Abstract: Techniques and mechanisms for providing access to a function with an input/output (I/O) device. In an embodiment, a main memory of a computer system including the I/O device stores a function-context data structure associating a function with a context for an access to the function. The I/O device stores a configuration for the I/O device to provide the function. In another embodiment, the software process exchanges information with the function-context data structure for the access to the function. The I/O device performs a synchronization of the function-context data structure and the configuration data structure with respect to one another, wherein the function-context data structure operates as a register level interface which interfaces the I/O device and the software process with one another.

Abstract: A disk array redundancy controller ensures integrity of a mirrored or RAID storage array supporting a host system and minimizes recovery time responsive to a storage volume failure by traversing caches of recently written blocks to identify partially flushed stripes of data and recovering the inconsistent stripes on each of the storage volumes based on a master copy derived from the scan of all pre-failure caches of the storage array. The storage array employs nonvolatile caches in conjunction with solid state drive (SSD) storage volumes, allowing post-failure recovery of recently written blocks. A cache depth at least sufficient to store the largest stripe, or set of blocks, from the host ensures recovery of the entire stripe from a collective scan of the caches of all storage volumes of the storage array.

Abstract: Techniques herein include systems and methods for driver-assisted BAR mapping that virtualize PCI functions, but without virtualizing the storage media itself. Such techniques make use of unused BARs (Base Address Registers) of a master (Operating system-facing) device to gain access to other PCIe logical instances, while still exposing only a single PCIe function (connection or channel) to system software. This technique provides a new concept of logical PCIe device instances through BAR mapping by making use of unused BARs to extend access to any number of PCIe instances or memory-mapped I/O devices behind a master device such that only a single PCIe function is exposed to system software. Embodiments can thus extend access to one or more additional storage devices through one level of BAR indirection. As a result, such techniques and embodiments enable the multiplication of storage capacity and performance through the aggregation of multiple, similar hardware components.

Abstract: A computer system may comprise a plurality of cores that may process the tasks determined by the operating system. A network device may direct a first set of packets to a first core using a flow-spreading technique such as receive side scaling (RSS). However, the operating system may re-provision a task from the first core to a second core to balance the load, for example, on the computer system. The operating system may determine an identifier of the second core using a new data field in the socket calls to track the identifier of the second core. The operating system may provide the identifier of the second core to a network device. The network device may then direct a second set of packets to the second core using the identifier of the second core.

Abstract: Techniques herein include systems and methods for driver-assisted BAR mapping that virtualize PCI functions, but without virtualizing the storage media itself. Such techniques make use of unused BARs (Base Address Registers) of a master (Operating system-facing) device to gain access to other PCIe logical instances, while still exposing only a single PCIe function (connection or channel) to system software. This technique provides a new concept of logical PCIe device instances through BAR mapping by making use of unused BARs to extend access to any number of PCIe instances or memory-mapped I/O devices behind a master device such that only a single PCIe function is exposed to system software. Embodiments can thus extend access to one or more additional storage devices through one level of BAR indirection. As a result, such techniques and embodiments enable the multiplication of storage capacity and performance through the aggregation of multiple, similar hardware components.

Abstract: Techniques and mechanisms for managing resources of an aggregate device which spans multiple physical devices of a computer platform. In an embodiment, an aggregation device coupled to a host bus of the computer platform receives resource information generated by a pre-boot software process of the computer platform. In another embodiment, the aggregation device, based on the received resource information, represents a resource in a first input/output (I/O) device to a host operating system (OS) as residing in the aggregation device, the first I/O device coupled to the aggregation device via a host bus for exchanging communications referencing a shared address space.

Abstract: An embodiment of the invention includes (i) receiving a core identifier that corresponds with a processor source core; (ii) receiving an input/output request, produced from the source core, that is associated with the core identifier; (iii) and directing an interrupt, which corresponds to the request, to the source core based on the core identifier. Other embodiments are described herein.

Abstract: An embodiment of the invention includes (i) receiving a core identifier that corresponds with a processor source core; (ii) receiving an input/output request, produced from the source core, that is associated with the core identifier; (iii) and directing an interrupt, which corresponds to the request, to the source core based on the core identifier. Other embodiments are described herein.

Abstract: Proxy nodes perform TCP/IP processing on behalf of application nodes, utilize lightweight protocols to communicate with application nodes, and communicate with network nodes and network clients using Transmission Control Protocol/Internet Protocol (TCP/IP).

Abstract: An embodiment of the invention includes (i) receiving a core identifier that corresponds with a processor source core; (ii) receiving an input/output request, produced from the source core, that is associated with the core identifier; (iii) and directing an interrupt, which corresponds to the request, to the source core based on the core identifier. Other embodiments are described herein.

Abstract: In one embodiment, the present invention includes a method for receiving a first packet associated with a first network flow in a first descriptor queue associated with a first hardware thread, receiving a marker in the first descriptor queue to indicate migration of the first network flow from the first hardware thread to a second hardware thread, and processing a second packet of the first network flow following the first packet in order in the second hardware thread.

Abstract: A computer system may comprise a plurality of cores that may process the tasks determined by the operating system. A network device may direct a first set of packets to a first core using a flow-spreading technique such as receive side scaling (RSS). However, the operating system may re-provision a task from the first core to a second core to balance the load, for example, on the computer system. The operating system may determine an identifier of the second core using a new data field in the socket calls to track the identifier of the second core. The operating system may provide the identifier of the second core to a network device. The network device may then direct a second set of packets to the second core using the identifier of the second core.

Abstract: In an embodiment, a method is provided that may include providing a first address space exclusively and coherently accessible by a first processor core partition in a platform. A second address space may be provided in this embodiment that is exclusively and coherently accessible by a second processor core partition in the platform. Also in this embodiment, a third address space in the platform may be provided that is accessible, at least in part, by both the first and second processor core partitions and may be to permit communication between the first and second processor core partitions of at least one packet and at least one descriptor associated with the at least one packet. The at least one descriptor may indicate, at least in part, one or more locations in the third address space to store, at least in part, the at least one packet. Of course, many alternatives, modifications, and variations are possible without departing from this embodiment.

Abstract: In an embodiment, a method is provided that may include providing a first address space exclusively and coherently accessible by a first processor core partition in a platform. A second address space may be provided in this embodiment that is exclusively and coherently accessible by a second processor core partition in the platform. Also in this embodiment, a third address space in the platform may be provided that is accessible, at least in part, by both the first and second processor core partitions and may be to permit communication between the first and second processor core partitions of at least one packet and at least one descriptor associated with the at least one packet. The at least one descriptor may indicate, at least in part, one or more locations in the third address space to store, at least in part, the at least one packet. Of course, many alternatives, modifications, and variations are possible without departing from this embodiment.

Abstract: In one embodiment, the present invention includes a method for receiving a first packet associated with a first network flow in a first descriptor queue associated with a first hardware thread, receiving a marker in the first descriptor queue to indicate migration of the first network flow from the first hardware thread to a second hardware thread, and processing a second packet of the first network flow following the first packet in order in the second hardware thread.

Abstract: Methods and apparatus relating to notification of out-of-order packets are described. In an embodiment, data such as a sequence number and a flow identifier may be extracted from a packet. The extracted data may be used to check the extracted sequence number against an expected sequence number and indicate that the packet is an out-of-order packet. Other embodiments are also disclosed.

Abstract: Receive side scaling in a network system may be improved by moving the task of adapting the load distribution from the operating system (“OS”) to the network device. A load feedback mechanism may be used for the OS to report per-core load to the network device. With per-core load information from the OS as well as its own knowledge of new flows, the network device is able to map new flows to the least-utilized cores by changing these cores' entries in an indirection table in the network device directly.

Abstract: Techniques are described that can be used to implement a network interface. A network interface may be communicatively coupled to a general purpose core or hardware thread. Various operations can be assigned to be performed by the general purpose core, thereby at least to provide flexible operation of the network interface. The general purpose core may be capable to issue inter-processor interrupts to other cores or hardware threads to request processing. The other cores or hardware threads may respond to inter-processor interrupts by executing one or more interrupt service routine. The other cores or hardware threads may be capable to process network protocol units received by the network interface.

Abstract: An apparatus and a system may include an adaptation module, a plurality of Direct Transport Interfaces (DTIs), a DTI accelerator, and a Transport Control Protocol/Internet Protocol (TCP/IP) accelerator. The adaptation module may provide a translated sockets call from an application program to one of the DTIs, where an included set of memory structures may couple the translated sockets call to the DTI accelerator, which may in turn couple the set of memory structures to the TCP/IP accelerator. An article may include data causing a machine to perform a method including: receiving an application program sockets call at the adaptation module, deriving a translated sockets call from the application program sockets call, receiving the translated sockets call at a DTI, coupling the translated sockets call to a DTI accelerator using a set of memory structures in the DTI, and coupling the set of memory structures to a TCP/IP accelerator.

Abstract: Mechanisms to implement memory management to enable protocol-aware asynchronous, zero-copy transmits. A transport protocol engine exposes interfaces via which memory buffers from a memory pool in operating system (OS) kernel space may be allocated to applications running in an OS user layer. The memory buffers may be used to store data that is to be transferred to a network destination using a zero-copy transmit mechanism, wherein the data is directly transmitted from the memory buffers to the network via a network interface controller. The transport protocol engine also exposes a buffer reuse API to the user layer to enable applications to obtain buffer availability information maintained by the protocol engine. In view of the buffer availability information, the application may adjust its data transfer rate.