Browse Prior Art Database

Pointer Collection and Reporting for Error Recovery Procedures Branching

IP.com Disclosure Number: IPCOM000117898D
Original Publication Date: 1996-Jul-01
Included in the Prior Art Database: 2005-Mar-31
Document File: 4 page(s) / 143K

Publishing Venue

IBM

Related People

Bailey, CE: AUTHOR [+4]

Abstract

Disclosed is a method of condensing many different types of error pointers in a multi-track magnetic tape system for use by the Pointer Control Circuit (PCC) of the Error Correction Code (ECC) decoder. Error pointers are categorized according to their type (definite or possible) and approximate location within a code block (first half or second half). This reduces the complexity of the PCC and still allows error pointers so as to optimize the effectiveness of the ECC decoder. The method also summarizes the distribution of errors on the tape into a concise format that can be handled easily by the microcode Error Recovery Procedures (ERP).

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 40% of the total text.

Pointer Collection and Reporting for Error Recovery Procedures Branching

      Disclosed is a method of condensing many different types of
error pointers in a multi-track magnetic tape system for use by the
Pointer Control Circuit (PCC) of the Error Correction Code (ECC)
decoder.  Error pointers are categorized according to their type
(definite or possible) and approximate location within a code block
(first half or second half).  This reduces the complexity of the PCC
and still allows error pointers so as to optimize the effectiveness
of the ECC decoder.  The method also summarizes the distribution of
errors on the tape into a concise format that can be handled easily
by the microcode Error Recovery Procedures (ERP).

Assume a system that:
  1.  uses a product code (an outer-code, inner-code code block)
where:
      o  the outer code is an error correction code
      o  and the inner code is an error detection code (generates
          error location pointers for the outer code),
  2.  interleaves the inner code rows or Minimum Interleave Elements
       (MIEs) over multiple product code code-blocks, which make up a
       Minimum Interleave Unit (MIU),
  3.  creates additional error pointers as a function of
      o  the modulation code and
      o  quality estimates of each track's timing,
  4.  employs a Deinterleaving Buffer to reassemble the MIEs back
into
       their original code block, along with their associated error
       pointers
  5.  and employs a Pointer Control Circuit (PCC) that prioritizes
and
       manages the pointers applied to the outer code decoder in
order
       to maximize its effectiveness.

      The Pointer Control Circuit in such a system does not need to
be concerned with the source of an error pointer.  It only needs to
know which pointers are for definite errors, which ones are for
possible errors, and in some cases, in which half of the code block
the error occurred.

      It is also useful to the microcode to know what the
distribution of errors was like on the tape.  It is important to know
if most of the errors occurred along certain tracks, or if there were
large clusters of errors that affected many tracks simultaneously.

This is useful for separating media defects from hardware problems.
It is also useful for tuning the error correction/pointer control
algorithms for obtaining the best error correction performance.
Since the various pointer detection circuits provide a large amount
of information in the course of reading a block, this information
must be condensed in order to minimize the storage required as well
as minimize the processing that must be done by the error recovery
procedures, but still provide the necessary information.

      The track logic of the read dataflow of such a system is
capable of performing the following error pointer indications:
  ICPTR     an error was detected by...