Abstract: One embodiment provides a method for enabling class-based credit flow control for a network node in communication with a link partner using an Ethernet communications protocol. The method includes receiving a control frame from the link partner. The control frame includes at least one field for specifying credit for at least one traffic class and the credit is based on available space in a receive buffer associated with the at least one traffic class. The method further includes sending data packets to the link partner based on the credit, the data packets associated with the at least one traffic class.

Abstract: A method of compiling code includes assigning an endian type to data. An endian flip operation is performed based on the endian type of the data and a target system. Other embodiments are described and claimed.

Abstract: An embodiment of the invention includes code, such as a compiler, that enables byte order dependent code to execute on opposite byte order dependent architectures or systems. The compiler analyzes source code and produces diagnostic reports that indicate where source code changes are desirable to produce “endian neutral” source code versions that are compatible with opposite byte order dependent architectures or systems. Such source code changes may be desirable for code portions that will produce implicit byte order changes or byte order border crossings. The modified source code that is generated may include the semantics of the desired endian conversion, as opposed to generated executable code that includes proper endian formats but which may limit the architectures to which the code is applicable.

Abstract: In one embodiment of the invention code (e.g., compiler, tool) may generate information so a first code portion, which includes a pointer value in a first endian format (e.g., big endian), can be properly initialized and executed on a platform having a second endian format (e.g., little endian). Also, various embodiments of the invention may identify problematic regions of code (e.g., source code) where a particular byte order is cast away through void pointers.

Abstract: An embodiment of the invention includes code, such as a compiler, that enables byte order dependent code to execute on opposite byte order dependent architectures or systems. The compiler analyzes source code and produces diagnostic reports that indicate where source code changes are desirable to produce “endian neutral” source code versions that are compatible with opposite byte order dependent architectures or systems. Such source code changes may be desirable for code portions that will produce implicit byte order changes or byte order border crossings. The modified source code that is generated may include the semantics of the desired endian conversion, as opposed to generated executable code that includes proper endian formats but which may limit the architectures to which the code is applicable.

Abstract: In one embodiment of the invention code (e.g., compiler, tool) may generate information so a first code portion, which includes a pointer value in a first endian format (e.g., big endian), can be properly initialized and executed on a platform having a second endian format (e.g., little endian). Also, various embodiments of the invention may identify problematic regions of code (e.g., source code) where a particular byte order is cast away through void pointers.

Abstract: In a node to forward data on a switch fabric, a method that includes scheduling the forwarding of data associated with one of a plurality of traffic classes. The data is to be transmitted through one of a plurality of ports coupled to the switch fabric, each port to be associated with a queue to store data to be forwarded from that port. The scheduling is to include a two stage arbitration scheme. The first stage is to select one queue associated for each traffic class. The second stage is to select one queue from among the queues selected for each traffic class selected in the first stage.

Abstract: A method of compiling code includes assigning an endian type to data. An endian flip operation is performed based on the endian type of the data and a target system. Other embodiments are described and claimed.

Abstract: A method of compiling code includes assigning an endian type to data. An endian flip operation is performed based on the endian type of the data and a target system. Other embodiments are described and claimed.

Abstract: Method and apparatus to support efficient check-point and role-back operations for flow-controlled queues in network devices. The method and apparatus employ queue descriptors to manage transfer of data from corresponding queues in memory into a switch fabric. In one embodiment, each queue descriptor includes an enqueue pointer identifying a tail cell of a segment of data scheduled to be transferred from the queue, a schedule pointer identifying a head cell of the segment of data, and a commit pointer identifying a most recent cell in the segment of data to be successfully transmitted into the switch fabric. In another embodiment, the queue descriptor further includes a scheduler sequence number; and a committed sequence number that are employed in connection with transfers of data from queues containing multiple segments. The various pointers and sequence numbers are employed to facilitate efficient check-point and roll-back operations relating to unsuccessful transmissions into the switch fabric.

Abstract: In a node to forward data on a switch fabric, a method that includes scheduling the forwarding of data associated with one of a plurality of traffic classes. The data is to be transmitted through one of a plurality of ports coupled to the switch fabric, each port to be associated with a queue to store data to be forwarded from that port. The scheduling is to include a two stage arbitration scheme. The first stage is to select one queue associated for each traffic class. The second stage is to select one queue from among the queues selected for each traffic class selected in the first stage.

Abstract: A mechanism to process units of data associated with a dependent data stream using different threads of execution and a common data structure in memory. Accessing the common data structure in memory for the processing uses a single read operation and a single write operation. The folding of multiple read-modify-write memory operations in such a manner for multiple multi-threaded stages of processing includes controlling a first stage, which operates on the same data unit as a second stage to pass context state information to the second stage for coherency.

Abstract: Methods and apparatus, including computer program products, implementing techniques for monitoring a state of a device of a switched fabric network, the device including on-chip queues to store queue descriptors and a data buffer to store data packets, each queue descriptor having a corresponding data packet; detecting a first trigger condition to transition the device from a first state to a second state; and recovering space in the data buffer in response to the first trigger condition detecting, the recovering comprising selecting one or more of the on-chip queues for discard, and removing the data packets corresponding to queue descriptors in the selected one or more on-chip queues from the data buffer.

Abstract: Method and apparatus to support efficient check-point and role-back operations for flow-controlled queues in network devices. The method and apparatus employ queue descriptors to manage transfer of data from corresponding queues in memory into a switch fabric. In one embodiment, each queue descriptor includes an enqueue pointer identifying a tail cell of a segment of data scheduled to be transferred from the queue, a schedule pointer identifying a head cell of the segment of data, and a commit pointer identifying a most recent cell in the segment of data to be successfully transmitted into the switch fabric. In another embodiment, the queue descriptor further includes a scheduler sequence number; and a committed sequence number that are employed in connection with transfers of data from queues containing multiple segments. The various pointers and sequence numbers are employed to facilitate efficient check-point and roll-back operations relating to unsuccessful transmissions into the switch fabric.

Abstract: A method of compiling code includes assigning an endian type to data. An endian flip operation is performed based on the endian type of the data and a target system. Other embodiments are described and claimed.

Abstract: Method and apparatus to manage flow control for a network device are described.

Abstract: Method and apparatus to support expansion of compute engine code space by sharing adjacent control stores using interleaved addressing schemes. Instructions corresponding to an original instruction thread are partitioned into multiple interleaved sequences that are stored in respective control stores. During thread execution, instructions are retrieved from the control stores in a repeated order based on the interleaving scheme. For example, in one embodiment two compute engines share two control stores. Thus, instructions for a given thread are sequentially loaded from the control stores in an alternating manner. In another embodiment, four control stores are shared by four compute engines. In this case, the instructions in a thread are interleave using four stores, and each store is accessed every fourth instruction in the code sequence. Schemes are also provided for handling branching operations to maintain synchronized access to the control stores.

Abstract: A system and method for employing multiple hardware contexts and programming engines in a functional pipeline partitioned to facilitate high performance data processing. The system and method includes a parallel processor that assigns system functions for processing data including programming engines that support multiple contexts arranged to provide a functional pipeline by a functional pipeline control unit that passes functional data among the programming engines.

Abstract: A mechanism to process units of data associated with a dependent data stream using different threads of execution and a common data structure in memory. Accessing the common data structure in memory for the processing uses a single read operation and a single write operation. The folding of multiple read-modify-write memory operations in such a manner for multiple multi-threaded stages of processing includes controlling a first stage, which operates on the same data unit as a second stage to pass context state information to the second stage for coherency.

Abstract: In general, in one aspect, the disclosure describes a system to process packets received over a network. The system includes a receive process of at least one thread of a network processor to receive data of packets belonging to different flows. The system also includes a transmit process of at least one thread to transmit packets received by the receive process. A scheduler process of at least one thread populates at least one schedule of flow service based, at least in part, on quality of service characteristics associated with the different flows. The schedule identifies different flow candidates for service. The system also includes a shaper process of at least one thread to select from the candidate flows for service from the at least one schedule.