System for determining status of errors in a memory subsystem

Claim 1

In a data processing system, a subsystem for determining the status of an error that occurred in a multi-chip memory storage means comprising: (a) error detecting means for detecting an error, said error detecting means receiving data and check bits associated therewith for generating syndrome bits as a function of said data and of said check bits; and (b) error logging means for determining the status of all detected errors operatively connected to said error detecting means for receiving said syndrome bits therefrom, said error logging means further comprising monitoring means for distinguishing between hard errors and soft errors, and scrubbing means for sequentially accessing predetermined physical locations in said multi-chip memory storage means in which errors have been detected, including incrementing address counters for logging said predetermined physical locations in said multi-chip memory storage means in which said errors have occurred.

Claim 9

A method for determining whether an error that occurs in a multi-chip memory subsystem having a plurality of physical storage locations in a data processing system is a repeating error in order to assess imminent storage location failures, the steps comprising: (a) reading data and check bits associated therewith from a predetermined physical storage location; (b) generating syndrome bits as a function of said data and check bits; (c) analyzing said syndrome bits to determine if an error occurred; (d) logging said predetermined physical storage location associated with an error, if detected, and counting how many times an error occurs in a particular storage location in order to assess imminent storage location failures; and (e) scrubbing a plurality of physical storage locations.

Claim 10

A method for determining whether an error that occurs in a multi-chip memory subsystem having a plurality of physical storage locations in a data processing system is a repeating error in order to assess imminent storage location failures, the steps comprising: (a) reading data and check bits associated therewith from a predetermined physical storage location; (b) generating syndrome bits as a function of said data and check bits; (c) analyzing said syndrome bits to determine if an error occurred; (d) logging said predetermined physical storage location associated with an error, if detected, and counting how many times an error occurs in a particular storage location in order to assess imminent storage location failures; and (e) scrubbing a plurality of physical storage locations.