Browse Prior Art Database

Generalized Routing Process for Personal Computer And Communication Card Applications

IP.com Disclosure Number: IPCOM000120234D
Original Publication Date: 1991-Apr-01
Included in the Prior Art Database: 2005-Apr-02
Document File: 4 page(s) / 134K

Publishing Venue

IBM

Related People

Christol, J: AUTHOR [+2]

Abstract

The use of a Personal Computer with plugged Intelligent Cards is more and more generalized.

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

Generalized Routing Process for Personal Computer And Communication
Card Applications

      The use of a Personal Computer with plugged Intelligent
Cards is more and more generalized.

      Applications can be written under the Personal Computer
Operating System or under the Card Operating System according to
function location.

      It can be interesting to be able to route data between any
couple of these applications, without needing to know if they are
located in the Personal Computer or in a Card.

      Described is a Generalized Routing Process (GRP) allowing any
application in the system to communicate with any other one without
consideration of where it is located (Personal Computer or Card).

      Fig. 1 shows the functions location and the Interprocess
mechanisms which are described hereafter.  In this figure, the GRP is
represented as an OS/2* process running in an IBM PS/2*.

      As it can be seen, the PS/2 part of the memory (referred to as
Virtual Card) has the same logical interface as the Card part of the
memory:
   1. a QUEUE_IN (Qin) to place data to be routed to another
application;
   2. a QUEUE_OUT (Qout) to receive data routed from another
application.
   3. an interface composed of Programming Interface (PI) verbs,
written in C language to mask the supporting environment (Intel 186,
286, 386...)

      Using this set of resources, any application, located in the
PS/2 or in a Card, can enqueue data and dequeue data, the routing of
which is under the control of the GRP itself.

      The GRP is an OS/2 process which scans the Queues_In according
to a round-robin algorithm.  The sequence of scanning can be done by
systematic polling, with DosSleep pauses to let the system available
to the other applications.  Nevertheless, nothing will prevent an
interrupt based processing.

      The queues are created by the GRP:
      - at initialization time for the Queues_In (one Queue per Card,
and one queue for the PS/2 virtual card),
   - at user's application connection time for the Queues_Out (one
Queue per connected application).

      To use the GRP services, a user's application requires its
connection by submitting a CONNECT verb on the PI, passing its Name
and the required service_level.

      The GRP updates internal tables to relate the application_Name
with Queue_Out.  The nature of the physical interface to access this
Queue_Out is deducted from the Queue_In on which the CONNECT has been
received.

      To route data, the GRP uses the Queue_Out identifier which
uniquely represents a given application, as seen above, independentl...