http filter example

examples/networking/http_filter/README.md

+#Simple HTTP Filter: Project purpose
+
+Write an eBPF application that parses HTTP packets and extracts (and prints on screen) the URL contained in the GET/POST request.
+
+[https://github.com/netgroup-polito/ebpf-test](https://github.com/netgroup-polito/ebpf-test)
+
+#Usage Example
+
+```Shell
+$ sudo python http-parse-v2.py 
+GET /pipermail/iovisor-dev/ HTTP/1.1
+HTTP/1.1 200 OK
+GET /favicon.ico HTTP/1.1
+HTTP/1.1 404 Not Found
+GET /pipermail/iovisor-dev/2016-January/thread.html HTTP/1.1
+HTTP/1.1 200 OK
+GET /pipermail/iovisor-dev/2016-January/000046.html HTTP/1.1
+HTTP/1.1 200 OK
+```
+
+#Implementation using BCC
+
+eBPF socket filter.<br />
+Filters IP and TCP packets, containing "HTTP", "GET", "POST" in payload and all subsequent packets belonging to the same session, having the same (ip_src,ip_dst,port_src,port_dst).<br />
+Program is loaded as PROG_TYPE_SOCKET_FILTER and attached to a socket, bind to eth0. <br />
+Matching packets are forwarded to user space, others dropped by the filter.<br />
+<br />
+Python script reads filtered raw packets from the socket, if necessary reassembles packets belonging to the same session, and prints on stdout the first line of the HTTP GET/POST request. <br />
+
+# Usage
+
+Require:
+- BPF Compiler Collection [BCC](https://github.com/iovisor/bcc)
+- Follow [INSTALL](https://github.com/iovisor/bcc/blob/master/INSTALL.md) guide
+
+# To run:
+
+```Shell
+$ sudo python http-parse.py
+$ sudo python http-parse-v2.py
+```
\ No newline at end of file

examples/networking/http_filter/http-parse-v2.c

+#include <uapi/linux/ptrace.h>
+#include <net/sock.h>
+#include <bcc/proto.h>
+
+#define IP_TCP 	6   
+#define ETH_HLEN 14
+
+struct Key {
+	u32 src_ip;               //source ip
+	u32 dst_ip;               //destination ip
+	unsigned short src_port;  //source port
+	unsigned short dst_port;  //destination port	
+};
+
+struct Leaf {
+	int timestamp;            //timestamp in ns
+};
+
+//BPF_TABLE(map_type, key_type, leaf_type, table_name, num_entry) 
+//map <Key, Leaf>
+//tracing sessions having same Key(dst_ip, src_ip, dst_port,src_port)
+BPF_TABLE("hash", struct Key, struct Leaf, sessions, 1024);
+
+/*eBPF program.
+  Filter IP and TCP packets, having payload not empty
+  and containing "HTTP", "GET", "POST"  as first bytes of payload.
+  AND ALL the other packets having same (src_ip,dst_ip,src_port,dst_port)
+  this means belonging to the same "session"
+  this additional check avoids url truncation, if url is too long
+  userspace script, if necessary, reassembles urls splitted in 2 or more packets.
+  if the program is loaded as PROG_TYPE_SOCKET_FILTER
+  and attached to a socket
+  return  0 -> DROP the packet
+  return -1 -> KEEP the packet and return it to user space (userspace can read it from the socket_fd )
+*/
+int http_filter(struct __sk_buff *skb) {
+
+	u8 *cursor = 0;
+
+	struct ethernet_t *ethernet = cursor_advance(cursor, sizeof(*ethernet));
+	//filter IP packets (ethernet type = 0x0800)
+	if (!(ethernet->type == 0x0800)){
+		goto DROP;	
+	}
+
+	struct ip_t *ip = cursor_advance(cursor, sizeof(*ip));
+	//filter TCP packets (ip next protocol = 0x06)
+	if (ip->nextp != IP_TCP) {
+		goto DROP;
+	}
+
+	u32  tcp_header_length = 0;
+	u32  ip_header_length = 0;
+	u32  payload_offset = 0;
+	u32  payload_length = 0;
+	struct Key 	key;
+	struct Leaf leaf;
+
+	struct tcp_t *tcp = cursor_advance(cursor, sizeof(*tcp));
+
+	//retrieve ip src/dest and port src/dest of current packet
+	//and save it into struct Key
+	key.dst_ip = ip->dst;
+	key.src_ip = ip->src;
+	key.dst_port = tcp->dst_port;
+	key.src_port = tcp->src_port;
+
+	//calculate ip header length
+	//value to multiply * 4
+	//e.g. ip->hlen = 5 ; IP Header Length = 5 x 4 byte = 20 byte
+	ip_header_length = ip->hlen << 2;    //SHL 2 -> *4 multiply
+		
+	//calculate tcp header length
+	//value to multiply *4
+	//e.g. tcp->offset = 5 ; TCP Header Length = 5 x 4 byte = 20 byte
+	tcp_header_length = tcp->offset << 2; //SHL 2 -> *4 multiply
+
+	//calculate patload offset and lenght
+	payload_offset = ETH_HLEN + ip_header_length + tcp_header_length; 
+	payload_length = ip->tlen - ip_header_length - tcp_header_length;
+		  
+	//http://stackoverflow.com/questions/25047905/http-request-minimum-size-in-bytes
+	//minimum lenght of http request is always geater than 7 bytes
+	//avoid invalid access memory
+	//include empty payload
+	if(payload_length < 7){
+		goto DROP;
+	}
+
+	//load firt 7 byte of payload into payload_array
+	//direct access to skb not allowed
+	unsigned long payload_array[7];
+	int i = 0;
+	int j = 0;
+	for (i = payload_offset ; i < (payload_offset + 7) ; i++){
+		payload_array[j] = load_byte(skb , i);
+		j++;
+	}
+
+	//find a match with an HTTP message
+	//HTTP
+	if ( (payload_array[0] == 'H') && (payload_array[1] == 'T') && (payload_array[2] == 'T') && (payload_array[3] == 'P')){
+		goto HTTP_MATCH;
+	}
+	//GET
+	if ( (payload_array[0] == 'G') && (payload_array[1] == 'E') && (payload_array[2] == 'T') ){
+		goto HTTP_MATCH;
+	}
+	//POST
+	if ( (payload_array[0] == 'P') && (payload_array[1] == 'O') && (payload_array[2] == 'S') && (payload_array[3] == 'T')){
+		goto HTTP_MATCH;
+	}
+	//PUT
+	if ( (payload_array[0] == 'P') && (payload_array[1] == 'U') && (payload_array[2] == 'T') ){
+		goto HTTP_MATCH;
+	}
+	//DELETE
+	if ( (payload_array[0] == 'D') && (payload_array[1] == 'E') && (payload_array[2] == 'L') && (payload_array[3] == 'E') && (payload_array[4] == 'T') && (payload_array[5] == 'E')){
+		goto HTTP_MATCH;
+	}
+	//HEAD
+	if ( (payload_array[0] == 'H') && (payload_array[1] == 'E') && (payload_array[2] == 'A') && (payload_array[3] == 'D')){
+		goto HTTP_MATCH;
+	}
+
+	//no HTTP match
+	//check if packet belong to an HTTP session
+	struct Leaf * lookup_leaf = sessions.lookup(&key);
+	if(lookup_leaf){
+		//send packet to userspace
+		goto KEEP;
+	}
+	goto DROP;
+
+	//keep the packet and send it to userspace retruning -1
+	HTTP_MATCH:
+	//if not already present, insert into map <Key, Leaf>
+	leaf.timestamp = 0;
+	sessions.lookup_or_init(&key, &leaf);
+	sessions.update(&key,&leaf);
+	
+	//send packet to userspace returning -1
+	KEEP:
+	return -1;
+
+	//drop the packet returning 0
+	DROP:
+	return 0;
+
+}

examples/networking/http_filter/http-parse.c

+#include <uapi/linux/ptrace.h>
+#include <net/sock.h>
+#include <bcc/proto.h>
+
+#define IP_TCP 	6   
+#define ETH_HLEN 14
+
+/*eBPF program.
+  Filter IP and TCP packets, having payload not empty
+  and containing "HTTP", "GET", "POST" ... as first bytes of payload
+  if the program is loaded as PROG_TYPE_SOCKET_FILTER
+  and attached to a socket
+  return  0 -> DROP the packet
+  return -1 -> KEEP the packet and return it to user space (userspace can read it from the socket_fd )
+*/
+int http_filter(struct __sk_buff *skb) {
+
+	u8 *cursor = 0;
+
+	struct ethernet_t *ethernet = cursor_advance(cursor, sizeof(*ethernet));
+	//filter IP packets (ethernet type = 0x0800)
+	if (!(ethernet->type == 0x0800)){
+		goto DROP;	
+	}
+
+	struct ip_t *ip = cursor_advance(cursor, sizeof(*ip));
+	//filter TCP packets (ip next protocol = 0x06)
+	if (ip->nextp != IP_TCP) {
+		goto DROP;
+	}
+
+	u32  tcp_header_length = 0;
+	u32  ip_header_length = 0;
+	u32  payload_offset = 0;
+	u32  payload_length = 0;
+
+	struct tcp_t *tcp = cursor_advance(cursor, sizeof(*tcp));
+
+	//calculate ip header length
+	//value to multiply * 4
+	//e.g. ip->hlen = 5 ; IP Header Length = 5 x 4 byte = 20 byte
+	ip_header_length = ip->hlen << 2;    //SHL 2 -> *4 multiply
+		
+	//calculate tcp header length
+	//value to multiply *4
+	//e.g. tcp->offset = 5 ; TCP Header Length = 5 x 4 byte = 20 byte
+	tcp_header_length = tcp->offset << 2; //SHL 2 -> *4 multiply
+
+	//calculate patload offset and lenght
+	payload_offset = ETH_HLEN + ip_header_length + tcp_header_length; 
+	payload_length = ip->tlen - ip_header_length - tcp_header_length;
+		  
+	//http://stackoverflow.com/questions/25047905/http-request-minimum-size-in-bytes
+	//minimum lenght of http request is always geater than 7 bytes
+	//avoid invalid access memory
+	//include empty payload
+	if(payload_length < 7){
+		goto DROP;
+	}
+
+	//load firt 7 byte of payload into payload_array
+	//direct access to skb not allowed
+	unsigned long payload_array[7];
+	int i = 0;
+	int j = 0;
+	for (i = payload_offset ; i < (payload_offset + 7) ; i++){
+		payload_array[j] = load_byte(skb , i);
+		j++;
+	}
+
+	//find a match with an HTTP message
+	//HTTP
+	if ( (payload_array[0] == 'H') && (payload_array[1] == 'T') && (payload_array[2] == 'T') && (payload_array[3] == 'P')){
+		goto KEEP;
+	}
+	//GET
+	if ( (payload_array[0] == 'G') && (payload_array[1] == 'E') && (payload_array[2] == 'T') ){
+		goto KEEP;
+	}
+	//POST
+	if ( (payload_array[0] == 'P') && (payload_array[1] == 'O') && (payload_array[2] == 'S') && (payload_array[3] == 'T')){
+		goto KEEP;
+	}
+	//PUT
+	if ( (payload_array[0] == 'P') && (payload_array[1] == 'U') && (payload_array[2] == 'T') ){
+		goto KEEP;
+	}
+	//DELETE
+	if ( (payload_array[0] == 'D') && (payload_array[1] == 'E') && (payload_array[2] == 'L') && (payload_array[3] == 'E') && (payload_array[4] == 'T') && (payload_array[5] == 'E')){
+		goto KEEP;
+	}
+	//HEAD
+	if ( (payload_array[0] == 'H') && (payload_array[1] == 'E') && (payload_array[2] == 'A') && (payload_array[3] == 'D')){
+		goto KEEP;
+	}
+
+	//no HTTP match
+	goto DROP;
+
+	//keep the packet and send it to userspace retruning -1
+	KEEP:
+	return -1;
+
+	//drop the packet returning 0
+	DROP:
+	return 0;
+
+}

examples/networking/http_filter/http-parse.py

+#!/usr/bin/python
+#
+#Bertrone Matteo - Polytechnic of Turin
+#November 2015
+#
+#eBPF application that parses HTTP packets 
+#and extracts (and prints on screen) the URL contained in the GET/POST request.
+#
+#eBPF program http_filter is used as SOCKET_FILTER attached to eth0 interface.
+#only packet of type ip and tcp containing HTTP GET/POST are returned to userspace, others dropped
+#
+#python script uses bcc BPF Compiler Collection by iovisor (https://github.com/iovisor/bcc)
+#and prints on stdout the first line of the HTTP GET/POST request containing the url
+
+from __future__ import print_function
+from bcc import BPF
+
+import sys
+import socket
+import os
+
+# initialize BPF - load source code from http-parse.c
+bpf = BPF(src_file = "http-parse.c",debug = 0)
+
+#load eBPF program http_filter of type SOCKET_FILTER into the kernel eBPF vm
+#more info about eBPF program types
+#http://man7.org/linux/man-pages/man2/bpf.2.html
+function_http_filter = bpf.load_func("http_filter", BPF.SOCKET_FILTER)
+
+#create raw socket, bind it to eth0
+#attach bpf program to socket created
+BPF.attach_raw_socket(function_http_filter, "eth0")
+
+#get file descriptor of the socket previously created inside BPF.attach_raw_socket
+socket_fd = function_http_filter.sock
+
+#create python socket object, from the file descriptor
+sock = socket.fromfd(socket_fd,socket.PF_PACKET,socket.SOCK_RAW,socket.IPPROTO_IP)
+#set it as blocking socket
+sock.setblocking(True)
+
+while 1:
+  #retrieve raw packet from socket
+  packet_str = os.read(socket_fd,2048)
+
+  #DEBUG - print raw packet in hex format
+  #packet_hex = toHex(packet_str)
+  #print ("%s" % packet_hex)
+
+  #convert packet into bytearray
+  packet_bytearray = bytearray(packet_str)
+  
+  #ethernet header length
+  ETH_HLEN = 14 
+
+  #IP HEADER
+  #https://tools.ietf.org/html/rfc791
+  # 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+  # +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+  # |Version|  IHL  |Type of Service|          Total Length         |
+  # +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+  #
+  #IHL : Internet Header Length is the length of the internet header 
+  #value to multiply * 4 byte
+  #e.g. IHL = 5 ; IP Header Length = 5 * 4 byte = 20 byte
+  #
+  #Total Lenght: This 16-bit field defines the entire packet size, 
+  #including header and data, in bytes.
+
+  #calculate packet total lenght
+  total_lenght = packet_bytearray[ETH_HLEN + 2]               #load MSB
+  total_lenght = total_lenght << 8                            #shift MSB
+  total_lenght = total_lenght + packet_bytearray[ETH_HLEN+3]  #add LSB
+  
+  #calculate ip header lenght
+  ip_header_length = packet_bytearray[ETH_HLEN]               #load Byte
+  ip_header_length = ip_header_length & 0x0F                  #mask bits 0..3
+  ip_header_length = ip_header_length << 2                    #shift to obtain lenght
+
+  #TCP HEADER 
+  #https://www.rfc-editor.org/rfc/rfc793.txt
+  #  12              13              14              15  
+  #  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 
+  # +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+  # |  Data |           |U|A|P|R|S|F|                               |
+  # | Offset| Reserved  |R|C|S|S|Y|I|            Window             |
+  # |       |           |G|K|H|T|N|N|                               |
+  # +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+  #
+  #Data Offset: This indicates where the data begins.  
+  #The TCP header is an integral number of 32 bits long.
+  #value to multiply * 4 byte
+  #e.g. DataOffset = 5 ; TCP Header Length = 5 * 4 byte = 20 byte
+
+  #calculate tcp header lenght
+  tcp_header_lenght = packet_bytearray[ETH_HLEN + ip_header_length + 12]  #load Byte
+  tcp_header_lenght = tcp_header_lenght & 0xF0                            #mask bit 4..7
+  tcp_header_lenght = tcp_header_lenght >> 2                              #SHR 4 ; SHL 2 -> SHR 2
+  
+  #calculate payload offset
+  payload_offset = ETH_HLEN + ip_header_length + tcp_header_lenght
+  
+  #print first line of the HTTP GET/POST request
+  #line ends with 0xOD 0xOA (\r\n)
+  #(if we want to print all the header print until \r\n\r\n)
+  for i in range (payload_offset-1,len(packet_bytearray)-1):
+    if (packet_bytearray[i]== 0x0A):
+      if (packet_bytearray[i-1] == 0x0D):
+        break
+    print ("%c" % chr(packet_bytearray[i]), end = "")
+  print("")
+