Fork me on GitHub
简单版WAF代码学习

简单版WAF代码学习

 

网上代码只见了http://sourceforge.net/projects/sqlxsswaf/?source=directory

开始read!

一: 主函数

流程很清晰,

1.  整个WAF主函数体为死循环,在while(1)的代码段中,当代码处理完毕当前日志内容后,睡眠10ms,继续从get_pos处向后处理新内容。

2.  第二个while处理日志,当找到get或post为开头的日志内容后,对客户端发送来的命令进行检测,直到文档结尾。然后到达1中while循环的末尾

1
#define tailer "/var/log/apache2/access.log"
1
#define finder "GET"
1
#define finder2 "POST"
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
int main(void){
        fpos_t get_pos;
        printf("SQLXSSGRABBER HAS STARTED\n");
        while(1) {
        FILE *fp = fopen(tailer,"r");
        fsetpos(fp,&get_pos);
        if (fp != NULL){
                char max_line[LINE_MAX];
                while(fgets(max_line,sizeof(max_line),fp) != NULL)
                {
                        if (strstr(max_line,finder)||strstr(max_line,finder2)){
                            fgetpos(fp,&get_pos);
                            capture(max_line);         
                        }
                }
                fclose(fp);
            }
                else{
                   perror(tailer);
                    }
        sleep(10);
}
 
return 0;
}

 

 二: capture函数

  主要是一个中间层,把具体的正则式和匹配实现抽象成一个函数,供main使用。

  很新奇的一点是,全部case使用了FALL THROUTH,又从regex_roller开始控制,很方便,省去了许多修改功能的注释时间,不需要某个引擎又不想要删除的时候,直接放在regex_roller的前面即可。

1
#define att_1 "((\%3C)|<)[^\n]+((\%3E)|>)"<br>#define att_2 "((\%3C)|<)((\%2F)|\\/)*[a-z0-9\%]+((\%3E)|>)"<br>#define att_3 "((\%3C)|<)((\%69)|i|(\%49))((\%6D)|m|(\%4D))((\%67)|g|(\%47))[^\n]+((\%3E)|>)"<br>#define att_4 "((\%27)|('))union"<br>#define att_5 "((\%3D)|(=))[^\n]*((\%27)|(<a href="file://')%7c(//-//-)%7C(/%253B" target="_blank">\\')|(\\-\\-)|(\%3B</a>)|(;))"<br>#define att_6 "((\%27)|('))"
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
char *capture(char *log_line){
    int regex_roller =0;
     
    char *xss_para_regex = att_1;
    char *xss_simple_regex = att_2;
    char *css_img_regex = att_3;
    char *unionsql_regex = att_4;
    char *sqlmeta_regex = att_5;
    char *sqlmagicquote_regex = att_6;
     
    switch(regex_roller){
        //add as many more as you wish but dont forget to #define the regex above.
        case 0:
        cap_matcher(log_line,xss_para_regex,0);
        case 1:
        cap_matcher(log_line,xss_simple_regex,1);
        case 2:
        cap_matcher(log_line,css_img_regex,2);
        case 3:
        cap_matcher(log_line,unionsql_regex,3);
        case 4:
        cap_matcher(log_line,sqlmeta_regex,4);
        case 5:
        cap_matcher(log_line,sqlmagicquote_regex,5);
        default:
            break;
    }
 
    return 0;
}

 方便带来的后果就是如果检测出一段攻击代码后,会对这段代码继续进行其他可能性的检测,这是不需要的,同时底层函数功能过于复杂,把本该在中间层函数实现的功能带到了底层去。

三: 匹配引擎

  首先编译正则式,然后进行匹配,匹配成功,根据传入的规则编号,输出对应的攻击方式,阻塞IP,再利用邮件通知给管理员。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
char *cap_matcher(char *log_line, char *regex,int attack_type){
    pid_t pid;
    pcre *attack_regex;
    const char *error;
    int erroffset;
    int ovector[OVECCOUNT];
    int rc;
    attack_regex = pcre_compile(regex,0,&error,&erroffset,NULL);
    if (! attack_regex){
        fprintf(stderr,"PCRE compilation failed at expression offset %d: %s\n", erroffset, error);
        return (char *)1;
        }
    rc = pcre_exec(attack_regex,NULL,log_line,strlen(log_line),0,0,ovector,OVECCOUNT);
 
    if (rc < 3)
    {
        return (char *)1;
    }
    else{
        switch(attack_type){
            case 0:
                printf("Paranoid Xss Filter Detection\n");
                iptables_blockage(log_line);
                break;
            case 1:
                printf("Simple Xss Filter Detection\n");
                iptables_blockage(log_line);
                break;
            case 2:
                printf("Xss Img Filter Detection\n");
                iptables_blockage(log_line);
                break;
            case 3:
                printf("Sql Injection Union Filter Detection\n");
                iptables_blockage(log_line);
                break;
            case 4:
                printf("Sql Injection meta characters Filter Detection\n");
                iptables_blockage(log_line);
                break;
            case 5:
                printf("Sql Injection magic quote Filter Detection\n");
                iptables_blockage(log_line);
                break;
            default:
                break;
            }
        pid = fork();
        if (pid ==0){
          FILE *emails = popen("/usr/bin/mail -s 'WebAttack On server' root@localhost","w");
         fprintf(emails,"Attack FOUND %s ! in the log file.\n",log_line);
                        pclose(emails);
                        _exit(0);
                }
            }
 
    return 0;
}

 个人感觉作者在这里的代码逻辑有些混乱,应该在匹配成功后,结束代码,返回中间层的capture函数进行处理,在capture中用宏来替换代码

这样代码会少很多而且逻辑清晰:)

#define R(re,way,info) if(cap_matcher(log_line,#re,#way)){\
                        printf(##info);\
                        iptables_blockage(log_line);\
                        break;\
                    }

 四: 防火墙阻塞IP

apache的log文件如下格式:

127.0.0.1 - - [23/Sep/2011:15:27:36 +0800] "GET / HTTP/1.1" 200 44

因此以第一个空格为标准,获取IP后,调用iptables添加阻塞IP

1
2
3
4
5
6
7
8
9
10
11
12
13
14
void iptables_blockage(char *log_line){
    char *ip_address= malloc(100);
    char command[1000];
    int i;
    for (i =0; i <= 100 ; i++){
        if (isspace(log_line[i])){break;}
        ip_address[i] = log_line[i];
    }
    snprintf(command,sizeof(command),"/sbin/iptables -A INPUT -s %s -j DROP",ip_address);
    FILE *iptables_run = (FILE*)popen(command,"r");
    pclose(iptables_run);
 
    free(ip_address);
}

 

最后的邮件通知实现与iptables差不多,详细的可以参见前一篇文章,LINUX下C语言利用命令发邮件

posted on 2012-02-28 09:27  HackerVirus  阅读(478)  评论(0编辑  收藏  举报