Browse Prior Art Database

ECC Hardware and Microcode Test Method

IP.com Disclosure Number: IPCOM000046913D
Original Publication Date: 1983-Aug-01
Included in the Prior Art Database: 2005-Feb-07
Document File: 1 page(s) / 12K

Publishing Venue

IBM

Related People

Chandler, DA: AUTHOR [+2]

Abstract

This is a method of normal read-mode testing the error correction code (ECC) hardware/microcode on a fixed block disk file which has defect skip. Defect skip is a disk organization wherein each data sector is provided with a fixed number of additional bytes and the identifier field (ID) includes the address of the first byte to be skipped when a defect is to be accommodated. When used, the ID includes the address of the start of the fixed number of bytes to be skipped which causes the data sector to include the fixed number of additional bytes provided at the end of the field. The advantage of this method is that a program has complete control of the ECC written in addition to most of the data. Thus, a burst error can be forced in the data, in the ECC, or across the data and ECC.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 58% of the total text.

Page 1 of 1

ECC Hardware and Microcode Test Method

This is a method of normal read-mode testing the error correction code (ECC) hardware/microcode on a fixed block disk file which has defect skip. Defect skip is a disk organization wherein each data sector is provided with a fixed number of additional bytes and the identifier field (ID) includes the address of the first byte to be skipped when a defect is to be accommodated. When used, the ID includes the address of the start of the fixed number of bytes to be skipped which causes the data sector to include the fixed number of additional bytes provided at the end of the field. The advantage of this method is that a program has complete control of the ECC written in addition to most of the data. Thus, a burst error can be forced in the data, in the ECC, or across the data and ECC. The length of the burst can be varied so that not only correction, but also detection modes of ECC can be tested. It also has the advantage that the data is read entirely from the disk with a normal read sequence without a requirement to preload a register, or go into a diagnostic mode. The procedure requires a disk file with a defect skip function. The defect skip length must be at least as large as the number of bytes in the ECC, and the format must provide for the data to be unchanged during a write ID sequence and readable by a read data command after the write ID. The following four steps are required. 1. In a file wherein the defect skip field...