Skip to contentSkip to conversation

selinux: Fix error priority for bind with AF_UNSPEC on PF_INET6 socket

PATCH In review Mickaël Salaün 3 participants 3 messages Jan 3, 2024 – Jan 4, 2024

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 selinux: Fix error priority for bind with AF_UNSPEC on PF_INET6 socket

The IPv6 network stack first checks the sockaddr length (-EINVAL error) before checking the family (-EAFNOSUPPORT error).

This was discovered thanks to commit a549d055a22e ("selftests/landlock:
Add network tests").

Cc Eric Paris <[email protected]>
Cc Konstantin Meskhidze <[email protected]>
Cc Paul Moore <[email protected]>
Cc Stephen Smalley <[email protected]>
Fixes 0f8db8cc73df ("selinux: add AF_UNSPEC and INADDR_ANY checks to selinux_socket_bind()")

---

Changes since v2:
https://lore.kernel.org/r/[email protected]
* Add !PF_INET6 check and comments (suggested by Paul).
* s/AF_INET/PF_INET/g (cosmetic change).

Changes since v1:
https://lore.kernel.org/r/[email protected]
* Use the "family" variable (suggested by Paul).

---
 security/selinux/hooks.c | 7 +++++++
 1 file changed, 7 insertions(+)
security/selinux/hooks.c +7 -0
index feda711c6b7b..8b1429eb2db5 100644
@@ -4667,6 +4667,13 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in
 				return -EINVAL;
 			addr4 = (struct sockaddr_in *)address;
 			if (family_sa == AF_UNSPEC) {
+				if (family == PF_INET6) {
+					/* Length check from inet6_bind_sk() */
+					if (addrlen < SIN6_LEN_RFC2133)
+						return -EINVAL;
+					/* Family check from __inet6_bind() */
+					goto err_af;
+				}
 				/* see __inet_bind(), we only want to allow
 				 * AF_UNSPEC if the address is INADDR_ANY
 				 */
Signature · 1 line
2.43.0
Tested-by on v3 1/1
#1 lore raw
Quoted 6 lines · Mickaël Salaün

The IPv6 network stack first checks the sockaddr length (-EINVAL error) before checking the family (-EAFNOSUPPORT error).

This was discovered thanks to commit a549d055a22e ("selftests/landlock:
Add network tests").

Quoted 8 lines
Cc Eric Paris <[email protected]>
Cc Konstantin Meskhidze <[email protected]>
Cc Paul Moore <[email protected]>
Cc Stephen Smalley <[email protected]>
Fixes 0f8db8cc73df ("selinux: add AF_UNSPEC and INADDR_ANY checks to selinux_socket_bind()")

Thank you Mickaël for the patch. Tested patch on v6.7-rc8.

Quoted 32 lines

---

Changes since v2:
https://lore.kernel.org/r/[email protected]
* Add !PF_INET6 check and comments (suggested by Paul).
* s/AF_INET/PF_INET/g (cosmetic change).

Changes since v1:
https://lore.kernel.org/r/[email protected]
* Use the "family" variable (suggested by Paul).

---
 security/selinux/hooks.c | 7 +++++++
 1 file changed, 7 insertions(+)
security/selinux/hooks.c +7 -0
index feda711c6b7b..8b1429eb2db5 100644
@@ -4667,6 +4667,13 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in
 				return -EINVAL;
 			addr4 = (struct sockaddr_in *)address;
 			if (family_sa == AF_UNSPEC) {
+				if (family == PF_INET6) {
+					/* Length check from inet6_bind_sk() */
+					if (addrlen < SIN6_LEN_RFC2133)
+						return -EINVAL;
+					/* Family check from __inet6_bind() */
+					goto err_af;
+				}
 				/* see __inet_bind(), we only want to allow
 				 * AF_UNSPEC if the address is INADDR_ANY
 				 */
Signature · 2 lines
BR,
Muhammad Usama Anjum
maintainer
#2 lore raw
Quoted 16 lines · Muhammad Usama Anjum
Quoted 6 lines · Mickaël Salaün

The IPv6 network stack first checks the sockaddr length (-EINVAL error) before checking the family (-EAFNOSUPPORT error).

This was discovered thanks to commit a549d055a22e ("selftests/landlock:
Add network tests").

Quoted 8 lines
Cc Eric Paris <[email protected]>
Cc Konstantin Meskhidze <[email protected]>
Cc Paul Moore <[email protected]>
Cc Stephen Smalley <[email protected]>
Fixes 0f8db8cc73df ("selinux: add AF_UNSPEC and INADDR_ANY checks to selinux_socket_bind()")
Quoted 3 lines

Thank you Mickaël for the patch. Tested patch on v6.7-rc8.

This looks good to me, and since it is rather trivial I'm going to go ahead and merge this into selinux/next so it should go up to Linus during the upcoming merge window.

Thanks everyone!

Signature · 1 line
paul-moore.com