Browse Prior Art Database

File Access Method Which is File System Independent

IP.com Disclosure Number: IPCOM000100876D
Original Publication Date: 1990-Jun-01
Included in the Prior Art Database: 2005-Mar-16
Document File: 2 page(s) / 74K

Publishing Venue

IBM

Related People

Bealkowski, R: AUTHOR

Abstract

This article describes a file system independent access method which will increase computer system performance when an intelligent hardware subsystem is attached.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

File Access Method Which is File System Independent

       This article describes a file system independent access
method which will increase computer system performance when an
intelligent hardware subsystem is attached.

      The layout of data on a disk depends on the file system used.
Different types of operating systems define different data structures
which represent the mapping of a file onto a physical device.  A
method is required which will allow bus-masters and input/output
(I/O) devices to access files independent of a particular file
system.  This allows an I/O device peer-to-peer file transfer without
incurring any host overhead.

      A file cannot be retrieved from disk without knowing how and
where it is stored on the physical device. Normally, the operating
system provides a file system which manages disk storage and the
associated files.  Different types of operating systems define
different data structures which represent the mapping of a file onto
a physical device.  This poses a problem for a hardware level
subsystem which requires access to files.  An intelligent bus-master
printer attachment is one example of a hardware level subsystem.  The
hardware must be kept as a platform to support any operating system
implementation or operating system extensions.  Thus, hardware which
is operating system and file system dependent is unacceptable.

      The organization of a physical disk drive requires that data be
stored in blocks called sectors.  Two main methods of identifying
sectors are relative block address (RBA) and cylinder head record
(CHR) notations.  The RBA method numbers each sector sequentially,
providing a scheme which is independent of the topology of the disk....