(19)
`
`(12)
`
`Europäisches Patentamt
`
`European Patent Office
`
`Office européen des brevets
`
`*EP000986885B1*
`EP 0 986 885 B1
`
`(11)
`
`EUROPEAN PATENT SPECIFICATION
`
`(45) Date of publication and mention
`of the grant of the patent:
`12.01.2005 Bulletin 2005/02
`
`(21) Application number: 98926168.0
`
`(22) Date of filing: 01.06.1998
`
`(51) Int Cl.7: H04L 29/06
`
`(86) International application number:
`PCT/US1998/011201
`
`(87) International publication number:
`WO 1998/056150 (10.12.1998 Gazette 1998/49)
`
`(54) SHARED USE OF A NETWORK INTERFACE CARD BETWEEN HETEROGENEOUS COMPUTER
`SYSTEMS
`
`VON HETEROGENEN RECHNERSYSTEMEN MITBENUTZUNG EINER
`NETZSCHNITTSTELLENKARTE
`
`UTILISATION PARTAGEE D’UNE CARTE D’INTERFACE RESEAU ENTRE SYSTEMES
`INFORMATIQUES HETEROGENES
`
`(84) Designated Contracting States:
`DE GB
`
`(30) Priority: 02.06.1997 US 48723 P
`
`(43) Date of publication of application:
`22.03.2000 Bulletin 2000/12
`
`(73) Proprietor: UNISYS CORPORATION
`Blue Bell, PA 19424-0001 (US)
`
`(72) Inventors:
`• JOHNSON, Robert, A.
`Pottstown, PA 19464 (US)
`
`• EBERSOLE, Dwayne, E.
`Lancaster, PA 17602-4139 (US)
`• DISNEY, William, W.
`Downingtown, PA 19335 (US)
`
`(74) Representative: Modiano, Guido, Dr.-Ing. et al
`Modiano, Josif, Pisanty & Staub,
`Baaderstrasse 3
`80469 München (DE)
`
`(56) References cited:
`WO-A-97/01944
`
`Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give
`notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in
`a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art.
`99(1) European Patent Convention).
`
`Printed by Jouve, 75001 PARIS (FR)
`
`EP0 986 885B1
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 1
`
`

`

`EP 0 986 885 B1
`
`Description
`
`BACKGROUND
`
`5
`
`Field of the Invention
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`[0001] The present invention relates to the field of computer networking, and, more particularly, to apparatus and
`methods for allowing two computer systems to access a network via a shared network interface card installed on one
`of them.
`
`Description of the Prior Art
`
`[0002] The ability for heterogeneous computer systems to communicate with each other over a network using stand-
`ard and/or proprietary networking protocols is known. Most computer systems have some form of networking archi-
`tecture that enables the computer system to perform networking in accordance with those protocols. Such a networking
`architecture typically comprises both system software and hardware. Figure 1 is a block diagram illustrating the com-
`ponents of a networking architecture employed by a Unisys A Series enterprise server 10 in order to communicate
`with other hosts. or nodes, on a network 15.
`[0003] The A Series enterprise server 10 executes the Unisys MCP operating system 12, and has an I/O subsystem
`that comprises one or more I/O Modules (IOM) 14 housed within the A Series chassis. The IOM 14 implements a
`Unisys proprietary I/O bus architecture referred to as CS-BUS II or CS-Bus III (hereinafter "the CS Bus"). A plurality
`of card slots, e.g. slots 16a-d, are provided for connecting interface cards, referred to as "channel adapters", into the
`CS Bus. Different groups, or racks, of channel adapter slots are each controlled by a Channel Manager Unit (CMU)
`(e.g., CMUs 18a, 18b). An IOM can contain several CMUs, each of which controls a different rack of channel adapter
`card slots via the CS-Bus. The CMUs manage the physical and data layers of the I/O process.
`[0004] Channel adapter cards, which each may occupy one or more channel adapter card slots within the IOM 14,
`provide various connectivity solutions for the A Series enterprise server 10. For example, Unisys provides a channel
`adapter card that implements the Small Computer System Interface(SCSI) protocol for connecting SCSI peripherals
`to the enterprise server.
`[0005] For network connectivity, Unisys provides several channel adapters to support various physical networking
`protocols. These channel adapters are generally referred to as network processors (NP). For example, Unisys ICP22
`and ICP26 network processors are channel adapter cards that implement the Ethernet network protocol and can be
`used to connect an A Series enterprise server to an Ethernet network. Unisys also provides network processors for
`connectivity to FDDI and ATM networks. As shown in Figure 1, a number of different network processors (e.g., NPs
`20a, 20b, and 20c) can be installed in respective channel adapter slots (e.g., slots 16b, 16c, and 16d) of the IOM 14,
`in order to provide different network connectivity solutions.
`[0006] As shown in the more detailed view of network processor 20c (installed in channel adapter slot 16d), a network
`processor may comprise a plurality of different lines, e.g., Line0, Line1 ... LineN. A line represents a physical endpoint
`within a network. For example, the Unisys ICP22 network processor has two lines, each of which comprises a separate
`Ethernet connection - one line could be connected to one Ethernet network, and the other to a different Ethernet network.
`[0007] Each line of a network processor can have one station group defined on that line. A station group consists of
`one or more stations. A station is a logical endpoint that represents a logical dialog on that line. Thus, more than one
`logical dialog can take place over a given line of a network processor. This is achieved through multiplexing. For
`example, with a connection-oriented networking protocol, such as the Burroughs Network Architecture - Version 2
`protocol (BNAv2), one station may represent a logical dialog with one other BNAv2 host on the network, whereas
`another station may represent a logical dialog to a different BNAv2 host. As illustrated in Figure 1, for example, Station0
`of LineN may represent a logical dialog with BNAv2 host 22, and Station1 of LineN may represent a logical dialog with
`BNAv2 host 24. For networking protocols that are not connection-oriented, like the Internet Protocol (IP), only one
`station needs to be defined to handle all communications for that protocol stack. For example, in Figure 1, StationN of
`LineN could be defined as the logical endpoint for all IP traffic over LineN. A Local Area Network Station Group (LANSG)
`module 26, which comprises software executing on the network processor 20c, provides callable procedures for cre-
`ating and maintaining stations and station groups on the various lines of the network processor 20d and for sending
`and receiving data over them.
`[0008] Other software components that execute on the network processor 20d include a Queue Service Provider
`(QSP) module 28, which handles the multiplexing and demultiplexing of data for all stations defined on a given NP,
`and two stub modules - a Network Services Manager stub (NSM-stub) 30 and a Link Layer Manager stub (LLM-stub)
`32 - which interface with corresponding modules of a Core Network Services (CNS) software component 34, to and
`from modules within the MCP environment.
`
`2
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 2
`
`

`

`EP 0 986 885 B1
`
`[0009] Generally, a network processor (e.g., NP 20a, 20b, or 20c) implements the data link and physical layers of
`the 7-layer ISO Reference Model. Higher level networking protocols that a client application 46 may wish to employ in
`order to communicate with applications running on different hosts of the network 15, such as the BNAv2 and TCP/IP
`networking protocols, are implemented as network protocol providers on the A Series system 10. A network protocol
`provider is a software module that implements these higher level networking protocols. For example, Unisys provides
`both BNAv2 Host Resident Network Provider (HRNP) modules and TCP/IP HRNP modules. In the example of Figure
`1, a BNAv2 HRNP 42 and a TCP/IP HRNP 44 are shown.
`[0010] The Core Network Services (CNS) software 34 provides support for the network protocol providers 42, 44
`and handles the initialization and maintenance of network processors and the station groups defined thereon. Specif-
`ically, CNS 34 comprises a Network Services Manager (NSM) 36 that initializes and manages the network processors
`(e.g., 20a, 20b, 20c) installed in the system, and a Link Layer Manager (LLM) 38 that initializes and maintains the
`identity and attributes of each station group defined on a given network processor. Another component (not shown) of
`CNS 34 validates attributes associated with station groups and stations created on a network processor. These at-
`tributes are passed between the network processor and CNS 34 via a control dialog when the stations arc defined.
`Like the stub procedures for the NSM and LLM modules 36, 38, network processors also have a stub procedure (LLAH,
`not shown) that corresponds to the attribute handler of CNS 34. An NPSUPPORT software library 40, as well as portions
`of the MCP operating system 12, provide routines and procedure calls that serve as an interface between a network
`processor and the CNS 34 and network protocol providers 42, 44, and control loading of software to the NPs and
`dumping of their state.
`[0011] Each network processor has an associated identifier that uniquely identifies that network processor within the
`system 10. When a network processor is initialized and brought on-line, the NSM-stub 30 in the network processor
`interfaces with the NSM 36 of CNS 34 via a control dialog in order to pass its identifier to the NSM 36. The NSM 36
`manages the identifiers of all active network processors.
`[0012] Each station group and station defined for a given network processor also has a unique identifier associated
`with it. Via a control dialog established between the LLM-stub 32 on the network processor and the LLM 38 of CNS
`34, the station and station group identifiers are passed to the LLM 38 during initialization. Within the LLM 38, a station
`corresponds to a connection, and a station group corresponds to a connection group.
`[0013] As mentioned above, the ability to define multiple stations (i.e., a station group) on a single physical line of a
`network processor is achieved through multiplexing. Specifically, the QSP 28 in the network processor multiplexes
`inbound and outbound data for multiple stations on a given line. Moreover, the QSP is responsible for distributing
`request and response data between the NSM 36 and NSM-stub 30 and between the LLM 38 and LLM-stub 32. To that
`end, each entity on the network processor that receives outbound data from the MCP, including every station, the NSM-
`stub 30, and the LLM-stub 32, is assigned a unique Remote Queue Reference (RQR) by the QSP. The NSM-stub RQR
`is reported to the NSM 36 within CNS 34 via NPSUPPORT when the NP is loaded. The LLM-stub RQR is reported to
`the LLM 38 via the NSM 36 by the NSM-stub 30 when the NP initializes. All of the station RQRs are reported to the
`HRNPs as the stations open.
`[0014] When a client application is required to send data via network 15 to some other host or node on the network
`15, such as another BNAv2 Host or another TCP/IP host, it invokes the services of the appropriate network protocol
`provider, e.g., 42, 44. The network protocol provider 42, 44 determines the appropriate network processor and station
`on which the data is to be output, adds protocol headers, and makes a corresponding request to the MCP 12 that
`includes the identifier of the network processor and the RQR of the station. The data and associated RQR are passed
`from the MCP 12 to the QSP 28 on the network processor (e.g., network processor 20c), which, in combination with
`the LANSG module 26, sends the data out to the network 15 via the appropriate line (e.g., Line0, Line1, ... or LineN)
`as part of the logical dialog represented by the designated station.
`[0015] When data is received from the network 15 on a given line, the LANSG module 26 determines, from header
`information associated with the data, the station (i.e. logical dialog) for which the data is intended. The LANSG and
`QSP modules 26, 28, in combination with portions of the MCP 12 and NPSUPPORT library 40, pass the received data
`to the appropriate network protocol provider 42, 44 associated with that station, along with an indication of which station
`received the data. For example, one of the stations on LineN of the network processor 20c of Figure 1 (e.g., station0)
`may be defined as the logical endpoint for the BNAv2 HRNP 42, while a different station (e.g., station1) may be defined
`as the logical endpoint on which all IP traffic over LineN is received for the TCP/IP HRNP 44. When a frame of data is
`received from the network on LineN, the LANSG module 26 determines from header information which of the network
`protocol providers (i.e., stations) is intended to receive the data. This determination is performed in accordance with
`the methods described in commonly assigned, U.S. Patent No. 5,379,296, entitled "Method and Apparatus for Inter-
`facing a Workstation to a Plurality of Computer Platforms" (Johnson et al.).
`[0016]
`In addition to its use in A Series computers, the foregoing networking architecture is also employed in Unisys
`ClearPath HMP NX enterprise servers. A ClearPath HMP NX server comprises an A Series enterprise server tightly
`integrated with a server running Microsoft Window NT. Please note that "Microsoft," "Windows," and "Windows NT"
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`3
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 3
`
`

`

`EP 0 986 885 B1
`
`are registered trademarks of Microsoft Corporation. Additional information concerning the foregoing networking archi-
`tecture can be found in the following documents, each of which is available from Unisys Corporation, assignee of the
`present invention, and each of which is hereby incorporated by reference in its entirety:
`
`ClearPath HMP NX Series with Windows NT Network Services Implementation Guide (Part No. 4198 6670);
`BNA/CNS Network Implementation Guide, Volume 2: Configuration (Part No. 3789 7014);
`ClearPath HMP NX Series with Windows NT Implementations and Operations Guide (Part No. 8807 6542);
`ClearPath HMP NX Series with Windows NT Migration Guide (Part No. 8807 7730);
`Networking Capabilities Overview (Part No. 3789 7139)
`Networking Operations Reference Manual, Volumes 1 and 2: Commands and Inquiries (Part No. 3787 7917); and
`Networking Products Installation Guide (Part No. 4198 4840).
`
`[0017] Using a Unisys ICP22 network processor, which is an Ethernet-based channel adapter, it has been possible
`in the past for a Unisys A Series enterprise server to communicate with a workstation or personal computer (PC) over
`a network. An example of this ability is illustrated in Figure 2. In this example, the A Series enterprise server 10 com-
`municates with an Intel-based workstation 48 running the Microsoft Windows NT operating system (hereinafter "the
`NT server"). The A Series enterprise server is connected to the network via network processor 20a. which may, for
`example, be a Unisys ICP22 Ethernet-based network processor.
`[0018] The I/O subsystem of the NT server 48 comprises portions of the NT operating system kernel, an EISA or
`PCI bus 52, and appropriate device driver software. To provide network connectivity, a network interface card (NIC)
`50 is installed in an available bus slot on the NT server 48. The NT server may support one or both of the PCI and
`EISA bus standards. NICs are available for both bus standards.
`[0019] A NIC device driver 54 that typically is sold with the NIC card 50 is installed in the kernel space of the NT
`operating system. The NIC device driver 54 interfaces with a higher level network protocol provider, such as an imple-
`mentation of the TCP/IP protocol. Microsoft Corporation provides an implementation of the TCP/IP protocol in the form
`of a kernel level device driver, also referred to as a transport protocol driver, named TCPIP.SYS 58. TCPIP.SYS 58
`interfaces with the NIC device driver 54 via NDIS, an industry standard Network Driver Inferface Specification jointly
`developed by Microsoft and 3Com. NDIS defines an interface for communication between hardware-independent pro-
`tocol drivers, such as TCPIP.SYS, which implement the Data Link, Network, and Transport layers of the OSI model,
`and hardware-dependent NIC drivers which provide an interface to the NIC hardware and which correspond to the
`Physical Layer of the OSI model. A client program 60 on the NT server can communicate over the network 15 in
`accordance with the TCP/IP protocol by issuing suitable calls via the NT operating system to the TCPIP SYS protocol
`driver.
`[0020] Network interface cards and associated device drivers for NT servers are available from a number of Original
`Equipment Manufactures (OEMs). OEM NICs are available at relatively low cost for a variety of different network media
`standards, including Ethernet, Fast-Ethernet, etc. As new network standards evolve, OEMs are quick to design and
`produce NICs to support these standards. Because these NICs are developed for industry standard I/O bus architec-
`tures, such as EISA and PCI, which are found in the many computer systems today, the economies of scale result in
`fast cycle development times and extremely low prices for consumers.
`[0021] On the contrary, it takes significantly longer and costs significantly more to design and produce a new network
`processor for a proprietary bus architecture, such as the CS-BUS II architecture of Unisys A Series enterprise servers.
`Vendors of proprietary systems cannot achieve the same economies of scale as the open system NIC vendors, and
`network processors, or NIC cards, for proprietary systems therefore typically cost significantly more than their open
`systems counterparts. To avoid the costs associated with the development of NIC cards for proprietary systems, it
`would be desirable if it were possible for two computer systems to each access a network via a shared network interface
`card installed on one of the systems. The present invention provides such a capability.
`[0022] WO 97 01944 describes a mechanism that enables emulated environments within a host computer system
`to share a single TCP/IP protocol stack with the host system, but it does not teach or suggest a mechanism that enables
`two host environments, each having its own protocol stack; to share a single network interface card located in one of
`the systems. The mechanism described in WO 97 01944 does not address the same problem as the present invention.
`
`SUMMARY OF THE INVENTION
`
`[0023] The present invention is directed to methods and apparatus that enable a first network protocol provider,
`executing on a first computer system and having a first network address associated therewith, and a second network
`protocol provider, executing on a second computer system and having a second network address associated therewith,
`to both send and receive data over a network via a same network interface card installed on the second computer
`system. In accordance with a preferred embodiment thereof, the present invention comprises an interconnection that
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`4
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 4
`
`

`

`EP 0 986 885 B1
`
`couples the input/output (I/O) subsystem of the first computer system to the I/O subsystem of the second computer
`system and over which data can be transmitted between systems, and a router executing on the second computer
`system that routes data among the first network protocol provider (via the interconnection), the second network protocol
`provider. and the network interface card in a manner that enables the network interface card to be shared between the
`first and second network protocol providers. Specifically, when data is received from the network by the network inter-
`face card, the router routes the data to the first network protocol provider, via the interconnection, when a network
`address received with the data matches the first network address, but passes the received data to the second network
`protocol provider when the network address received with the data matches the second network address. Data received
`from the network with a broadcast address is routed to both the first (via the interconnection) and second network
`protocol providers. Outgoing data from each network protocol provider is routed to the network via that same network
`interface card. However, when outgoing data from one of the first and second network protocol providers is addressed
`to the other network protocol provider, a "loopback" capability enables the router to route that data directly from one
`network protocol provider to the other, bypassing the network interface card. Also, outgoing data from one of the first
`and second network protocol providers that has a broadcast address is routed to the network via the network interface
`card, and also to the other network protocol provider via the loopback path (bypassing the network interface card). In
`all cases, data routed between the first and second computer systems is transmitted over the interconnection. Addi-
`tionally, it is understood that the present invention can be used to route data to and from a plurality of network protocol
`providers on the first computer system and/or a plurality of network protocol providers on the second system, via the
`same network interface card.
`[0024] Preferably, the router simulates the operation of a network protocol provider when interfacing with a device
`driver of the network interface card, and the router simulates the operation of a device driver of the network interface
`card when interfacing with the first and second network protocol providers, whereby the functionality of the router is
`transparent to the network interface card and to the first and second network protocol providers.
`[0025] The interconnection between the I/O subsystem of the first computer system and the I/O subsystem of the
`second computer system preferably comprises a physical connection between the I/O subsystems over which data
`can be transmitted between them, and an interconnection device driver on the second computer system that controls
`access by the second computer system to the physical connection. The interface between the interconnection device
`driver and other components on the second computer system is preferably implemented in the form of a procedure
`registration mechanism. In this manner, different interconnection device drivers can be installed on the second computer
`system for different physical connections, in a manner that is transparent to the other components of the present
`invention. For example, when the first and second computer systems are separate physical units, the physical con-
`nection may comprise suitable hardware (e.g., interface boards) installed in available slots of the I/O buses of each
`system and a cable that provide a connection between them. Alternatively, where the first computer system is emulated
`within the second system, the physical connection may be emulated within the second system in the form of a memory-
`to-memory connection.
`[0026] One advantage of the present invention is that it can be used to enable a computer system that has a propri-
`etary I/O subsystem to access a network via a standard network interface card developed for a more open, or widely
`used I/O subsystem such as that found in systems that support the Microsoft Windows NT operating system. This
`avoids the need to develop network interface cards for the proprietary system as new networking protocols and stand-
`ards evolve. Instead, the present invention allows the proprietary system to share network interface cards designed
`for the open system.
`[0027] Additional features and advantages of the present invention will become evident hereinafter.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`[0028] The foregoing summary, as well as the following detailed description of the preferred embodiment, is better
`understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there
`is shown in the drawings an embodiment that is presently preferred, it being understood, however, that the invention
`is not limited to the specific methods and instrumentalities disclosed. In the drawings:
`
`Figure 1 is a block diagram illustrating the components of a prior art networking architecture employed by Unisys
`A Series enterprise servers in order to communicate with other hosts, or nodes, on a network;
`Figure 2 is a block diagram showing a prior art method by which a Unisys A Series enterprise server can commu-
`nicate via a network with a server running Microsoft Windows NT;
`Figure 3 is a block diagram illustrating one embodiment of apparatus that enables two computer systems to share
`a single network interface card installed on one of them, in accordance with the present invention;
`Figure 4 is a block diagram illustrating an alternative embodiment of an interconnection of the apparatus of Figure 3;
`Figure 5 is a block diagram illustrating yet another embodiment of the interconnection of the apparatus of Figure 3;
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`5
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 5
`
`

`

`EP 0 986 885 B1
`
`Figures 6A-F are flow diagrams that further illustrate the operation of the interconnections illustrated in Figures 3-5;
`Figure 7 is a diagram illustrating the different data paths over which the router of the present invention may route
`incoming and outgoing network data;
`Figure 8 is a block diagram illustrating further details of an implementation of the router of the present invention,
`in accordance with a preferred embodiment;
`Figures 9A-9C comprise a pseudocode listing illustrating further details of one embodiment of the routing func-
`tionality of the router of the present invention and also one embodiment of a method of the present invention;
`Figures 10A-D are flow diagrams illustrating further details of another embodiment of the routing functionality of
`the router of the present invention, and also another embodiment of the method of the present invention.
`
`DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
`
`[0029] The present invention is directed to methods and apparatus that enable a first network protocol provider,
`executing on a first computer system and having a first network address associated therewith, and a second network
`protocol provider, executing on a second computer system and having a second network address associated therewith,
`to both send and receive data over a network via a same network interface card installed on the second computer
`system. In accordance with a preferred embodiment thereof, the present invention comprises an interconnection that
`couples the input/output (I/O) subsystem of the first computer system to the I/O subsystem of the second computer
`system and over which data can be transmitted between systems, and a router executing on the second computer
`system that and a router executing on the second computer system that routes data among the first network protocol
`provider (via the interconnection), the second network protocol provider. and the network interface card in a manner
`that enables the network interface card to be shared between the first and second network protocol providers.
`[0030]
`In one embodiment, described more fully hereinafter, the methods and apparatus of the present invention
`may be implemented as part of a Cooperative Networking Platform (CNP) (sometimes also referred to "NX/Network
`Services" or "NNS") provided as a feature of Unisys ClearPath HMP NX enterprise servers, in which, as mentioned
`above, a Unisys A Series enterprise server is tightly integrated with an Intel-based server running Microsoft Window
`NT. In that embodiment, the A Series enterprise server comprises the first computer system and the NT server com-
`prises the second computer system. As embodied in that environment, the present invention allows a network protocol
`provider (e.g., BNAv2 HRNP or TCP/IP HRNP) on the A Series server to send and receive data over a network via a
`standard network interface card installed in the NT server. A network protocol provider (e.g., TCPIP.SYS) on the NT
`server can also access the network via that NIC, and thus, the NIC is shared by both systems. This capability is
`advantageous because NICs for NT servers are widely available at relatively low cost. The need to develop proprietary
`NICs (i.e.. network processors) for the A Series server, and the significant costs associated therewith, can be avoided.
`[0031] The methods and apparatus of the present invention, or certain aspects or portions thereof, may take the
`form of program code (i.e., instructions) embodied in tangible media, such as floppy diskettes, CD-ROMs, hard drives,
`or any other machine-readable storage medium, wherein, when the program code is loaded into and executed by a
`machine, such as a computer, the machine becomes an apparatus for practicing the invention. The methods and
`apparatus of the present invention may also be embodied in the form of program code that is transmitted over some
`transmission medium, such as over electrical wiring or cabling, through fiber optics, or via any other form of transmis-
`sion, wherein, when the program code is received and loaded into and executed by a machine, such as a computer,
`the machine becomes an apparatus for practicing the invention. When implemented on a general purpose processor,
`the program code combines with the processor to provide a unique apparatus that operates analogously to specific
`logic circuits.
`[0032] Referring now to the drawings, wherein like numerals represent like elements throughout, Figure 3 is a block
`diagram illustrating one embodiment of the present invention, in which the methods and apparatus of the present
`invention are implemented as part of a Cooperative Networking Platform (CNP) deployed on a Unisys ClearPath HMP
`NX computer system ("the ClearPath system"). As shown, the ClearPath system comprises a Unisys A Series enter-
`prise server 100 and an Intel-based server 102 running Windows NT 102 ("the NT server"). In this embodiment, the A
`Series enterprise server 100 defines a first computer system and the NT server 102 defines a second computer system.
`A first network protocol provider 44 is provided on the A Series system 100, in this case a TCP/IP HRNP, and it has a
`network address (i.e., IP address) associated with it. A second network protocol provider 58 is provided on the NT
`server 102, in this case TCPIP.SYS (available from Microsoft Corporation), and it has its own network address (i.e.,
`IP address) associated with it that defines a second network address in this embodiment. Other network protocol
`providers may be installed on the A Series and NT servers as well. For example, on the A Series server, a BNAv2
`HRNP 42 may be provided. Note, however, that because the BNAv2 protocol is a Unisys proprietary protocol, the
`BNAv2 HRNP 42 does not have an IP address associated with it. A network interface card (NIC) 50 is installed in a
`slot of the I/O bus (e.g., EISA or PCI) of the NT server 102. In this embodiment, any LAN-type NIC that is compatible
`with Windows NT can be employed. Preferably, the NIC supports the Fast-Ethernet networking protocol (e.g., 100Base-
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`6
`
`Petitioners Microsoft Corporation and HP Inc. - Ex. 1023, p. 6
`
`

`

`EP 0 986 885 B1
`
`T). NICs of this type are available from numerous vendors and original equipment manufacturers (OEMs). NICs sup-
`porting other physical media types, such as Ethernet/802.3, FDDI, or Gigabit Ethernet, can alternatively be employed.
`Typically, a NIC vendor will supply a device driver with the NIC, which is installed in the kernel space of the operating
`system so that other entities on the system can access the networking functionality of the NIC. The NIC 50 of the
`exemplary system of Figure 3 has a device driver 54 ("<nicdrv>.sys") that is installed in the Windows NT kernel space,
`as shown.
`[0033] According to the present invention, apparatus and methods are provided that enable both the first and second
`network protocol providers 44, 58 to access the network 15 via the same NIC 50 installed on the second computer
`system 102. That is, the present invention enables the NIC 50 to be shared between the two systems 100, 102.
`[0034] Apparatus of the present invention comprises an interconnection that couples the I/O subsystem of the A
`Series server 100 to the I/O subsystem of the NT server 102 so that data can be transmitted between the two servers,
`and a router 80 that executes on the second computer system (i.e., the NT server 102).