Ethical Hacking Learn to find vulnerabilities before the bad guys do! Gain real world hands on hacking experience in our state of the art hacking lab. Course designed and taught by expert instructors with years of penetration testing experience. 12 student maximum in every class. Certification attempt included in every package. | Computer Forensics Training at InfoSec Institute Gain the in-demand skills of a certified computer examiner, learn to recover trace data left behind by fraud, theft, and cybercrime perpetrators. Discover the source of computer crime and abuse at your organization so that it never happens again. All of our class sizes are guaranteed to be 12 students or less to facilitate one-on-one interaction with one of our expert instructors. |

| Subject: | Sending multipart/form-data requests from Flash (with arbitrary headers) |
|---|---|
| Date: | Thu, 10 Aug 2006 09:25:32 +0200 |
Hello lists, In my original "Forging HTTP request headers with Flash" paper (http://www.securityfocus.com/archive/1/441014), I mentioned forcing multipart/form-data input format to ensure that Flash's LoadVars isn't used to forge the request. However, there's a work-around for the attacker - using Flash's XML object. The attacker can forge almost any practical input for multipart/form-data (as long as it's part of a well-formed XML document): var req:XML=new XML("<foo>\r\n--bar\r\nContent-Disposition: form-data; name=field1\r\n\r\ndata1\r\n--bar\r\nContent-Disposition: form-data; name=field2\r\n\r\ndata2\r\n--bar--\r\n</foo>"); req.addRequestHeader("Content-Type","multipart/form-data; boundary=\"bar\""); req.addRequestHeader("Referer","http://some.site/"); req.addRequestHeader("Expect","something"); req.send("http://www.vuln.site/path/to/script.cgi","_blank"); It seems that the capabilities of the XML object (XML.send), in terms of forging HTTP request headers, are similar to those of LoadVars (LoadVars.send). The above code works well in IE 6.0 (Flash 7 and 8). The reader may note that there's a bit of data in the body before the first MIME part (that's the <foo> start-element) as well as a bit of data after the last MIME part (</foo>). As it happens, the multipart Content-Type allows such "preamble" (data before the first multipart boundary delimiter) and "epilogue" (data after the last multipart boundary delimiter) - see RFC 2046 section 5.1.1, and indeed application engines (e.g. PHP) silently ignore those parts. One shortcoming of the technique is that it cannot express all possible data. This is due to the fact that the XML object stores its XML data in an XML tree, and then serializes this tree to an XML document upon invocation of the send() method. This serialization doesn't duplicate the exact XML input (byte-wise, although logically the same tree would result), and it also guarantees that the serialized data is a well formed XML document. For example, any double quote marks will be converted to their XML entity representation ("), likewise for several other symbols. Theoretically, this means that it's still possible to distinguish a Flash submission from a browser submission, by forcing data with those symbols to be sent, e.g. using hidden fields with those values. But I wouldn't recommend this practice as it doesn't sound like "The Right Thing" - it's way too implementation specific. Bottom line: multipart/form-data submissions can be forged almost completely, headers and all, using Flash 7/8 (in IE 6.0). -Amit ------------------------------------------------------------------------- Sponsored by: Watchfire Watchfire was recently named the worldwide market leader in Web application security assessment tools by both Gartner and IDC. Download a free trial of AppScan today and see why more customers choose AppScan then any other solution. Try it today! https://www.watchfire.com/securearea/appscancamp.aspx?id=701500000008VnB --------------------------------------------------------------------------
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Environment for testing WebApp Security Scanners, mikeiscool |
|---|---|
| Next by Date: | Unable to disable browser caching in Firefox through HTTP headers, smith . norton |
| Previous by Thread: | Ruby On Rails 1.1.5 Released to Address Critical Vulnerability, bugtraq |
| Next by Thread: | Unable to disable browser caching in Firefox through HTTP headers, smith . norton |
| Indexes: | [Date] [Thread] [Top] [All Lists] |