Abstract: A network interface device includes a bus interface that communicates over a bus with a host processor and memory, and a network interface that sends and receive data packets carrying data over a packet network. A protocol processor conveys the data between the network interface and the memory via the bus interface while performing protocol offload processing on the data packets in accordance with multiple different application flows. The bus interface queues the data for transmission over the bus in a plurality of queues that are respectively assigned to the different application flows, and transmits the data over the bus according to the queues.

Abstract: A method for communication is disclosed and may include, in a network interface device, parsing a portion of a TCP segment into one or more portions of Internet Small Computer Systems Interface (iSCSI) Protocol Data Units (PDUs). A header and/or a payload for one or more of the parsed iSCSI PDUs may be recovered. Concurrent with parsing of a remaining portion of the TCP segment to recover a remaining portion of PDUs, the recovered header may be evaluated and/or the recovered payload may be routed external to the network interface device for processing. The evaluating and the routing may occur independently of the parsing within the network interface device. Respective separate physical processors may be used for the parsing and the recovering. The respective separate processors for recovering may be used for the evaluating and the routing.

Abstract: Certain aspects of a method and system for transparent transmission control protocol (TCP) offload with best effort direct placement of incoming traffic are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) processor without transferring state information to a host processor every time a TCP segment is received. When an event occurs that terminates the collection of TCP segments, the NIC processor may generate a new aggregated TCP segment based on the collected TCP segments. If a placement sequence number corresponding to the generated new TCP segment for the particular network flow is received before the TCP segment is received, the generated new TCP segment may be transferred directly from the memory to the user buffer instead of transferring the data to a kernel buffer, which would require further copy by the host stack from kernel buffer to user buffer.

Abstract: A network controller in a communication device may be operable to provide pass-through communication of local host-management traffic between a local host and a management controller within the communication device, wherein the local host may be operable to utilize its network processing resources during communication of the local host-management traffic. The network controller may use packet filtering to provide the pass-through communication, wherein the network controller may utilize a plurality filtering rules during filtering of packets received in the network controller. The filtering rules may specify packet processing and/or forwarding actions by said network controller based on one or more specified conditions. The specified conditions may based on one or more match criteria; wherein the match criteria comprising source address, destination address, and/or traffic type data in the received packets.

Abstract: Certain aspects of a method and system for transparent transmission control protocol (TCP) offload are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) processor without transferring state information to a host system. The collected TCP segments may be buffered in a coalescer. The coalescer may verify that the network flow associated with the collected TCP segments has an entry in a flow lookup table (FLT). When the FLT is full, the coalescer may close a current entry and assign the network flow to the available entry. The coalescer may also update information in the FLT. When an event occurs that terminates the collection of TCP segments, the coalescer may generate a single aggregated TCP segment based on the collected TCP segments. The aggregated TCP segment and state information may be communicated to the host system for processing.

Abstract: Certain aspects of a method and system for transparent transmission control protocol (TCP) offload with best effort direct placement of incoming traffic are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) processor without transferring state information to a host processor every time a TCP segment is received. When an event occurs that terminates the collection of TCP segments, the NIC processor may generate a new aggregated TCP segment based on the collected TCP segments. If a placement sequence number corresponding to the generated new TCP segment for the particular network flow is received before the TCP segment is received, the generated new TCP segment may be transferred directly from the memory to the user buffer instead of transferring the data to a kernel buffer, which would require further copy by the host stack from kernel buffer to user buffer.

Abstract: A method for communication includes inputting from a host processor to a network interface device a sequence of work requests indicative of operations to be carried out by the network interface device with respect to a plurality of the connections. The device looks ahead through the sequence in order to identify at least first and second operations that are to be carried out with respect to one of the connections in response to first and second work requests, respectively, wherein the second work request does not immediately follow the first work request in the sequence. The device loads the context data for the one of the connections from a host memory into a context cache, and performs at least the first and second operations sequentially while the context data are held in the cache.

Abstract: Certain aspects of a method and system for transparent transmission control protocol (TCP) offload are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) processor without transferring state information to a host system. The collected TCP segments may be buffered in a coalescer. The coalescer may verify that the network flow associated with the collected TCP segments has an entry in a flow lookup table (FLT). When the FLT is full, the coalescer may close a current entry and assign the network flow to the available entry. The coalescer may also update information in the FLT. When an event occurs that terminates the collection of TCP segments, the coalescer may generate a single aggregated TCP segment based on the collected TCP segments. The aggregated TCP segment and state information may be communicated to the host system for processing.

Abstract: Certain aspects of a method and system for transparent TCP offload with a user space library are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) without transferring state information to a host system. When an event occurs that terminates the collection of TCP segments, a single aggregated TCP segment based on the collected TCP segments may be generated. The aggregated TCP segment may be posted directly to a user space library, bypassing kernel processing of the aggregated TCP segment.

Abstract: Apparatus for receiving a sequence of Transmission Control Protocol (TCP) segments, including a parsing machine which is adapted to parse at least one TCP segment so as to recover an Internet Small Computer Systems Interface Protocol Data Unit (iSCSI PDU), the PDU including a header and at least part of a payload. The apparatus further includes at least one analysis machine which is adapted to receive and evaluate the header and to receive and route the at least part of the payload for the iSCSI PDU, the parsing machine and the at least one analysis machine operating substantially autonomously.

Abstract: A method for cache management includes assigning a respective cache line in a cache of a processing device to each of a plurality of processing flows in the processing device, and loading respective context data relating to each of the processing flows from a memory into the respective cache line that is assigned thereto. Respective activity levels of the processing flows are monitored. Responsively to detecting an absence of activity of a processing flow, and prior to receiving a request to overwrite the cache line, the context data are written back to the memory from the respective cache line that is assigned to the processing flow.

Abstract: Methods and systems for latency optimized ATS usage are disclosed. Aspects of one method may include communicating a memory access request using an untranslated address and also an address translation request using the same untranslated address, where the translation request may be sent without waiting for a result of the memory access request. The memory access request and the address translation request may be made in either order. A translation agent may be used to translate the untranslated address, and the translated address may be communicated to the device that made the memory access request. The translated address may also be used to make the memory access. Accordingly, by communicating the translated address without having to wait for completion of the memory access, or vice versa, the requesting device may reduce latency for memory accesses when using untranslated addresses.

Abstract: Certain aspects of a method and system for transparent TCP offload with a user space library are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) without transferring state information to a host system. When an event occurs that terminates the collection of TCP segments, a single aggregated TCP segment based on the collected TCP segments may be generated. The aggregated TCP segment may be posted directly to a user space library, bypassing kernel processing of the aggregated TCP segment.

Abstract: An off-load engine for processing a packet conveyed between a target and a network over a transport connection, the packet including a payload and a header. The engine includes a payload buffer, for holding data exchanged between the off-load engine, the network, and the target for inclusion in the payload, and a packet processor, for processing context of the transport connection. The engine also includes a context buffer, for holding the context processed by the packet processor, a memory access multiplexer, which is coupled to convey the data in the payload buffer and the context in the context buffer to and from a memory that stores both the data and the context, and an arbiter, which controls the multiplexer by arbitrating among payload requests to convey the data between the payload buffer and the memory and context requests to convey the context between the context buffer and the memory.

Abstract: Certain aspects of a method and system for transparent transmission control protocol (TCP) offload with transmit and receive coupling are disclosed. Aspects of a method may include collecting at least one received TCP segment for a determined network flow via a network interface card (NIC) processor. The state information for the received TCP segment and state information for transmitted TCP segments for the determined network flow may be stored at the NIC without transferring state information for the received TCP segment and the state information for the transmitted TCP segments to a host system communicatively coupled to the NIC. A new TCP segment comprising the collected TCP segments may be generated after a termination event occurs. The generated new TCP segment, new state information for the generated new TCP segment, and the state information for the transmitted TCP segments may be communicated to the host system for TCP offload.

Abstract: Methods and systems for direct device access are disclosed. Aspects of one method may include a plurality of GOSs directly accessing a first network interface device, where the first network interface device may provide access to a network. One or more of the GOSs may be migrated to directly access a second network interface device, based on state information for each of the GOSs, where the state information may be maintained by the host. The GOSs may communicate data to a device coupled to the network by direct accessing the first and/or second network interface device. Similarly, the first and/or second network interface device may communicate data received from a device coupled to the network to one or more of the plurality of GOSs via direct access of the first and/or second network interface device.

Abstract: Certain aspects of a method and system for protocol offload in paravirtualized systems may be disclosed. Exemplary aspects of the method may include preposting of application buffers to a front-end driver rather than to a NIC in a paravirtualized system. The NIC may be enabled to place the received offloaded data packets into a received data buffer corresponding to a particular GOS. A back-end driver may be enabled to acknowledge the placed offloaded data packets. The back-end driver may be enabled to forward the received data buffer corresponding to the particular GOS to the front-end driver. The front-end driver may be enabled to copy offloaded data packets from a received data buffer corresponding to a particular guest operating system (GOS) to the preposted application buffers.

Abstract: Certain aspects of a method and system for user space TCP offload are disclosed. Aspects of a method may include offloading transmission control protocol (TCP) processing of received data to an on-chip processor. The received data may be posted directly to hardware, bypassing kernel processing of the received data, utilizing a user space library. If the received data is not cached in memory, an application buffer comprising the received data may be registered by the user space library. The application buffer may be pinned and posted to the hardware.

Abstract: Certain aspects of a method and system for transparent TCP offload with a user space library are disclosed. Aspects of a method may include collecting TCP segments in a network interface card (NIC) without transferring state information to a host system. When an event occurs that terminates the collection of TCP segments, a single aggregated TCP segment based on the collected TCP segments may be generated. The aggregated TCP segment may be posted directly to a user space library, bypassing kernel processing of the aggregated TCP segment.

Abstract: Methods and systems for protocol offload and direct I/O with I/O sharing in a virtualized network environment are disclosed. Aspects of one method may include a plurality of GOSs sharing a single network interface circuitry, or a network interface card, (NIC) that may provide access to a network. The NIC may directly handle processing of data to be transmitted to the network and/or data received from the network for each of the GOSs without a TGOS for the GOSs handling the data to be transmitted to the network and/or data received from the network. The data may be copied directly from a buffer in the single NIC to an application buffer for one of the plurality of GOSs and/or directly from an application buffer for one of the plurality of GOSs to the buffer in the single NIC.