| 摘要 |
PURPOSE:To reduce the possibility of danger causing an error due to an intermittent fault during retrial and causing a system down and job abort, etc., to which the retrial of instruction of the error is impossible by executing the retrial of instruction that became an error with an another normal logic unit without using the logic unit that caused the error. CONSTITUTION:When the fault is occurred in a logic unit 11 and detected by a supervisory means 14, the operation of the logic unit 11 is suspended and the internal state is preserved, information the logic unit 11 has a fault is given to a diagnosing processor 2 through a diagnosing interface 3. In response to this, the diagnosing processor 2 starts an internal state reading means 21, and the internal state of the logic unit 11 preserved by a supervisory means 14 is read out by the means 21. Then, the diagnosing processor 2 judges the fault detected from the internal state read out by a judging means 22. When it is judged that retrial is possible, and function degrade operation is not possible, a normal logic unit 13 is made to the reexecution and continuation of succeeding processing. |
