An arbitration system for a shared address, data and control bus provides burst mode operations for transferring data between a peripheral device and memory via a bus master. The arbitration system is responsive to high priority bus activities, such as memory refresh cycles and DMA cycles to temporarily transfer control of the shared bus from the bus master to a circuit controlling the high priority activity. After the high priority activity is completed, the arbitration system returns control of the shared bus to the bus master so that the associated peripheral device may continue operating in the burst mode. This transfer of control occurs without requiring the time overhead of arbitrating priority between bus masters having active bus requests. The arbitration system further includes timing circuits to assure that a bus master transferring data in the burst mode does not retain control of the shared bus for an excessive amount of time.

Shared memory bus system for arbitrating access control among contending memory refresh circuits, peripheral controllers, and bus masters

A dual-bus architecture that includes a high-speed system bus, called the NexBus (20), and a slower peripheral bus, called the alternate bus or AB (25). The NexBus and AB are coupled by control logic (45) which includes an arbiter (50) and an alternate bus interface (ABI) (60). The ABI is treated as a master for both the NexBus and the AB. While it would be possible to have the adapter always request the AB (which also requires the NexBus), that would slow down NexBus operations to the bandwidth of the AB. This problem is avoided by providing two request lines for each adapter, -NREQ (NexBus only) and -AREQ (both buses), having the adapter normally assert -NREQ first. However, if the addressed device is on the AB, the ABI automatically detects this fact and attempts to do a crossing transfer to the AB, even though the request was for the NexBus only. If the ABI is unable to do the crossing transfer because the AB was busy, the ABI automatically causes the NexBus adapter to retry the request using the -AREQ line. Thus the slower AB is only accessed when actually necessary.

Crossing transfers for maximizing the effective bandwidth in a dual-bus architecture

A fast arbiter for handling a large number of types of resources with multiple instances of each type of resource is provided. During a first cycle a request logic circuit broadcasts a request for a preselected type of resource onto a broadcast medium. During a second cycle a grant logic circuit broadcasts a queue position onto the broadcast medium for the preselected type of resource. Also, during the second cycle the request logic circuit stores the queue position. After an asynchronous wait for an instance of the requested type of resource to become free, a third cycle begins wherein the first grant logic circuit broadcasts an indication that a free instance of the preselected type of resource is available. During a fourth cycle the requester is granted access to the free instance of the preselected type of resource.

Fast arbiter having easy scaling for large numbers of requesters, large numbers of resource types with multiple instances of each type, and selectable queuing disciplines

A processor apparatus for use in a multiprocessor computer system having a main memory storing a plurality of data items and being coupled to a bus operating according of a SNOOPY protocol. The processor apparatus includes a processor, a primary cache, a backup cache and a bus interface. The backup cache memory a first TAG store comprising a plurality of VALID indicators, one VALID indicator for each of the data items currently contained in the backup cache memory. The primary cache memory includes a second TAG store comprising a plurality of address indicators and a plurality of VALID indicators, one address indicator and one VALID indicator for each of the data items currently contained in the primary cache memory. The interface includes a duplicate TAG store coupled to the primary cache memory, the duplicate TAG store consisting of a copy of the address indicators of the second TAG store. The bus interface is coupled to the processor, the backup cache memory and to the bus. The bus interface operates in accordance with a SNOOPY protocol to monitor transactions on the bus for write transactions affecting data items in the corresponding backup cache memory having set VALID indicators. The bus interface will invalidate or update each VALID data item of the backup cache memory when there is a write transaction affecting data item and assert an invalidate signal for an affected data item indicated by the address indicators of the duplicate TAG store. The invalidate signal causes the VALID indicator in the second TAG store for the affected data item to be cleared.

Duplicate tag store for a processor having primary and backup cache memories in a multiprocessor computer system

A multiprocessor computer system is provided having a multiplicity of sub-systems and a main memory coupled to a system controller. An interconnect module, interconnects the main memory and sub-systems in accordance with interconnect control signals received from the system controller. At least two of the sub-systems are data processors, each having a respective cache memory that stores multiple blocks of data and a respective master cache index. Each master cache index has a set of master cache tags (Etags), including one cache tag for each data block stored by the cache memory. Each data processor includes a master interface having master classes for sending memory transaction requests to the system controller. The system controller includes memory transaction request logic for processing each memory transaction request by a data processor. The system controller maintains a duplicate cache index having a set of duplicate cache tags (Dtags) for each data processor. Each data processor has a writeback buffer for storing the data block previously stored in a victimized cache line until its respective writeback transaction is completed and an Nth+1 Dtag for storing the cache state of a cache line associated with a read transaction which is executed prior to an associated writeback transaction of a read-writeback transaction pair. Accordingly, upon a cache miss, the interconnect may execute the read and writeback transactions in parallel relying on the writeback buffer or Nth+1 Dtag to accommodate any ordering of the transactions.

Parallelized coherent read and writeback transaction processing system for use in a packet switched cache coherent multiprocessor system

A method and apparatus for selectively invalidating tag data related to data stored in high speed processor cache memory systems. The tag data to be invalidated, due to processor operations and cache memory misses, is stored in two tag stores and indicia related to the tag data to be invalidated is stored in a purge queue. Further improvement in system performance is provided by selective tag data and indicia elimination.

Duplicate tag store purge queue

A method and apparatus for granting, to a select processor in a multiprocessor computing system, exclusive access to a bus for issuance of address, data and command signals thereover, wherein each processor includes bus lock request and bus lock assert elements which provide corresponding bus request and bus hold signals which are recognized by corresponding elements included in other processors connected to the bus. The bus lock according to the present invention assures the processor having lock status of privacy on the bus necessary to complete a specified operation without interruption from the other processors.

Multiprocessor bus locking system with a winning processor broadcasting an ownership signal causing all processors to halt their requests

A decentralized, pipelined, synchronous bus arbitration scheme which allows almost completely fair arbitration between multiple devices competing for the use of a communication bus while allowing the device that last used the bus faster access to the bus if no other device is competing for its use. The arbitration method and apparatus according to the present invention allows all competing devices equal access to the bus, with the exception that when bus requests are posted simultaneously, the device with the higher priority will always be granted use of the bus first.

Andrew Milia

Papers

Duplicate tag store purge queue

Multiprocessor bus locking system with a winning processor broadcasting an ownership signal causing all processors to halt their requests

Quasi-fair arbitration scheme with default owner speedup