Abstract: A renaming register file complex for saving power is described. A mapping unit transforms an instruction register number (IRN) to a logical register number (LRN). The renaming register file maps an LRN to a physical register number (PRN), there being a greater number of physical registers than addressable by direct use of the IRN. The renaming register file uses a content addressable memory (CAM) to provide the mapping function. The renaming register file CAM further uses current processor state information to selectively enable tag comparators to minimize power in accessing registers. When a tag comparator is not enabled it remains in a low power state. A processor using a renaming register file with low power features is also described.

Abstract: Hazard detection is simplified by converting a conditional instruction, operative to perform an operation if a condition is satisfied, into an emissary instruction operative to evaluate the condition and an unconditional base instruction operative to perform the operation. The emissary instruction is executed, while the base instruction is halted. The emissary instruction evaluates the condition and reports the condition evaluation back to the base instruction. Based on the condition evaluation, the base instruction is either launched into the pipeline for execution, or it is discarded (or a NOP, or null instruction, substituted for it). In either case, the dependencies of following instructions may be resolved.

Abstract: A processor includes a return stack circuit used for predicting procedure return addresses for instruction pre-fetching, wherein a return stack controller determines the number of return levels associated with a given return instruction, and pops that number of return addresses from the return stack. Popping multiple return addresses from the return stack permits the processor to pre-fetch the return address of the original calling procedure in a chain of successive procedure calls. In one embodiment, the return stack controller reads the number of return levels from a value embedded in the return instruction. A complementary compiler calculates the return level values for given return instructions and embeds those values in them at compile-time. In another embodiment, the return stack circuit dynamically tracks the number of return levels by counting the procedure calls (branches) in a chain of successive procedure calls.

Abstract: When a branch misprediction in a pipelined processor is discovered, if the mispredicted branch instruction is not the last uncommitted instruction in the pipelines, older uncommitted instructions are checked for dependency on a long latency operation. If one is discovered, all uncommitted instructions are flushed from the pipelines without waiting for the dependency to be resolved. The branch prediction is corrected, and the branch instruction and all flushed instructions older than the branch instruction are re-fetched and executed.

Abstract: Data from a source domain operating at a first data rate is transferred to a FIFO in another domain operating at a different data rate. The FIFO buffers data before transfer to a sink for further processing or storage. A source side counter tracks space available in the FIFO. In disclosed examples, the initial counter value corresponds to FIFO depth. The counter decrements in response to a data ready signal from the source domain, without delay. The counter increments in response to signaling from the sink domain of a read of data off the FIFO. Hence, incrementing is subject to the signaling latency between domains. The source may send one more beat of data when the counter indicates the FIFO is full. The last beat of data is continuously sent from the source until it is indicated that a FIFO position became available; effectively providing one more FIFO position.

Abstract: A method and system for ensuring orderly forward progress in granting snoop castout requests. Masters may include a tag (“request tag”) in their transfer requests to a bus macro. The request tag indicates the order of the request issued by the master. If the bus macro determines that the transfer request is snoopable, then the bus macro broadcasts a snoop request that includes the request tag. If a snoop controller determines that the address in the snoop request is a hit to a modified coherency granule in an associated cache, then the master associated with that snoop controller transmits a castout request to the bus macro that includes the request tag associated with the snoop request. The bus macro uses the request tag to determine whether the castout request is a response to the oldest in a series of pipelined snoop requests to be serviced.

Abstract: A method of operating a request FIFO of a system on a chip (SoC) in which a requests in a first position that has been granted and which subsequently receives a retry from the intended target is automatically re-ordered with respect to the other requests below it in the request FIFO. Each issued requests is tagged to either enable or disable a re-order feature. When a request that is tagged as re-order enabled is granted, the FIFO logic monitors the response provided for the request. If the response is a retry, the request is removed from the first position of the request FIFO and the next sequential request is moved into the first position. The removed requests may be re-ordered within the request FIFO or sent back to the initiator. In the former implementation, controller logic reorders the first request within the request FIFO.

Abstract: A data processing system with a snooper that is capable of dynamically enabling and disabling its snooping capabilities (i.e., snoop detect and response). The snooper is connected to a bus controller via a plurality of interconnects, including a snooperPresent signal, a snoop response signal and a snoop detect signal. When the snooperPresent signal is asserted, subsequent snoop requests are sent to the snooper, and the snooper is polled for a snoop response. Each snooper is capable of responding at different times (i.e., each snooper operates with different snoop latencies). The bus controller individually tracks the snoop response received from each snooper with the snooperPresent signal enabled. Whenever the snooper wishes to deactivate its snooping capabilities/operations, the snooper de-asserts the snooperPresent signal. The bus controller recognizes this as an indication that the snooper is unavailable.

Abstract: A random access memory circuit comprises a plurality of memory cells and at least one decoder coupled to the memory cells, the decoder being configurable for receiving an input address and for accessing one or more of the memory cells in response thereto. The random access memory circuit further comprises a plurality of sense amplifiers operatively coupled to the memory cells, the sense amplifiers being configurable for determining a logical state of one or more of the memory cells. A controller coupled to at least a portion of the sense amplifiers is configurable for selectively operating in at least one of a first mode and a second mode. In the first mode of operation, the controller enables one of the sense amplifiers corresponding to the input address and disables the sense amplifiers not corresponding to the input address. In the second mode of operation, the controller enables substantially all of the sense amplifiers.

Abstract: A method for prioritizing snoop pushes in a data processing system that schedules requests within a request FIFO. Each new request that is received is placed in the last position of the request FIFO and the request FIFO typically grants request based solely on the order within the request FIFO. As prior requests are sequentially granted the subsequent requests move closer to a first position of the request FIFO. However, when a snoop push is received at the request FIFO, the snoop push is automatically inserted at the first position of the request FIFO ahead of all yet to be granted requests within the request FIFO.

Abstract: A method of designing a system on a chip (SoC) to operate with varying latencies and frequencies. A layout of the chip is designed with specific placement of devices, including a bus controller, initiator, and target devices. The time for a signal to propagate from a source device to a destination device is determined relative to a default propagation time. A pipeline stage is then inserted into a bus path between said source device and destination device for each additional time the signal takes to propagate. Each device (i.e., initiators, targets, and bus controller) is designed with logic to control a protocol that functions with a variety of response latencies. With the additional logic, the devices do not need to be changed when pipeline stages are inserted in the various paths. Registers are utilized as the pipeline stages that are inserted within the paths.

Abstract: A method and system for ensuring orderly forward progress in granting snoop castout requests. Masters may include a tag (“request tag”) in their transfer requests to a bus macro. The request tag indicates the order of the request issued by the master. If the bus macro determines that the transfer request is snoopable, then the bus macro broadcasts a snoop request that includes the request tag. If a snoop controller determines that the address in the snoop request is a hit to a modified coherency granule in an associated cache, then the master associated with that snoop controller transmits a castout request to the bus macro that includes the request tag associated with the snoop request. The bus macro uses the request tag to determine whether the castout request is a response to the oldest in a series of pipelined snoop requests to be serviced.

Abstract: A method and system for reducing power in a snooping cache based environment. A memory may be coupled to a plurality of processing units via a bus. Each processing unit may comprise a cache controller coupled to a cache associated with the processing unit. The cache controller may comprise a segment register comprising N bits where each bit in the segment register may be associated with a segment of memory divided into N segments. The cache controller may be configured to snoop a requested address on the bus. Upon determining which bit in the segment register is associated with the snooped requested address, the segment register may determine if the bit associated with the snooped requested address is set. If the bit is not set, then a cache search may not be performed thereby mitigating the power consumption associated with a snooped request cache search.

Abstract: A method and system for performing variable-sized memory coherency transactions. A bus interface unit coupled between a slave and a master may be configured to receive a request (master request) comprising a plurality of coherency granules from the master. Each snooping unit in the system may be configured to snoop a different number of coherency granules in the master request at a time. Once the bus interface unit has received a collection of sets of indications from each snooping logic unit indicating that the associated collection of coherency granules in the master request have been snooped by each snooping unit and that the data at the addresses for the collection of coherency granules snooped has not been updated, the bus interface unit may allow the data at the addresses of those coherency granules not updated to be transferred between the requesting master and the slave.

Abstract: A data processing system with a snooper that is capable of dynamically enabling and disabling its snooping capabilities (i.e., snoop detect and response). The snooper is connected to a bus controller via a plurality of interconnects, including a snooperPresent signal, a snoop response signal and a snoop detect signal. When the snooperPresent signal is asserted, subsequent snoop requests are sent to the snooper, and the snooper is polled for a snoop response. Each snooper is capable of responding at different times (i.e., each snooper operates with different snoop latencies). The bus controller individually tracks the snoop response received from each snooper with the snooperPresent signal enabled. Whenever the snooper wishes to deactivate its snooping capabilities/operations, the snooper de-asserts the snooperPresent signal. The bus controller recognizes this as an indication that the snooper is unavailable.

Abstract: A method of designing a system on a chip (SoC) to operate with varying latencies and frequencies. A layout of the chip is designed with specific placement of devices, including a bus controller, initiator, and target devices. The time for a signal to propagate from a source device to a destination device is determined relative to a default propagation time. A pipeline stage is then inserted into a bus path between said source device and destination device for each additional time the signal takes to propagate. Each device (i.e., initiators, targets, and bus controller) is designed with logic to control a protocol that functions with a variety of response latencies. With the additional logic, the devices do not need to be changed when pipeline stages are inserted in the various paths. Registers are utilized as the pipeline stages that are inserted within the paths.

Abstract: A method for prioritizing snoop pushes in a data processing system that schedules requests within a request FIFO. Each new request that is received is placed in the last position of the request FIFO and the request FIFO typically grants request based solely on the order within the request FIFO. As prior requests are sequentially granted the subsequent requests move closer to a first position of the request FIFO. However, when a snoop push is received at the request FIFO, the snoop push is automatically inserted at the first position of the request FIFO ahead of all yet to be granted requests within the request FIFO.

Abstract: A method of operating a request FIFO of a system on a chip (SoC) in which a requests in a first position that has been granted and which subsequently receives a retry from the intended target is automatically re-ordered with respect to the other requests below it in the request FIFO. Each issued requests is tagged to either enable or disable a re-order feature. When a request that is tagged as re-order enabled is granted, the FIFO logic monitors the response provided for the request. If the response is a retry, the request is removed from the first position of the request FIFO and the next sequential request is moved into the first position. The removed requests may be re-ordered within the request FIFO or sent back to the initiator. In the former implementation, controller logic reorders the first request within the request FIFO.

Abstract: A method and system for performing variable-sized memory coherency transactions. A bus interface unit coupled between a slave and a master may be configured to receive a request (master request) comprising a plurality of coherency granules from the master. Each snooping unit in the system may be configured to snoop a different number of coherency granules in the master request at a time. Once the bus interface unit has received a collection of sets of indications from each snooping logic unit indicating that the associated collection of coherency granules in the master request have been snooped by each snooping unit and that the data at the addresses for the collection of coherency granules snooped has not been updated, the bus interface unit may allow the data at the addresses of those coherency granules not updated to be transferred between the requesting master and the slave.

Abstract: A method and system for reducing power in a snooping cache based environment. A memory may be coupled to a plurality of processing units via a bus. Each processing unit may comprise a cache controller coupled to a cache associated with the processing unit. The cache controller may comprise a segment register comprising N bits where each bit in the segment register may be associated with a segment of memory divided into N segments. The cache controller may be configured to snoop a requested address on the bus. Upon determining which bit in the segment register is associated with the snooped requested address, the segment register may determine if the bit associated with the snooped requested address is set. If the bit is not set, then a cache search may not be performed thereby mitigating the power consumption associated with a snooped request cache search.