Abstract: An apparatus includes a processor, a first interface configured to connect to a bus of the apparatus, a second interface configured to communicate over a packet network, and circuitry. The circuitry is configured to, in a first operational mode, exchange data between the processor and one or more remote devices over the packet network, via the second interface, and in a second operational mode, monitor the bus using the first interface, detect a predefined trigger event occurring on the bus and, in response to detecting the trigger event, log one or more transactions on the bus that are adjacent to the trigger event and generate one or more protocol-analysis packets comprising at least part of the logged transactions.

Abstract: Apparatuses and methods are described that provide for a mechanism for allocating physical device memory for one or more virtual functions. In particular, a memory allocating framework is provided to utilize device memory more efficiently by mapping at least one target location of the physical memory in a Base Address Register (BAR) associated with the virtual function from a plurality of available target locations based on an allocation request. The memory allocating framework is further configured to compare an indication associated with the requesting virtual function to an identifier of the requested target location. Moreover, the memory allocating framework is further configured to allow the simultaneous use of more than one virtual function at a time while providing isolation between multiple virtual functions.

Abstract: An apparatus includes an interface and memory acquisition circuitry. The interface is configured to communicate over a bus operating in accordance with a bus protocol, which supports address-translation transactions that translate between bus addresses in an address space of the bus and physical memory addresses in an address space of a memory. The memory acquisition circuitry is configured to read data from the memory by issuing over the bus, using the bus protocol, one or more requests that (i) specify addresses to be read in terms of the physical memory addresses, and (ii) indicate that the physical memory addresses in the requests have been translated from corresponding bus addresses even though the addresses were not obtained by any address-translation transaction over the bus.

Abstract: A computing system comprises one or more cores. Each core comprises a processor. In some implementations, each processor is coupled to a communication network among the cores. In some implementations, a switch in each core includes switching circuitry to forward data received over data paths from other cores to the processor and to switches of other cores, and to forward data received from the processor to switches of other cores. Also disclosed are techniques for implementing hierarchical serial interconnects such as a PCI Express switch topology over a coherent mesh interconnect.

Abstract: A network interface device is connected to a host computer by having a memory controller, and a scatter-gather offload engine linked to the memory controller. The network interface device prepares a descriptor including a plurality of specified memory locations in the host computer, incorporates the descriptor in exactly one upload packet, transmits the upload packet to the scatter-gather offload engine via the uplink, invokes the scatter-gather offload engine to perform memory access operations cooperatively with the memory controller at the specified memory locations of the descriptor, and accepts results of the memory access operations.

Abstract: A network communication device includes a host interface, which is coupled to communicate with a host processor, having a memory, so as to receive a work request to convey one or more data blocks over a network. The work request specifies a memory region of a given data size, and at least one data integrity field (DIF), having a given field size, is associated with the data blocks. Network interface circuitry is configured to execute an input/output (I/O) data transfer operation responsively to the work request so as to transfer to or from the memory a quantity of data that differs from the data size of the memory region by a multiple of the field size, while adding the at least one DIF to the transferred data or removing the at least one DIF from the transferred data.

Abstract: A network communication device includes a host interface, which is coupled to communicate with a host processor, having a memory, so as to receive a work request to convey one or more data blocks over a network. The work request specifies a memory region of a given data size, and at least one data integrity field (DIF), having a given field size, is associated with the data blocks. Network interface circuitry is configured to execute an input/output (I/O) data transfer operation responsively to the work request so as to transfer to or from the memory a quantity of data that differs from the data size of the memory region by a multiple of the field size, while adding the at least one DIF to the transferred data or removing the at least one DIF from the transferred data.

Abstract: A network interface adapter includes a memory interface, for coupling to a memory containing a first data packet composed in accordance with a first communication protocol, and a network interface, for coupling to a packet communication network. Packet processing circuitry in the adapter reads the first data packet from the memory via the memory interface, computes a checksum of the first data packet, inserts the checksum in the first data packet in accordance with the first communication protocol, and encapsulates the first data packet in a payload of a second data packet in accordance with a second communication protocol applicable to the packet communication network, so as to transmit the second data packet over the network via the network interface. The circuitry likewise computes checksums of incoming encapsulated data packets from the network.