                assertTrue(test.getValue() == i);
                test = test.getNode();
            }


    }

    @Test
    public void test_reverseTwoElement() throws Exception {
        Node node2 = new Node(2, null);
        Node start = new Node(1, node2);


        start = reversalService.reverse(start);
        Node test = start;
        for (int i = 2; i >=1 ; i -- ){
            assertTrue(test.getValue() == i);
            test = test.getNode();
        }


    }
}