We present a consistency analysis approach to assist the Linux community in verifying the correctness of authorization hook placement in the Linux Security Modules (LSM) framework. The LSM framework consists of a set of authorization hooks inserted into the Linux kernel to enable additional authorizations to be performed (e.g., for mandatory access control). When compared to system call interposition, authorization within the kernel has both security and performance advantages, but it is more difficult to verify that placement of the LSM hooks ensures that all the kernel's security-sensitive operations are authorized. Static analysis has been used previously to verified mediation (i.e., that some hook mediates access to a security-sensitive operation), but that work did not determine whether the necessary set of authorizations were checked. In this paper, we develop an approach to test the consistency of the relationships between security-sensitive operations and LSM hooks. The idea is that whenever a security-sensitive operation is performed as part of specifiable event, a particular set of LSM hooks must have mediated that operation. This work demonstrates that the number of events that impact consistency is manageable and that the notion of consistency is useful for verifying correctness. We describe our consistency approach for performing verification, the implementation of run-time tools that implement this approach, the anomalous situations found in an LSM-patched Linux 2.4.16 kernel, and an implementation of a static analysis version of this approach.

We present the concept of an access control space and investigate how it may be useful in managing access control policies. An access control space represents the permission assignment state of a subject or role. For example, the set of permissions explicitly assigned to a role defines its specified subspace, and the set of constraints precluding assignment to that role defines its prohibited subspace. In analyzing these subspaces, we identify two problems: (1) often a significant portion of an access control space has unknown assignment semantics, which indicates that the policy is underspecified; and (2) often high-level assignments and constraints that are easily understood result in conflicts, where resolution often leads to significantly more complex specifications. We have developed a prototype system, called Gokyo, that computes access control spaces. Gokyo identifies the unknown subspace to assist system administrators in developing more complete policy specifications. Also, Gokyo identifies conflicting subspaces and enables system administrators to resolve conflicts in a variety of ways in order to preserve the simplicity of constraint specification. We demonstrate Gokyo by analyzing a Web server policy example and examine its utility by applying it to the SELinux example policy. Even for the extensive SELinux example policy, we find that only eight additional expressions are necessary to resolve Apache administrator policy conflicts.

We present runtime tools to assist the Linux community in verifying the correctness of the Linux Security Modules (LSM) framework. The LSM framework consists of a set of authorization hooks inserted into the Linux kernel to enable additional authorizations to be performed (e.g., for mandatory access control). When compared to system call interposition, authorization within the kernel has both security and performance advantages, but it is more difficult to verify that placement of the LSM hooks ensures that all the kernel's security-sensitive operations are authorized. We have examined both static and runtime analysis techniques for this verification, and have found them to be complementary. Static analysis is more complex to implement and tends to generate more false positives, but coverage of all type-safe execution paths is possible. Runtime analysis lacks the code and input coverage of static analysis, but tends to be simpler to gather useful information. The major simplifying factor in our runtime verification approach is that we can leverage the fact that most of the LSM hooks are properly placed to identify misplaced hooks. Our runtime verification tools collect the current LSM authorizations and find inconsistencies in these authorizations. We describe our approach for performing runtime verification, the design of the tools that implement this approach, and the anomalous situations found in an LSM-patched Linux 2.4.16 kernel.

We present the concept of an access control space and investigate how it may be useful in managing access control policies. An access control space represents the permission assignment state of a subject. We identify subspaces that have meaningful semantics. For example, the set permissions explicitly assigned to a subject defines its specified subspace, and constraints define the prohibited subspace. In analyzing these subspaces, we identify two problems: (1) often a significant portion of the access control space has unknown assignment semantics, meaning that it is not defined whether an assignment in this space should be permitted or not, and (2) often high-level assignments and constraints that are easily understood result in conflicts where permissions are both specified and prohibited. To solve these problems, we have developed a tool, called Gokyo, that enables definition and analysis of access control spaces. Gokyo computes the unknown subspace to show system administrators the ambiguous region and enable them to reduce it. Gokyo identifies conflicting subspaces and enables system administrators to handle subspaces as exceptions, if desired. We demonstrate the utility of Gokyo by analyzing a web server policy example.

This paper makes the claim that secure extensibility of operating systems is not only desirable but also achievable. We claim that OS extensibility should be done at user-level to avoid the security problems inherent in other approaches. We furthermore claim (backed up by some initial results) that user-level extensibility is possible at a performance that is similar to in-kernel extensions. Finally, user-level extensions allow the use of modern software engineering techniques.