A paper (Divide and Conquer - HTTP Response Splitting, Web Cache Poisoning Attacks, and Related Topics) was released to describe various attacks that target web users through web application, browser, web/application server and proxy implementations. These attacks are described under the general category of HTTP Response Splitting and involve abusing various input validation flaws in these implementations to split HTTP responses into multiple parts in such a way that response data may be misrepresented to client users. Exploitation would occur by injecting variations of CR/LF sequences into parts of HTTP response headers that the attacker may control or influence. The general consequences of exploitation are that an attacker may misrepresent web content to the client, potentially enticing the user to trust the content and take actions based on this false trust. While the various implementations listed in the paper contribute to these attacks, this issue will most likely be exposed...
A paper (Divide and Conquer - HTTP Response Splitting, Web Cache Poisoning Attacks, and Related Topics) was released to describe various attacks that target web users through web application, browser, web/application server and proxy implementations. These attacks are described under the general category of HTTP Response Splitting and involve abusing various input validation flaws in these implementations to split HTTP responses into multiple parts in such a way that response data may be misrepresented to client users. Exploitation would occur by injecting variations of CR/LF sequences into parts of HTTP response headers that the attacker may control or influence. The general consequences of exploitation are that an attacker may misrepresent web content to the client, potentially enticing the user to trust the content and take actions based on this false trust. While the various implementations listed in the paper contribute to these attacks, this issue will most likely be exposed through web applications that do not properly account for CR/LF sequences when accepting user-supplied input that may be returned in server responses. This vulnerability could also aid in exploitation of cross-site scripting vulnerabilities.
