Browse Prior Art Database

Bus Unit Controller Freeze on Program Input/Outpug Error

IP.com Disclosure Number: IPCOM000114854D
Original Publication Date: 1995-Feb-01
Included in the Prior Art Database: 2005-Mar-30
Document File: 2 page(s) / 63K

Publishing Venue

IBM

Related People

Arndt, RL: AUTHOR [+2]

Abstract

The general structure of a computer system is shown in the Figure. In such a system, data integrity may be a concern. If so, error detection and sometimes error recovery is built in to the system. The various components of the system include, both hardware (for example parity checking) and software. Where error recoverability is a concern: 1. The system design must assure that the hardware cannot get in a state from which the software cannot recover. 2. The hardware must assure that errors in the data must not get propagated to a place where the data will be used as good data when it is not.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 53% of the total text.

Bus Unit Controller Freeze on Program Input/Outpug Error

      The general structure of a computer system is shown in the
Figure.  In such a system, data integrity may be a concern.  If so,
error detection and sometimes error recovery is built in to the
system.  The various components of the system include, both hardware
(for example parity checking) and software.  Where error
recoverability
is a concern:
  1.  The system design must assure that the hardware cannot get in a
       state from which the software cannot recover.
  2.  The hardware must assure that errors in the data must not get
       propagated to a place where the data will be used as good data
       when it is not.

      Various I/O subsystems have taken various approaches to
facilitate error recoverability and data integrity.  For example,
some systems have implemented synchronous error handling where the
error is reported by the end of the instruction that caused the
error.  This allows the software to recover and error propagation to
be prevented because the software has not continued on past the point
where the error occurred.  The disadvantage of such systems is that
Programmed I/O (PIO) operations stall the processor and thus degrade
the system performance somewhat.  Other systems take the approach
that this problem is to be handled as a cooperative effort between
the I/O device hardware and the device driver.  Although this latter
approach can be made to work...