Method for Improved Diagnostic Fault Isolation of Intermittent Multiple Bit Failures in an Interleaved Memory Design
Original Publication Date: 1987-Mar-01
Included in the Prior Art Database: 2005-Feb-01
An interleaved memory design employing a Single Error Correction/Double Error Detection (SEC/DED) Error Correcting Code (ECC) scheme has poor intermittent error isolation beyond Single Bit Error correction. A software interface which provides a diagnostic Field Replaceable Unit (FRU) isolation of intermittent multiple bit errors (MBEs) works as follows. This software interface is made between Retry/Recovery microcode and an Error Log Analysis (ELA) routine designed to isolate main storage machine checks. The Recovery software is required to preserve error counters that indicate the total number of bits, on a per array card basis, which have appeared to be defective at the time of recovery.