Abstract: A computer processor includes a replay system to replay instructions which have not executed properly and a first event pipeline coupled to the replay system to process instructions including any replayed instructions. A second event pipeline is provided to perform additional processing on an instruction. The second event pipeline has an ability to detect one or more faults occurring therein. The processor also includes a synchronization circuit coupled between the first event pipeline and the second event pipeline to synchronize faults occurring in the second event pipeline to matching instruction entries in the first event pipeline.

Abstract: According to one aspect of the invention, a method is provided in which store addresses of store instructions dispatched during a last predetermined number of cycles are maintained in a first data structure of a first processor. It is determined whether a load address of a first load instruction matches one of the store addresses in the first data structure. The first load instruction is replayed if the load address of the first load instruction matches one of the store addresses in the first data structure.

Abstract: According to one aspect of the invention, a method is provided in which store addresses of store instructions dispatched during a last predetermined number of cycles are maintained in a first data structure of a first processor. It is determined whether a load address of a first load instruction matches one of the store addresses in the first data structure. The first load instruction is replayed if the load address of the first load instruction matches one of the store addresses in the first data structure.

Abstract: A request received from a requester to access a processor cache or register file or the like is buffered, by storing requestor identification, request type, address, and a status of the request. This buffered request may be forwarded to the cache if it has the highest priority among a number of buffered requests that also wish to access the cache. The priority is a function of at least the requestor identification, the requester type, and the status of the request. For buffered requests which include a read, the buffered request is deleted after, not before, receiving an indication that the requestor has received the data read from the cache.

Abstract: A method of controlling operations by one or more processors includes granting ownership of a memory location having data stored therein to a first processor and performing, in an atomic manner by the first processor, a read operation to load the data from the memory location to a register, a modify operation to modify the data in the register, and a write operation to store the data from the register to the memory location. The method also prevents other operations directed towards the data by a second processor while the read, modify, and write operations are performed by the first processor, and vice versa. Ownership of the memory location is released after performing the read, modify, and write operations so as to allow the first or second processors to perform subsequent atomic operations.

Abstract: A write combining buffer that supports snoop requests includes a first cache memory and a second cache memory. The apparatus also includes a write combining buffer, coupled to the first and second cache memories, to combine data from a plurality of store operations. Each of the plurality of store operations is to at least a part of a cache line, and the write combining buffer can be snooped in response to requests initiated external to the apparatus.

Abstract: A system for write-combining uncacheable stores includes a memory order buffer, which receives first and second stores, and a data cache address and control, which receives the first and second stores from the memory order buffer. One of the memory order buffer and the data cache address and control determines whether the first and second stores are uncacheable and whether the first and second stores are contiguous in memory. If those conditions are satisfied, the data cache address and control write-combines the first and second stores before committing them to memory. The system may also apply additional conditions to determine whether the stores should be write-combined, for example requiring a minimum size for each store.

Abstract: An integrated, on-chip thermal management system providing closed-loop temperature control of an IC device and methods of performing thermal management of an IC device. The thermal management system comprises a temperature detection element, a power modulation element, a control element, and a visibility element. The temperature detection element includes a temperature sensor for detecting die temperature. The power modulation element may reduce the power consumption of an IC device by directly lowering the power consumption of the IC device, by limiting the speed at which the IC device executes instructions, by limiting the number of instructions executed by the IC device, or by a combination of these techniques. The control element allows for control over the behavior of the thermal management system, and the visibility element allows external devices to monitor the status of the thermal management system.

Abstract: The present invention provides a method for maintaining cache coherency while minimizing the power consumption. The method includes operating a first processor in a reduced power mode. While the first processor is operating in a reduced power mode, certain portions of the internal logic in the first processor remain clocked so that the first processor continues to monitor transactions on the system bus. The second processor runs a transaction on the system bus to request data. In the event that the first processor determines that the transaction by the second processor is requesting cache data that is stored in the first processor in a modified state, the first processor signals the second processor. After the current bus cycle is completed, the first processor writes back the modified cache line on the system bus and second processor re-runs the transaction on the system bus.

Abstract: A method for maintaining cache coherency while minimizing the power consumption. The method includes operating a first processor in a reduced power mode. While the first processor is operating in a reduced power mode, certain portions of the internal logic in the first processor remain clocked so that the first processor continues to monitor transactions on the system bus. The second processor runs a transaction on the system bus to request data. In the event that the first processor determines that the transaction by the second processor is requesting cache data that is stored in the first processor in a modified state, the first processor signals the second processor. After the current bus cycle is completed, the first processor writes back the modified cache line on the system bus and second processor re-runs the transaction on the system bus.

Abstract: A multiprocessing system maintains cache coherency during a reduced power mode of operation. The multiprocessing system has a first and a second processor coupled to the bus to perform data transactions with the main memory. During the reduced power mode of operation, the internal clock signal of the second processor is decoupled from a portion of the internal logic of the second processor while remaining coupled to a portion of the internal logic of the second processor that is used to monitor and respond to the traffic on the external bus to maintain cache coherency. During the reduced power mode of operation, the second processor continues to perform snoop and write-back processes to maintain a cache coherent multiprocessing system.