초록 ▼

A multiprocessor computer includes a fault detection scheme which detects and identifies the failure of one of the processors. Each processor is assigned a write location, which may be a unique register. During normal computer operation, each processor intermittently performs a test and stores the r

A multiprocessor computer includes a fault detection scheme which detects and identifies the failure of one of the processors. Each processor is assigned a write location, which may be a unique register. During normal computer operation, each processor intermittently performs a test and stores the results of the test in the assigned write location. The stored results are compared to expected results, and an error signal is generated if the stored results differ from the expected results to indicate that one of the processors has failed.

대표청구항 ▼

1. A method of determining a processor failure in a multiprocessor computer, the method comprising the acts of:(a) identifying each processor in the computer; (b) assigning a unique write location to each identified processor; (c) directing each processor to perform a given test and to store a resul

1. A method of determining a processor failure in a multiprocessor computer, the method comprising the acts of:(a) identifying each processor in the computer; (b) assigning a unique write location to each identified processor; (c) directing each processor to perform a given test and to store a result of the test in the respective unique write location; (d) comparing the result stored in each unique write location to a respective expected result; and (e) indicating a failure of a particular one of the identified processors in response to the stored result being different than the expected result.