Method and apparatus for indicating uncorrectable data errors

• US 5,953,351 A
• Filed: 09/15/1995
• Issued: 09/14/1999
• Est. Priority Date: 09/15/1995
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Topps, Avery Cox, Hicks, Dwain Alan
• Primary Examiner(s): MOISE, EMMANUEL LIONEL
• Application Number:
  US08/529,125
• Time in Patent Office: 1,460 Days
• Field of Search: 371/40.1, 371/40.2, 371/51.1, 371/37.12, 371/37.5, 371/37.6, 371/37.7, 371/40.11, 371/40.12, 371/40.13, 371/40.17, 371/40.3, 371/42, 371/47.1, 395/182.04, 395/185.07, 395/183.19, 395/185.01, 395/185.05, 395/185.09, 395/280
• US Class Current: 714/763
• CPC Class Codes: G06F 11/1024 Identification of the type of error

Abstract

A method and apparatus for identifying data that contains an uncorrectable error may be accomplished in a computer that includes a memory unit operably coupled to a processor. The memory unit includes an error detection circuit that, when an uncorrectable storage error is detected, produces transmit check bits indicating that the data being transmitted includes an uncorrectable storage error. The processor, which includes a check bit decoder, upon receiving the transmit check bits, interprets the transmit check bits to identify the uncorrectable error. When the uncorrectable error is identified, the check bit decoder provides a data error signal to a processing core of the processor, thereby interrupting the processing core which avoids a system error and the need to reboot the computer.

Claims

1. A method for indicating to a processor that data having an uncorrectable error is to be received, the method comprising the steps of:

a) when a data transfer request is received by a memory unit, performing a storage error detection process on a data block of the memory unit, wherein the data transfer request identifies the data block;

b) when the storage error detection process indicates an uncorrectable error, performing an error coding process to produce transmit check bits which indicate that the data block includes the uncorrectable error;

c) routing the data and the transmit check bits from the memory unit to the processor; and

d) when the data and the transmit check bits are received, performing an error decode process based on the transmit check bits to determine that the data includes the uncorrectable error, such that the data having the uncorrectable error is identified.

2. The method of claim 1, wherein the step of performing the storage error detection process further comprises the steps of:

reading data and stored check bits from the data block;

calculating first check bits from the data;

when the first check bits do not match the stored check bits, identifying a storage error in the data; and

determining whether the storage error is correctable or uncorrectable.

3. The method of claim 2 further comprises:

when the first check bits match the stored check bits, performing the error coding process to produce the transmit check bits which indicate that the data does not include a storage error.

4. The method of claim 3, wherein the step of performing the error decode process further comprises the steps of:

generating receive check bits from the data;

comparing the receive check bits with the transmit check bits;

when the received check bits match the transmit check bits, utilizing the data.

5. The method of claim 3, wherein the step of performing the error decode process further comprises the steps of:

generating receive check bits from the data;

comparing the receive check bits with the transmit check bits;

when the receive check bits do not match the transmit check bits, identifying a transmission error; and

determining whether the transmission error is correctable or uncorrectable.

6. The method of claim 5 further comprises the steps of:

when the transmission error is correctable, correcting the transmission error to produce corrected data; and

utilizing the corrected data.

7. The method of claim 5 further comprises the step of:

when the transmission error is uncorrectable, interrupting the processor such that the processor does not process the data.

8. The method of claim 2 further comprises the steps of:

when the storage error is correctable, correcting the storage error to produce a corrected data; and

performing the error encode process on the corrected data to produce the transmit check bits which indicate that the corrected data does not include the storage error.

9. The method of claim 8, wherein the step of performing the error decode process further comprises the steps of:

generating receive check bits from the corrected data;

comparing the receive check bits with the transmit check bits; and

when the receive check bits match the transmit check bits, utilizing the corrected data.

10. The method of claim 8, wherein the step of performing the error decode process further comprises the steps of:

generating receive check bits from the corrected data;

comparing the receive check bits with the transmit check bits; and

when the receive check bits do not match the transmit check bits, determining whether a transmission error is correctable or uncorrectable.

11. The method of claim 1, wherein the step of performing the storage error detection process further comprises the step of performing the storage error detection process using a first error correction process, and wherein the step of performing the error decode process further comprises the step of performing the error decode process using a second error correction process.

12. The method of claim 1, wherein the step of performing the error decode process further comprises the steps of:

generating receive check bits from the data;

comparing the receive check bits with the uncorrectable check bits; and

when the receive check bits do not match the uncorrectable check bits, determining that the data includes the uncorrectable error.

13. The method of claim 12, wherein the step of determining that the data includes the uncorrectable error further comprises the steps of:

decoding the transmit check bits to identify an entry; and

determining that the entry is not contained within the data block.

14. The method of claim 1 further comprises the steps of:

when the storage error detection process indicates a correctable storage error, performing, by the memory unit, the error coding process to produce transmit check bits, wherein the transmit check bits indicate the storage error;

transmit the data and the transmit check bits to the processor;

performing, by the processor, the error decode process to determine the storage error; and

correcting, by the processor, the storage error to produce corrected data.

15. A data processing system comprising:

memory unit that includes;

data storage that stores data and stored check bits;

storage error detection circuit operably coupled to receive the data and the stored check bits, wherein the storage error detection circuit, using a first error detection process, produces an uncorrectable error signal when an uncorrectable storage error is detected;

check bit encoder operably coupled to the storage error detection circuit, wherein the check bit encoder produces transmit check bits when the uncorrectable error signal is received, wherein the transmit check bits indicate the uncorrectable storage error;

a processor operably coupled to the memory unit, wherein the processor includes;

a processing core that processes data, wherein the processing core does not process the data when a data error signal is received; and

transmit check bit decoder operably coupled to receive a representation of the transmit check bits and produces, in response thereto, the data error signal when the transmit check bits indicate that the data includes the uncorrectable storage error.

16. The data processing system of claim 15, wherein the check bit encoder further comprises means for generating the transmit check bits from the data to indicate that the data does not include the uncorrectable storage error.

17. The data processing system of claim 16, wherein the check bit encoder further comprises a multiplexor having inputs to receive the transmit check bits and inverted transmit check bits, wherein the multiplexor outputs, when the uncorrectable storage error exists, the inverted transmit check bits as the transmit check bits to indicate the uncorrectable storage error.

18. The data processing system of claim 15, wherein the processor further comprises a transmit error detection circuit operably coupled to receive the data, wherein the transmit error detection circuit determines whether the data contains a transmission error and generates the data error signal, when the transmission error is uncorrectable.

19. The data processing system of claim 18, wherein the processor further comprises a data correction circuit operably coupled to the transmit error detection circuit and the transmit check bit decoder, wherein the data correction circuit corrects the data when either the transmit error detection circuit or the transmit check bit decoder indicate a correctable error.

20. A memory unit comprising:

data storage for storing data and stored check bits;

storage error detection circuit operably coupled to the data storage, wherein the storage error detection circuit produces uncorrectable error signal when an uncorrectable storage error is detected within the data; and

check bit encoder operably coupled to the storage error detection circuit, wherein the check bit encoder produces transmit check bits that indicate the uncorrectable storage error when the uncorrectable error signal is received.

21. A processor comprising:

a processing core that processes data based upon operational codes, wherein the processing core does not process the data when a data error signal is received; and

transmit check bit decoder operably coupled to receive transmit check bits and produces, in response thereto, the data error signal when the transmit check bits indicate that the data includes an uncorrectable error.