Final answer:
Statements B) and D) do not accurately define the requirements of a security kernel. Process isolation is related but not a core requirement, while implementing a method to circumvent security directly contradicts the purpose of a security kernel.
Step-by-step explanation:
The question asks which statements do not define the requirements of a security kernel. The security kernel is a critical part of computer security, often implemented within operating systems to enforce access control policies. There are two statements here that do not accurately define the requirements of a security kernel:
- The reference monitor should provide process isolation.
- A method to circumvent the security should be implemented by the reference monitor.
While the reference monitor does indeed provide process isolation, this is not a defining requirement of the security kernel itself. Instead, it is more accurately described as a function that the security kernel may control or influence through the reference monitor. The more critical misstatement is the suggestion that a method to circumvent security should be implemented. This goes against the fundamental principles of security systems, where the utmost effort is made to prevent circumvention rather than enable it. Verification of the reference monitor's correctness and a comprehensive verification of the security kernel are both true requirements for a security kernel.