CVE-2020-5249

In Puma (RubyGem) before 4.3.3 and 3.12.4, if an application using Puma allows untrusted input in an early-hints header, an attacker can use a carriage return character to end the header and inject malicious content, such as additional headers or an entirely new response body. This vulnerability is known as HTTP Response Splitting. While not an attack in itself, response splitting is a vector for several other attacks, such as cross-site scripting (XSS). This is related to CVE-2020-5247, which fixed this vulnerability but only for regular responses. This has been fixed in 4.3.3 and 3.12.4.

In Puma (RubyGem) before 4.3.3 and 3.12.4, if an application using Puma allows untrusted input in an early-hints header, an attacker can use a carriage return character to end the header and inject malicious content, such as additional headers or an entirely new response body. This vulnerability is known as HTTP Response Splitting. While not an attack in itself, response splitting is a vector for several other attacks, such as cross-site scripting (XSS). This is related to CVE-2020-5247, which fixed this vulnerability but only for regular responses. This has been fixed in 4.3.3 and 3.12.4.

### Impact If an application using Puma allows untrusted input in an early-hints header, an attacker can use a carriage return character to end the header and inject malicious content, such as additional headers or an entirely new response body. This vulnerability is known as [HTTP Response Splitting](https://owasp.org/www-community/attacks/HTTP_Response_Splitting). While not an attack in itself, response splitting is a vector for several other attacks, such as cross-site scripting (XSS). This is related to [CVE-2020-5247](https://github.com/puma/puma/security/advisories/GHSA-84j7-475p-hp8v), which fixed this vulnerability but only for regular responses. ### Patches This has been fixed in 4.3.3 and 3.12.4. ### Workarounds Users can not allow untrusted/user input in the Early Hints response header. ### For more information If you have any questions or comments about this advisory: * Open an issue in [puma](https://github.com/puma/puma) * Email us a project maintainer. [Email addresses are listed in our Code of Conduct](https://github.com/puma/puma/blob/master/CODE_OF_CONDUCT.md#enforcement).

### Impact If an application using Puma allows untrusted input in an early-hints header, an attacker can use a carriage return character to end the header and inject malicious content, such as additional headers or an entirely new response body. This vulnerability is known as [HTTP Response Splitting](https://owasp.org/www-community/attacks/HTTP_Response_Splitting). While not an attack in itself, response splitting is a vector for several other attacks, such as cross-site scripting (XSS). This is related to [CVE-2020-5247](https://github.com/puma/puma/security/advisories/GHSA-84j7-475p-hp8v), which fixed this vulnerability but only for regular responses. ### Patches This has been fixed in 4.3.3 and 3.12.4. ### Workarounds Users can not allow untrusted/user input in the Early Hints response header. ### For more information If you have any questions or comments about this advisory: * Open an issue in [puma](https://github.com/puma/puma) * Email us a project maintainer. [Email addresses are listed in our Code of Conduct](https://github.com/puma/puma/blob/master/CODE_OF_CONDUCT.md#enforcement).

En Puma (RubyGem) versiones anteriores a 3.3 y 3.12.4, si una aplicación que usa Puma permite una entrada no confiable en un encabezado early-hint, un atacante puede usar un carácter retorno de carro para finalizar el encabezado e inyectar contenido malicioso, tales como encabezados adicionales o un cuerpo de respuesta completamente nuevo. Esta vulnerabilidad se conoce como División de Respuesta HTTP. Si bien no es un ataque en sí mismo, la división de la respuesta es un vector para varios otros ataques, tales como un cross-site scripting (XSS). Esto está relacionado con CVE-2020-5247, que corrigió esta vulnerabilidad pero solo para respuestas regulares. Esto se ha corregido en las versiones 4.3.3 y 3.12.4.