Browse Prior Art Database

Memory Tag Processing And Checking

IP.com Disclosure Number: IPCOM000099553D
Original Publication Date: 1990-Feb-01
Included in the Prior Art Database: 2005-Mar-15
Document File: 2 page(s) / 53K

Publishing Venue

IBM

Related People

Thornley, JS: AUTHOR

Abstract

Disclosed is a circuit in which memory access tags (point- ers and counters) are updated. The tags are processed in such a way as to provide trace diagnostic aids and validity checking of tags with minimal logic. The tags provide the necessary information to control main memory access in either random or contiguous sequences. The random access is sometime referred to as a "scatter gather" operation. The number of hardware tags is determined by the system requirements.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 74% of the total text.

Memory Tag Processing And Checking

       Disclosed is a circuit in which memory access tags
(point- ers and counters) are updated. The tags are processed in such
a way as to  provide trace diagnostic aids and validity checking of
tags with minimal logic.  The tags provide the necessary information
to control main memory access in either random or contiguous
sequences.  The random access is sometime referred to as a "scatter
gather" operation. The number of hardware tags is determined by the
system requirements.

      Each of the tags contains a main memory page pointer, line
pointer (each page contains a given number of lines of data), and
line count (the number of lines to be transferred to/from main
memory). The information in the tags is supplied by the system
software.

      When the tag array (1) is loaded by software, the tag adr reg
(3) points to the beginning of the tag array (1). The first tag value
is then read out and placed into the tag reg (4). The tag information
is used by the hardware to select a page of main memory and a line
within the  page. Each time a line of data (to/from main memory) and
a tag is processed,  the decodes (7) and (8) check for end conditions
and validity. The validity is a simple process of testing overflow
(7) and line count zero (8).

      The conventional method of testing for validity is done at the
beginning of the operation.  The process of testing for a condition
of the line count being large enough to...