Abstract: A message routing method includes receiving a plurality of messages at a routing node. The routing node is configured to receive instructions indicating an offset, criteria associated with the offset, and a group of interested nodes. The routing node examines a portion of a data payload of a received message based on the offset. If the information at the offset matches the criteria, the routing node routes the message to each of the group of interested nodes. Thus, the routing node can route messages to different groups of destination nodes depending on information in the data payload of received messages, thereby providing a flexible way to route messages over a network.

Abstract: A message routing method includes receiving a plurality of messages at a routing node. The routing node is configured to route each received message based on the message type. Accordingly, the routing node is able to establish a peer-to-peer connection between the message source and a destination node for a designated message type, while establishing a server-client connection between the message source and a server for other message types. The routing node can also route messages to different groups of destination nodes depending on the message type, thereby providing a flexible way to route messages over a network.

Abstract: A method of securing network authentication information at a data processing device includes determining a boot source from which to boot the device and comparing the boot source to an expected source. If the boot source is not the expected source, access to the network authentication information is inhibited, such as by disabling access to the portion of memory that stores the authentication information. Further, if the boot source is the expected source, boot code authentication information is retrieved from memory and verified during the boot sequence. If the device authentication information is not authenticated, access to the network authentication information is inhibited. Accordingly, access to the network authentication information is allowed only if the data processing device is booted from an expected source, and only if the boot code is authenticated, thereby reducing the likelihood of unauthorized access to the network authentication information.

Abstract: A method for diagnosing and correcting errors at a data processing system is disclosed includes detecting at a first device of the system, such as a network interface device, an error at a second device of the system, such as a data processor. In response to detecting the error, the first device communicates a help request via a network. In response to the help request, the first device receives diagnostic and error correction routines from a remote system. The first device executes the routines and provides information to the remote system to diagnose and correct errors at the second device.

Abstract: An interceptor program at a computer device intercepts certain externally initiated communications intended for the peer program before those communications are placed on the network stack. Further, the interceptor program stores the received communications in a buffer in application space, rather than kernel space. The interceptor program can then intercept polling requests from the peer program inquiring if data for the peer program is located at the network stack, and provide the buffered communications to the peer program in response. Because the communications are stored in application space, rather than kernel space, the data can be rapidly provided to the peer program, improving the speed with which the program can respond to the client communications.

Abstract: A method of communicating with a network interface includes providing a packet to the network interface, where the packet includes an address field indicating a destination of the packet. The network interface analyzes the address field, and determines if it reflects an address associated with the network interface. If not, the network interface provides the packet to a network. If the network interface determines the address field reflects an address associated with the interface, it provides information in the packet to an application executing at the network interface. Accordingly, information targeted to an application can be communicated by associating an address, such as a network address, with the network interface, allowing for communication of the information without extensive processing of each packet at the interface.

Abstract: A method for receiving a first indication from a host device is comprised. The device is a first port of the host device that has been reserved fro a network device. The method also poses as the host device at the network device in response to a receipt of the first indication.

Abstract: A method for diagnosing and correcting errors at a data processing system is disclosed includes detecting at a first device of the system, such as a network interface device, an error at a second device of the system, such as a data processor. In response to detecting the error, the first device communicates a help request via a network. In response to the help request, the first device receives diagnostic and error correction routines from a remote system. The first device executes the routines and provides information to the remote system to diagnose and correct errors at the second device.

Abstract: A method of managing communication of messages via a network includes storing messages received from an application in a queue at a transport layer. The transport layer monitors the amount of information stored at the queue. When the amount of information exceeds a threshold amount, the transport layer provides a warning message to the application. In an embodiment, the transport layer provides the warning message before the queue assigned to store messages from the application is full, so that the application can continue to provide messages to the transport layer after the warning message. The application can take appropriate action in response to the message, such as reducing the amount or frequency of information provided at the transport layer, thereby reducing the likelihood of a communication bottleneck at the transport layer.

Abstract: A method for diagnosing and correcting errors at a data processing system is disclosed includes detecting at a first device of the system, such as a network interface device, an error at a second device of the system, such as a data processor. In response to detecting the error, the first device communicates a help request via a network. In response to the help request, the first device receives diagnostic and error correction routines from a remote system. The first device executes the routines and provides information to the remote system to diagnose and correct errors at the second device.

Abstract: A method of authenticating a data processing device includes receiving a request to authenticate the data processing device. In response, an authentication key is accessed an authenticated at an authentication module. The authentication key is stored at a storage module that is located within the same integrated circuit package as the authentication module, so that the authentication key can be communicated to the module without exposing the key to unauthorized probing. The integrated circuit package also includes a tamper detection module to determine whether a memory of the data processing device has been accessed. In response to determining the memory has been accessed, the tamper detection module instructs the authentication module to not authenticate the data processing device.

Abstract: A method of transforming messages for communication via a network includes receiving a message from an application being executed at a processor. The message is received at a connection object, which transforms the received message based on characteristics of the connection object. The characteristics can be determined via a negotiation between the source and target of the message. The connection object provides the transformed message to a protocol layer, which forms packets based on the message. The connection object thus transforms each message independent of the transmission protocol used to communicate the message to the target, thereby improving communication bandwidth and efficiency.

Abstract: Techniques are disclosed for setting network communication parameters at a network interface based on the types of applications being executed at one or more computer devices that interface with the network. Thus, for example, the network interface can set the communication bandwidth, priority, or combination thereof, for each executing application based on the application type. By setting the network communication parameters for each application based on the application type, the applications can communicate with the network more efficiently.

Abstract: Securing network authentication information at a data processing device includes determining a boot source from which to boot the device and comparing the boot source to an expected source. If the boot source is not the expected source, access to the network authentication information is inhibited, such as by disabling access to the portion of memory that stores the network authentication information. Further, if the boot source is the expected source, boot code authentication information is retrieved from memory and verified during the boot sequence. If the boot code authentication information is not authenticated, access to the network authentication information is inhibited. Accordingly, access to the network authentication information is allowed only if the data processing device is booted from an expected source, and only if the boot code is authenticated, thereby reducing the likelihood of unauthorized access to the network authentication information.

Abstract: A message routing method includes receiving a plurality of messages at a routing node. The routing node is configured to receive instructions indicating an offset, criteria associated with the offset, and a group of interested nodes. The routing node examines a portion of a data payload of a received message based on the offset. If the information at the offset matches the criteria, the routing node routes the message to each of the group of interested nodes. Thus, the routing node can route messages to different groups of destination nodes depending on information in the data payload of received messages, thereby providing a flexible way to route messages over a network.

Abstract: A message routing method includes receiving a plurality of messages at a routing node. The routing node is configured to receive instructions indicating an offset, criteria associated with the offset, and a group of interested nodes. The routing node examines a portion of a data payload of a received message based on the offset. If the information at the offset matches the criteria, the routing node routes the message to each of the group of interested nodes. Thus, the routing node can route messages to different groups of destination nodes depending on information in the data payload of received messages, thereby providing a flexible way to route messages over a network.

Abstract: A method is disclosed that receives a function call at an application program interface. The method selects a first processor of a plurality of processors to execute the function call. The method further executes a first transmit function associated with a bus coupled to the first processor. The first transmit function includes a function parameter associated with the function call.

Abstract: In general, in one aspect, the disclosure describes a method of searching for an address within a collection of addresses. The method includes identifying a section storing an ordered sub-set of a collection of addresses. The identifying is based on a comparison of the address with at least one section boundary value. The method also includes searching the identified section for a match for the address.

Abstract: The state of the communication socket is maintained at a network device and communicated to a device driver for the network device. The device driver can send messages to an interface program in application space, which stores the state information in dedicated application space memory. In response to a query from a peer program requesting the state of the communication socket, the interface program retrieves the state information from the dedicated memory. Because the state information and the interface program are located in application space, the number of kernel transitions required to determine the socket state is reduced, improving communication efficiency.

Abstract: Provided are a method, system, and program for managing data transmission from a source to a destination through a network. The destination imposes a window value on the source which limits the quantity of data packets which can be sent from the source to the destination without receiving an acknowledgment of being received by the destination. In one embodiment, the source imposes a second window value, smaller than the destination window value, which limits even further the quantity of data packets which can be sent from the source to the destination without receiving an acknowledgment of being received by the destination. In another embodiment, a plurality of direct memory access connections are established between the source and a plurality of specified memory locations of a plurality of destinations.