Abstract: A system and method for establishing a passive recovery environment for an operating system kernel. A cross reference between an instruction address range and recovery routine address is established during binding of the kernel. Normal process initiation and termination includes no recovery overhead. In the case of system failure, the failing instruction address is used to access the recovery routine cross reference tables to determine recovery routine address. The error detection process then initiates recovery routine processing.

Abstract: A system and method for preventing deadlock in a multiprocessor computer system executing instructions requiring multiple resources. The system detects potential deadlock situations where a multi-resource instruction is blocked from obtaining one of the resources. A multi-resource instruction global lock is provided that can be held by at most one processor. Upon conflict detection, the processor attempts to acquire the multi-resource instruction global lock and, if successful, resumes resource acquisition. The use of a global lock serializes multiple resource requests and assures that the processor holding the lock can eventually acquire all required resources without deadlock with another processor. The preferred embodiment acquires the global lock on an exception basis to minimize the overhead impact. However, an alternate embodiment which uses the global lock in each multiple resource instruction could also be implemented.

Abstract: A method for assembling swap blocks of virtual pages for transfer between a backing store and physical memory in a data processing system. The method includes segregating all virtual pages resident in physical memory between an active list and an inactive list. Virtual pages are then assigned to the inactive list by groups where each page in a group belongs to a single owning task or object. From a group, virtual pages are assigned to a swap block based upon correlation of most recent use. The swap block may then be paged out in a single operation to a backing store. A list of the group members is kept to permit page in to physical memory upon reference to a member of the swap block.

Abstract: A method for serializing access to computer system resources without disabling interrupts in critical code sections or requiring excessive use of spin locks. A queue lock is introduced that allows an interrupt process to enqueue and be processed in turn without a spin lock and to block process code access to a particular resource until all interrupts are processed. Process level code is blocked from accessing a locked resource while requests for accesses from interrupt level code are queued in a deferred work queue which is processed prior to the release of the blocking lock. Establishment of deferred work queue means that processing can continue without disablement of interrupts and without significant overhead consumed by processes holding spin locks.