摘要:
A method and system for a multi-phase net refresh on a bus bridge interconnect, the interconnect comprising a number of nodes, a bus bridge, and a number of buses, are described. In one embodiment, a primary bus is acquired by communicating with other bus bridges on the buses. A secondary bus is breached to acquire the secondary bus. In addition, the primary bus and the secondary bus are committed.
摘要:
In a first embodiment, an applications programming interface (API) implements and manages isochronous and asychronous data transfer operations between an application and a bus structure. During an asynchronous transfer the API includes the ability to transfer any amount of data between one or more local data buffers within the application and a range of addresses over the bus structure using one or more asynchronous transactions. An automatic transaction generator may be used to automatically generate the transactions necessary to complete the data transfer. The API also includes the ability to transfer data between the application and another node on the bus structure isochronously over a dedicated channel. During an isochronous data transfer, a buffer management scheme is used to manage a linked list of data buffer descriptors. During isochronous transfer of data, the API provides implementation of a resynchronization event in the stream of data allowing for resynchronization by the application to a specific point within the data. Implementation is also provided for a callback routine for each buffer in the list which calls the application at a predetermined point during the transfer of data. An isochronous API of the preferred embodiment presents a virtual representation of a plug, using a plug handle, to the application. The isochronous API notifies a client application of any state changes on a connected plug through the event handle. The isochronous API also manages buffers utilized during a data operation by attaching and detaching the buffers to the connected plug, as appropriate, to mange the data flow.
摘要:
In a first embodiment, an applications programming interface (API) implements and manages isochronous and asynchronous data transfer operations between an application and a bus structure. During an synchronous transfer the API includes the ability to transfer any amount of data between one or more local data buffers within the application and a range of addresses over the bus structure using one or more asynchronous transactions. An automatic transaction generator may be used to automatically generate the transactions necessary to complete the data transfer. The API also includes the ability to transfer data between the application and another node on the bus structure isochronously over a dedicated channel. During an isochronous data transfer, a buffer management scheme is used to manage a linked list of data buffer descriptors. During isochronous transfers of data, the API provides implementation of a resynchronization event in the stream of data allowing for resynchronization by the application to a specific point within the data. Implementation is also provided for a callback routine for each buffer in the list which calls the application at a predetermined point during the transfer of data. An isochronous API of the preferred embodiment presents a virtual representation of a plug, using a plug handle, to the application. The isochronous API notifies a client application of any state changes on a connected plug through the event handle. The isochronous API also manages buffers utilized during a data operation by attaching and detaching the buffers to the connected plug, as appropriate, to manage the data flow.
摘要:
In a first embodiment, an applications programming interface (API) implements and manages isochronous and asynchronous data transfer operations between an application and a bus structure. During an asynchronous transfer the API includes the ability to transfer any amount of data between one or more local data buffers within the application and a range of addresses over the bus structure using one or more, asynchronous transactions. An automatic transaction generator may be used to automatically generate the transactions necessary to complete the data transfer. The API also includes the ability to transfer data between the application and another node on the bus structure isochronously over a dedicated channel. During an isochronous data transfer, a buffer management scheme is used to manage a linked list of data buffer descriptors. During isochronous transfers of data, the API provides implementation of a resynchronization event in the stream of data allowing for resynchronization by the application to a specific point within the data. Implementation is also provided for a callback routine for each buffer in the list which calls the application at a predetermined point during the transfer of data. An isochronous API of the preferred embodiment presents a virtual representation of a plug, using a plug handle, to the application. The isochronous API notifies a client application of any state changes on a connected plug through the event handle. The isochronous API also manages buffers utilized during a data operation by attaching and detaching the buffers to the connected plug, as appropriate, to manage the data flow.
摘要:
A system and method for multi-level context switching in an electronic network comprises a control state machine configured to implement a data priority scheme, a return address generator configured to hold and release return addresses for interrupted instruction modules in accordance with the data priority scheme and context information from the electronic network, and a processor configured to process data from the electronic network in accordance with the data priority scheme and the context information. Receive registers stores data received from the electronic network. The control state machine includes a switch address generator and a program counter select. The switch address generator outputs a switch address, which is an address for a first instruction for a selected-context instruction module. The return address generator holds and releases the return addresses, which are addresses of next consecutive instructions, when an instruction module is interrupted for a context switch. The program counter select outputs a switch address, a return address, or a next consecutive address to select the appropriate instruction in the memory for execution by the processor.
摘要:
A system and method for context switching in an electronic network comprises a memory configured to store instruction modules, each instruction module corresponding to a context, a processor that executes the instruction modules, and a control state machine. The control state machine selects one of the instruction modules for execution by the processor according to context information from the electronic network. The control state machine includes a switch address generator, a return address register, and a program counter select. The switch address generator outputs a switch address, which is an address for a first instruction for a selected-context instruction module. The return address register stores a return address, which is an address of a next consecutive instruction, when an instruction module is interrupted for a context switch. The program counter select outputs the switch address, the return address, or a next consecutive address to select the appropriate instruction in the memory for execution by the processor.
摘要:
A system and method for multi-level context switching in an electronic network comprises a control state machine configured to implement a data priority scheme, a return address generator configured to hold and release return addresses for interrupted instruction modules in accordance with the data priority scheme and context information from the electronic network, and a processor configured to process data from the electronic network in accordance with the data priority scheme and the context information. Receive registers stores data received from the electronic network. The control state machine includes a switch address generator and a program counter select. The switch address generator outputs a switch address, which is an address for a first instruction for a selected-context instruction module. The return address generator holds and releases the return addresses, which are addresses of next consecutive instructions, when an instruction module is interrupted for a context switch. The program counter select outputs a switch address, a return address, or a next consecutive address to select the appropriate instruction in the memory for execution by the processor.
摘要:
A system and method for fast data transfers in an electronic network comprises a data transfer engine configured to independently execute data transfer instructions and a processor configured to forward the data transfer instructions to the data transfer engine. The data transfer engine preferably executes the data transfer instructions while the processor proceeds to execute a next instruction. The data transfer engine includes a transmit engine that executes transmit transfer instructions and a receive engine that executes receive transfer instructions. The transmit engine and the receive engine operate independently and thus may operate concurrently.
摘要:
An apparatus for dispatching a processing element to a program location based on a channel number of received data includes a channel pointer register having a number of storage locations each with a channel number field, a valid bit field and a corresponding instruction pointer field. When an isochronous channel is allocated for use for reception, the host device programs the channel number and a corresponding instruction pointer value into a storage location. When a storage location is programmed, a valid bit within that storage location is also preferably set. The corresponding instruction pointer value points to a series of instructions which are to be used to process data received on that isochronous channel. When isochronous data is then received, the channel number on which the data is received is compared to the channel numbers within the valid storage locations in the channel pointer register. If one of the channel numbers within a valid storage location matches the channel number of the received data, then the corresponding instruction pointer value is output and the data is processed according to a series of instructions beginning at the location specified by the corresponding instruction pointer value. Otherwise, if the channel number of the received data does not match any of the channel numbers within valid storage locations then a default instruction pointer value is output and the received data is processed according to a series of instructions beginning at the location specified by the default instruction pointer value.
摘要:
A disk drive system includes a disk controller system having separate head positioning and data transfer subsystems and supporting up to four disk drives, two of which may be removable and two of which may connect through a standard ST506 interface. The system provides a read and write protected cylinder for removable drives, a read only cylinder for removable drives which may store a unique serial number as well as other write protected data, and a high capacity, high speed sector buffer which allows continuous transfers of data to or from noninterleaved sectors and supports concurrent disk and system accesses.