Kernel panic

Abstract: We propose an intrusion prevention system called WHIPS that controls, entirely in kernel mode, the invocation of the critical system calls for the Windows OS security. WHIPS is implemented as a kernel driver, also called kernel module, by using kernel structures of the Windows OS. It is integrated without requiring changes to either the kernel data structures or to the kernel algorithms. WHIPS is also transparent to the application processes that continue to work correctly without source code changes or recompilation. A working prototype has been implemented as a kernel extension and it is applicable to all the Windows NT family OS, e.g. Windows 2000/XP/2003. The WHIPS first contribution is to apply the system call interposition technique to the Windows OS, which is not open source. It is not straightforward to apply this technique to Windows OS, also because Windows kernel structures are hidden from the developer, and furthermore, its kernel documentation is poor.

Abstract: Linux kernel oops is invoked when the kernel detects an erroneous state inside itself. It kills an offending process and allows Linux to continue its operation under a compromised reliability. We investigate how reliable Linux is after a kernel oops in this paper. To investigate the reliability after a kernel oops, we analyze the scope of error propagation through an experimental campaign of fault injection in Linux 2.6.38. The error propagation scope is process-local if an error is confined in the process context that activated it, while the scope is kernel-global if an error propagates to other processes' contexts or global data structures. If the scope is process-local, Linux can be reliable even after a kernel oops. Our findings are twofold. First, the error propagation scope is mostly process-local. Thus, Linux remains consistent after a kernel oops in most cases. Second, Linux stops its execution before accessing inconsistent states when kernel-global errors occur because synchronization primitives prevent the inconsistent states from being accessed by other processes.

Abstract: The UNIX® operating system, developed by AT&T Bell Laboratories, has become a standard operating system gaining rapid acceptance because of its superior flexibility, portability, and a number of support tools to increase programmer productivity. However, UNIX was originally designed for multitasking and time-sharing, and therefore conventional UNIX does not have an adequate response time and data throughput needed to support real-time applications.Many attempts have been made to adapt the UNIX kernel to provide a real-time environment. MODCOMP has developed REAL/IX operating system, which is a fully preemptive, low latency UNIX kernel. This paper discusses real-time performance of REAL/IX and compares it to MASSCOMP RTU operating system.

Abstract: Exception handling is a powerful abstraction that can be used to help manage errors and support the construction of reliable operating systems. Using exceptions to notify system components about exceptional conditions also reduces coupling of error handling code and increases the modularity of the system. We explore the benefits of incorporating exception handling into the Choices operating system in order to improve reliability. We extend the set of exceptional error conditions in the kernel to include critical kernel errors such as invalid memory access and undefined instructions by wrapping them with language-based software exceptions. This allows developers to handle both hardware and software exceptions in a simple and unified manner through the use of an exception hierarchy. We also describe a catch-rethrow approach for exception propagation across protection domains. When an exception is caught by the system, generic recovery techniques like policy-driven micro-reboots and restartable processes are applied, thus increasing the reliability of the system.