Chapter 5 - OSF/1 Exceptions and Interrupts (III)

Abstract

This chapter presents OSF/1 exceptions and interrupts. Exceptions are relevant primarily to the currently executing process. Exception service routines execute in response to exception conditions caused by software. All exception service routines execute in kernel mode on the kernel stack. Exception conditions consist of faults, arithmetic traps, and synchronous traps. A fault occurs during an instruction and leaves the registers and memory in a consistent state such that elimination of the fault condition and subsequent re-execution of the instruction gives correct results. An arithmetic trap occurs at the completion of the operation that caused the exception. A synchronous trap occurs at the completion of the operation that caused the exception. No instructions can be issued between the completion of the operation that caused the exception and the trap. The processor arbitrates interrupt requests. When the interrupt priority level (IPL) of an outstanding interrupt is greater than the current IPL, the processor raises IPL to the level of the interrupt and dispatches to entInt, the interrupt entry to the OS. Interrupts are serviced in kernel mode on the kernel stack. Interrupts can come from one of four sources: (1) I/O devices, (2) the clock, (3) performance counters, or (4) machine checks.