IPV4 RAW AF INET TCP server client program with Poll system call

IPv4 RAW_AF_INET TCP

• In this program, you are going to learn

• How to create a Socket ?

• How to write a data ?

• How to read a data ?

• How to use socket APIs ?

  • socket

  • write

  • read

Topics in this section,

• IPV4 RAW_AF_INET TCP SOCKET FAQs

• Step 1: Sequence Diagram for SERVER.c

• Step 2: Program for Server.c

• Step 3: Compile and Execute Server.c

• Step 4: Sequence Diagram for CLIENT.c

• Step 5: Program for Client.c

• Step 6: Compile and Execute Client.c

• Summary

IPV4 RAW_AF_INET TCP SOCKET : FAQs

Let us answer few basic questions in this socket

What does socket(AF_INET, SOCK_RAW, IPPROTO_TCP) do?

See Answer

This call creates a raw socket in the AF_INET address family for direct access to TCP packets.

When is it appropriate to use SOCK_RAW sockets with TCP?

See Answer

It’s suitable for tasks like packet capturing, network monitoring, or implementing custom protocols where direct access to TCP packets is needed.

Can this socket be used for regular TCP communication?

See Answer

While technically possible, it’s not recommended for regular communication due to increased complexity and potential security risks.

How does a raw TCP socket differ from a regular TCP socket?

See Answer

A raw TCP socket provides direct access to the TCP layer, allowing for manual packet manipulation, whereas regular TCP sockets handle packet details internally.

What are some use cases for raw TCP sockets?

See Answer

Use cases include network sniffing, packet analysis, security auditing, and developing custom network protocols.

How can I capture and analyze TCP packets using raw sockets?

See Answer

You can use the raw socket to capture TCP packets and analyze them using packet analysis tools like Wireshark.

Is error checking needed after creating the socket?

See Answer

Yes, checking for errors ensures that the socket is created successfully before proceeding with further operations.

Why is it important to check the return value of read() and write() in socket programming?

See Answer

It detects issues such as network errors or closed connections.

What is the purpose of the poll system call?

See Answer

To block and wait for activity on one or more file descriptors.

How does poll differ from poll in terms of usability?

See Answer

poll is more efficient than poll for monitoring multiple file descriptors.

What types of file descriptors can be monitored using poll?

See Answer

sockets, files, timerfd, socketpair, message_queue, Namedpipes and shared_memory.

How does poll handle a set of file descriptors with different states (e.g., reading, writing, exception)?

See Answer

It uses different structures for each state in the pollfd array.

How do you handle errors when using the poll system call?

See Answer

Check the return value for -1 to detect errors, Use perror to print error messages.

How does poll handle a set of file descriptors with different states (e.g., reading, writing, exception)?

See Answer

Array of pollfd Structures:

Before calling poll, you need to create an array of pollfd structures, where each structure represents a file descriptor and its associated events.

struct pollfd fds[NUM_FDS];

NUM_FDS is the number of file descriptors you want to monitor.

Initialize pollfd Structures:

For each file descriptor you want to monitor, initialize the corresponding pollfd structure with the following information:

fd: The file descriptor to monitor. events: The events of interest (e.g., POLLIN for readability, POLLOUT for writability). revents: Initially set to zero. After the poll call, this field is updated to indicate the events that occurred.

fds[0].fd = fd1;

fds[0].events = POLLIN;

fds[0].revents = 0;

fds[1].fd = fd2;

fds[1].events = POLLIN;

fds[1].revents = 0;

Call poll:

After initializing the pollfd array, call the poll function, providing the array, the number of file descriptors, and a timeout

int ready_fds = poll(fds, NUM_FDS, timeout_ms);

ready_fds will contain the number of file descriptors that are ready.

How does poll Checking Ready File Descriptors?

See Answer

After the poll call, loop through the pollfd array and check the revents field for each file descriptor to determine which events occurred.

for (int i = 0; i < NUM_FDS; ++i) {
        if (fds[i].revents & POLLIN) {
                // File descriptor i is ready for reading
        }

        if (fds[i].revents & POLLOUT) {
                // File descriptor i is ready for writing
        }
        // Check other events if needed (e.g., POLLERR, POLLHUP)
}

What does it mean if poll returns 0?

See Answer

No file descriptors are ready within the specified timeout.

Step 1: Sequence Diagram for SERVER.c

Step 2 : program for server.c

• There are many functions used in socket. We can classify those functions based on functionalities.

  • Create Socket

  • Bind Socket

  • Connect Socket

  • Poll

  • Write data_packet

  • Read data_packet

  • Close socket

• socket() is used to create a new socket. For example,

sock_fd = socket(AF_INET, SOCK_RAW, IPPROTO_TCP);

• bind() is used to associate the socket with a specific address and port. For example,

ret = bind(sock_fd, (struct sockaddr*)servaddr, sizeof(struct sockaddr_in));

• connect() is used in network programming to establish a connection from a client to a server. For example,

ret = connect(sock_fd, (struct sockaddr*)client_addr, sizeof(struct sockaddr_in));

• poll() is used for monitoring multiple file descriptors to see if I/O is possible on any of them.

ret = poll(fds, 1, 1000);

• read system call in C is commonly used to read data from a file descriptor, such as a socket.

ret = read(sock_fd, recvbuffer, sizeof(recvbuffer));

• write system call in C is used to write data to a file descriptor, such as a socket.

ret = write(sock_fd, buffer, sizeof(buffer));

• close is used to close the socket To free up system resources associated with the socket. For example,

(void)close(sock_fd);

• See the full program below,

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <signal.h>
#include <errno.h>
#include <linux/ip.h>
#include <netinet/tcp.h>
#include <poll.h>

#define PORT 50000
#define PORT_CLIENT 50001

struct sockaddr_in 
*servaddr = NULL,
*client_addr = NULL;
int sock_fd;

struct pseudo_iphdr {
    unsigned int source_ip_addr;
    unsigned int dest_ip_addr;
    unsigned char fixed;
    unsigned char protocol;
    unsigned short tcp_len;
};

unsigned short in_cksum (
uint16_t * addr, int len)
{
   int nleft = len;
   unsigned int sum = 0;
   unsigned short *w = addr;
   unsigned short answer = 0;

   while (nleft > 1) {
       sum += *w++;
       nleft -= 2;
   }

   if (nleft == 1) {
     *(unsigned char *) 
     (&answer) = 
    * (unsigned char *) w;
         sum += answer;
    }

    sum = (sum >> 16) + 
    (sum & 0xffff);
    sum += (sum >> 16); 
    answer = (unsigned short) ~sum;
    return (answer);
}

void interrupt_handler (
int signum) {
     (void)close(sock_fd);
     free(client_addr);
     exit(0);
}

void validate_convert_addr(
char *ip_str,
struct sockaddr_in *sock_addr)
{
  if (ip_str == NULL) {
   perror("Invalid ip_str\n");
   exit(EXIT_FAILURE);
 }

 if (sock_addr == NULL) {
   perror("Invalid sock_addr\n");
   exit(EXIT_FAILURE);
 }

 printf("IP Address: %s\n", ip_str);

 if (inet_pton(AF_INET, ip_str, 
 &(sock_addr->sin_addr)) <= 0) {
    perror("Invalid address\n");
    exit(EXIT_FAILURE);
  }
}

int main (int argc, char *argv[])
{
  char buffer[1024] = 
  {0};
  unsigned char recvbuffer[1024] = 
  {0};
  int length, ret;
  char *string = 
  "Hello client";
  struct tcphdr *tcp_hdr = NULL;
  char *string_data = NULL;
  char *recv_string_data = NULL;
  char *csum_buffer = NULL;
  struct pseudo_iphdr 
  csum_hdr;
  struct pollfd fds[1];

  if (argc != 2) {
     printf("%s<ip-addr>\n",
     argv[0]);
     exit(EXIT_FAILURE);
  }

  signal (SIGINT, 
  interrupt_handler);
  signal (SIGTERM, 
  interrupt_handler);

  sock_fd = socket(AF_INET, 
            SOCK_RAW, 
            IPPROTO_TCP);

  if(0 > sock_fd) {
     printf("unable to create\n");
     exit(0);
  }

  servaddr = (struct sockaddr_in *)malloc(
  sizeof(struct sockaddr_in));
        
  if (servaddr == NULL) {
     printf("could not allocate memory\n");
     goto end;
  }

  servaddr->sin_family = AF_INET;
  servaddr->sin_port = PORT;
  validate_convert_addr(argv[1], 
  servaddr);

  ret = bind(sock_fd, 
  (struct sockaddr *)servaddr, 
  sizeof(struct sockaddr_in));

  if (ret < 0) {
    printf("bind\n");
    goto end1;
  }

  client_addr = (struct sockaddr_in *)malloc(
  sizeof(struct sockaddr_in));
  
  if (client_addr == NULL) {
    printf("allocation memory\n");
    goto end2;
  }

  client_addr->sin_family = AF_INET;
  client_addr->sin_port =
  PORT_CLIENT;
  validate_convert_addr(argv[1],
  client_addr);

  ret = connect(sock_fd, 
  (struct sockaddr *)client_addr, 
  sizeof(struct sockaddr_in));
        
  if (ret != 0) {
     printf("error %d", errno);
     printf("connect returned error\n");
     goto end2;
  }
      
  string_data = (char *) 
  (buffer + sizeof(struct tcphdr));
  strncpy(string_data, string, 
  strlen(string));

  tcp_hdr = (struct tcphdr *)buffer;
  tcp_hdr->source = htons(PORT);
  tcp_hdr->dest = 
  htons(PORT_CLIENT);
  tcp_hdr->ack_seq = 0x0; 
  tcp_hdr->doff = 5; 
  tcp_hdr->syn = 1; 
  tcp_hdr->window = htons(200); 

  csum_buffer = (char *)calloc((
  sizeof(struct pseudo_iphdr) + 
  sizeof(struct tcphdr) + 
  strlen(string_data)), 
  sizeof(char));

  if (csum_buffer == NULL) {
     printf("allocate csum buffer\n");
     goto end1;
  }

  csum_hdr.source_ip_addr = 
  inet_addr("192.168.1.11");
  csum_hdr.dest_ip_addr = 
  inet_addr("192.168.1.14");
  csum_hdr.fixed = 0;
  csum_hdr.protocol = 
  IPPROTO_TCP;
  csum_hdr.tcp_len = 
  htons(sizeof(struct tcphdr) + 
  strlen(string_data) + 1);

  memcpy(csum_buffer, (char *)&csum_hdr, 
  sizeof(struct pseudo_iphdr));
  memcpy(csum_buffer + 
  sizeof(struct pseudo_iphdr), buffer, 
  (sizeof(struct tcphdr) + 
  strlen(string_data) + 1));

  tcp_hdr->check = (in_cksum(
  (unsigned short *) csum_buffer,
  (sizeof(struct pseudo_iphdr)+ 
  sizeof(struct tcphdr) + strlen(string_data) + 1)));

  printf("checksum is %x", tcp_hdr->check);
  free (csum_buffer);

  memset(fds, 0, sizeof(fds));
  fds[0].fd = sock_fd;
  fds[0].events = POLLIN;

  while (1) {
   ret = poll(fds, 1, 1000);           

   if (ret == -1) {
       perror("poll\n");
       break;
   }

   if (fds[0].revents & POLLIN) {
      memset(recvbuffer, 0, 
      sizeof(recvbuffer));

      ret = read(sock_fd, recvbuffer, 
      sizeof(recvbuffer));

      if (ret == -1) {
        perror("read error");
        break;
      }

      tcp_hdr = (struct tcphdr *)
      (recvbuffer + 
      sizeof (struct iphdr));
      recv_string_data = (char *) 
      (recvbuffer + 
      sizeof (struct iphdr) + 
      sizeof (struct tcphdr));
      
      if (PORT == ntohs(
	tcp_hdr->dest)) {
        printf("Received : %s\n", 
	recv_string_data);
      }

      ret = write(sock_fd, buffer, 
      sizeof(buffer));
      
      if (ret == -1) {
        perror("write error");
        break;
      }
    }
   }

end2:
     free(client_addr);
end1:
     free(servaddr);
end:
     (void)close(sock_fd);

     return 0;
}

Step 3: Compile and Execute server.c

$ gcc -o server server.c

$ sudo ./server 127.0.0.1

IP Address: 127.0.0.1
IP Address: 127.0.0.1
checksum is c945
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
Received : Hello server
^C

Step 4 : Sequence Diagram for CLIENT.c

Step 5: program for client.c

• There are many functions used in socket. We can classify those functions based on functionalities.

  • Create Socket

  • Bind Socket

  • Connect Socket

  • Poll

  • Write data_packet

  • Read data_packet

  • Close socket

• socket is used to create a new socket. For example,

sockfd = socket(AF_INET, SOCK_RAW, IPPROTO_TCP);

• bind() is used to associate the socket with a specific address and port. For example,

ret = bind(sockfd, (struct sockaddr*)clientaddr, sizeof(struct sockaddr_in));

• connect() is used in network programming to establish a connection from a client to a server. For example,

ret = connect(sockfd, (struct sockaddr*)serveraddr, sizeof(struct sockaddr_in));

• poll() is used for monitoring multiple file descriptors to see if I/O is possible on any of them.

ret = poll(fds, 2, -1);

• read system call in C is commonly used to read data from a file descriptor, such as a socket.

ret = read(sockfd, recvbuffer, sizeof(recvbuffer));

• write system call in C is used to write data to a file descriptor, such as a socket.

ret = write(sockfd, buffer, sizeof(buffer));

• close is used to close the socket To free up system resources associated with the socket. For example,

(void)close(sockfd);

• See the full program below,

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <signal.h>
#include <errno.h>
#include <netinet/tcp.h>
#include <linux/ip.h>
#include <poll.h>

#define PORT 50001
#define SERVER_PORT 50000

struct sockaddr_in 
*serveraddr = NULL, 
*clientaddr;
int sockfd;

struct pseudo_iphdr {
   unsigned int source_ip_addr;
   unsigned int dest_ip_addr;
   unsigned char fixed;
   unsigned char protocol;
   unsigned short tcp_len;
};

unsigned short in_cksum (
uint16_t * addr, int len)
{
   int nleft = len;
   unsigned int sum = 0;
   unsigned short *w = addr;
   unsigned short answer = 0;

   while (nleft > 1) {
       sum += *w++;
       nleft -= 2;
   }

   if (nleft == 1) {
      *(unsigned char *) (&answer) = 
      * (unsigned char *) w;
      sum += answer;
   }

   sum = (sum >> 16) + 
   (sum & 0xffff); 
   sum += (sum >> 16); 
   answer = (unsigned short) ~sum;
   return (answer);
}

void interrupt_handler (
int signum) 
{
  close(sockfd);
  free(clientaddr);
  exit(0);
}

void validate_convert_addr(
char *ip_str,
struct sockaddr_in *sock_addr)
{
  if (ip_str == NULL) {
   perror("Invalid ip_str\n");
   exit(EXIT_FAILURE);
 }

 if (sock_addr == NULL) {
   perror("Invalid sock_addr\n");
   exit(EXIT_FAILURE);
 }

 printf("IP Address: %s\n", ip_str);

 if (inet_pton(AF_INET, ip_str,
 &(sock_addr->sin_addr)) <= 0) {
    perror("Invalid address\n");
    exit(EXIT_FAILURE);
  }
}

int main (int argc, char *argv[])
{
   char buffer[1024] = 
   {0};
   unsigned char recvbuffer[1024] = 
   {0};
   int length, ret;
   char *string = 
   "Hello server";
   struct tcphdr *tcp_hdr = NULL;
   char *string_data = NULL;
   char *recv_string_data = NULL;
   char *csum_buffer = NULL;
   struct pseudo_iphdr csum_hdr;
   int error;
   struct pollfd fds[2];

   if (argc != 2) {
     printf("%s<ip-addr>\n",
     argv[0]);
     exit(EXIT_FAILURE);
   }

   signal (SIGINT, 
   interrupt_handler);
   signal (SIGTERM, 
   interrupt_handler);

   sockfd = socket(AF_INET, 
            SOCK_RAW, 
            IPPROTO_TCP);
      
   if(0 > sockfd) {
     printf("create socket\n");
     exit(0);
   }

   clientaddr = (struct sockaddr_in *)malloc(
   sizeof(struct sockaddr_in));
      
   if (clientaddr == NULL) {
      printf("allocate memory\n");
      goto end;
   }

   clientaddr->sin_family = AF_INET;
   clientaddr->sin_port = PORT;
   validate_convert_addr(argv[1],
   clientaddr);

   ret = bind(sockfd, 
   (struct sockaddr *)clientaddr, 
   sizeof(struct sockaddr_in));

   if (ret < 0) {
      printf(" bind\n");
      goto end1;
   }

   serveraddr = (struct sockaddr_in *)malloc(
   sizeof(struct sockaddr_in));
      
   if (serveraddr == NULL) {
      printf("allocate memory\n");
      goto end2;
   }
   
   serveraddr->sin_family = AF_INET;
   serveraddr->sin_port = SERVER_PORT;
   validate_convert_addr(argv[1],
   serveraddr);

   ret = connect(sockfd, 
   (struct sockaddr *)serveraddr, 
   sizeof(struct sockaddr_in));
      
   if (ret != 0) {
      printf("error %d", errno);
      printf("connect returned error\n");
      goto end2;
   }

   string_data = (char *) 
   (buffer + sizeof(struct tcphdr));
   strncpy(string_data, string, 
   strlen(string));

   tcp_hdr = (struct tcphdr *)buffer;
   tcp_hdr->source = htons(PORT);
   tcp_hdr->dest = htons(SERVER_PORT);
   tcp_hdr->ack_seq = 0x0;
   tcp_hdr->doff = 5; 
   tcp_hdr->syn = 1;
   tcp_hdr->window = htons(200);

   csum_buffer = (char *)calloc(
   (sizeof(struct pseudo_iphdr) + 
   sizeof(struct tcphdr) + 
   strlen(string_data)), sizeof(char));
      
   if (csum_buffer == NULL) {
      printf("allocate csum buffer\n");
      goto end1;
   }

   csum_hdr.source_ip_addr = 
   inet_addr("192.168.1.14");
   csum_hdr.dest_ip_addr = 
   inet_addr("192.168.1.11");
   csum_hdr.fixed = 0;
   csum_hdr.protocol = 
   IPPROTO_TCP;
   csum_hdr.tcp_len = 
   htons(sizeof(struct tcphdr) + 
   strlen(string_data) + 1);

   memcpy(csum_buffer, 
   (char *)&csum_hdr, 
   sizeof(struct pseudo_iphdr));
   memcpy(csum_buffer + 
   sizeof(struct pseudo_iphdr), 
   buffer, (sizeof(struct tcphdr) + 
   strlen(string_data) + 1));

   tcp_hdr->check = (in_cksum(
   (unsigned short *) csum_buffer,
   (sizeof(struct pseudo_iphdr)+ 
   sizeof(struct tcphdr) + 
   strlen(string_data) + 1)));

   printf("checksum is %x", 
   tcp_hdr->check);
      
   free (csum_buffer);

   memset(fds, 0, sizeof(fds));

   fds[0].fd = sockfd;
   fds[0].events = POLLIN; 

   while (1) {
     ret = write(sockfd, 
     buffer, sizeof(buffer));
           
     if (ret < 0) {
        perror("read error");
        break;
     }
          
     ret = poll(fds, 2, -1);

     if (ret < 0) {
        perror("poll\n");
        break;
     }

     if (fds[0].revents & POLLIN) {
       memset(recvbuffer, 0, 
       sizeof(recvbuffer));
                
       ret = read(sockfd, 
       recvbuffer, sizeof(recvbuffer));
                
       if (ret < 0) {
          perror("read error");
          break;
       }

       tcp_hdr = (struct tcphdr *)
       (recvbuffer + sizeof (struct iphdr));
       recv_string_data = (char *) 
       (recvbuffer + sizeof (struct iphdr) + 
       sizeof (struct tcphdr));
                
       if (PORT == ntohs(tcp_hdr->dest)) {
           printf("Received : %s\n", 
           recv_string_data);
       }
     }
   }

end2:
      free(serveraddr);
end1:
      free(clientaddr);
end:
      (void)close(sockfd);

      return 0;
}

Step 6: Compile and Execute client.c

$ gcc -o client client.c

$ sudo ./client 127.0.0.1

IP Address: 127.0.0.1
IP Address: 127.0.0.1
checksum is c135
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
Received : Hello client
^C

Enhanced Socket Flexibility with AF_INET and PF_INET Domains

Default Domain:

By default, the socket is configured to work in the AF_INET domain, handling all types of network data.

Additional Domain Support:

We expand the socket’s capabilities to also function in the PF_INET domain, allowing it to operate similarly to AF_INET.

Socket Creation:

We set up a network connection point known as a socket using socket(PF_INET, SOCK_RAW, IPPROTO_TCP).

Working Scenario:

Despite the change in domain to PF_INET, the socket continues to operate the same way, handling general network data.

Summary

Socket API	Learning
socket	Create a new socket
poll	Monitor multiple file descriptors (usually sockets) for read, write, or error conditions.
write	used to write data to a file descriptor, such as a socket.
read	used to read data from a file descriptor, such as a socket.

See Also

• previous topic

  • IPV4 RAW AF INET TCP server client program with Select system call

• current topic

  • IPV4 RAW AF INET TCP server client program with Poll system call

• Next topic

  • IPV4 RAW AF INET TCP server client program with Epoll system call

• Other sockets

  • ipv4_raw_af_inet_udp

  • ipv6_af_inet6_tcp

  • ipv6_af_inet6_udp

  • //TODO ipv6_raw_af_inet6_icmp

  • //TODO ipv6_raw_af_inet6_tcp

  • //TODO ipv6_raw_af_inet6_udp

  • // TODO ipv6_raw_af_inet6_raw

  • raw_af_packet_raw_htons_ETH_P_ALL

  • raw_af_packet_tcp_htons_ETH_P_ALL

  • raw_af_packet_udp_htons_ETH_P_ALL

• Other IPCs

  • Message_queues

  • NamedPipes

  • Netlink

  • Shared_Memory

  • Shared_Memory_2_FDS

  • SocketPair

  • Timerfd