International Business Machines Corporation

International Business Machines Corpation

Richard Nicholas

Bernard Charles Drerup

Eric Eugene Retter

Jeffrey A Stuecheli

Giora Biran

Thuong Quang Truong

Eric Francis Robinson

Hien Le

Jason Alan Cox

Robert John Dorsey

Luis E De La Torre

Jyoti Gupta

Kenneth Lee Wright

Barry Joe Wolford

Guy L Guthrie

Derek E Williams

William John Starke

Warren Edward Maule

Luis Alfonso Lastras-Montano

Hugh Shen

Naresh Nayar

James A O'Connor

Stephen J Powell

Jason Frederick Cantin

Michael Raymond Trombley

Robert Christopher Dixon

Joab Daniel Henderson

Ram Raghavan

Kirk Edward Morrow

Luke Murray

Prasanna Srinivasan

Daniel James Henderson

Alexander Michael Taft

Michael J Mueller

Kanwal Bahri

A data processing system includes a plurality of processor cores each supported by a respective one of a plurality of vertical cache hierarchies. Based on receiving on a system fabric a cache injection request requesting injection of a data into a cache line identified by a target real address, the data is written into a cache in a first vertical cache hierarchy among the plurality of vertical cache hierarchies. Based on a value in a field of the cache injection request, a distribute field is set in a directory entry of the first vertical cache hierarchy. Upon eviction of the cache line the first vertical cache hierarchy, a determination is made whether the distribute field is set. Based on determining the distribute field is set, a lateral castout of the cache line from the first vertical cache hierarchy to a second vertical cache hierarchy is performed.

Distribution of injected data among caches of a data processing system

Address error detection including a method that receives a read address corresponding to a read location in a memory. Data is read from the read location in the memory. The data is transformed at a computer based on the data and the read address to produce read data. Error correction codes (ECC) bits associated the read data are read from the read location in the memory. The ECC bits were generated based on the write data. It is determined whether the read data has an address error responsive to the read data and the ECC bits associated with the write data. An error is generated in response to determining that the read address has an address error.

Address error detection

An approach for saving power in a memory subsystem that uses memory access idle timer to enable low power mode and memory scrub operation within computing system has been provided. The computing system determines that a memory subsystem is switched out of low power operation mode due to a memory scrub operation. In addition, the computing system bypasses the low power operation mode of an idle timer of the memory subsystem such that the memory subsystem is returned to the low power operation mode upon completion of the memory scrub operation. The computing system further sets a scrub flag of the memory subsystem to a high state, and clears the scrub flag to a low state to track if the idle timer should be bypassed.

Method and apparatus for mitigating effects of memory scrub operations on idle time power savings modes

Address error detection including a method that receives write data and a write address, the write address corresponding to a location in a memory. Error correction code (ECC) bits are generated based on the received write data. The write data is transformed at a computer based on the write address and the write data, to produce transformed write data. The transforming is configured to cause an ECC to detect an address error during a read operation to the write address in response to a mismatch between either the write address or the read address and data read from the location. The transformed write data and the ECC bits are written to the location in memory.

A method for maintaining cache coherency operates in a data processing system with a system memory and a plurality of processing units (PUs), each PU having a cache, and each PU coupled to at least another one of the plurality of PUs. A first PU receives a first data block for storage in a first cache of the first PU. The first PU stores the first data block in the first cache. The first PU assigns a first coherency state and a first tag to the first data block, wherein the first coherency state is one of a plurality of coherency states that indicate whether the first PU has accessed the first data block. The plurality of coherency states further indicate whether, in the event the first PU has not accessed the first data block, the first PU received the first data block from a neighboring PU.

System and method for cache coherency in a multiprocessor system

A method for managing data operates in a data processing system with a system memory and a plurality of processing units (PUs), each PU having a cache comprising a plurality of cache lines, each cache line having one of a plurality of coherency states, and each PU coupled to at least another one of the plurality of PUs. A first PU selects a castout cache line of a plurality of cache lines in a first cache of the first PU to be castout of the first cache. The first PU sends a request to a second PU, wherein the second PU is a neighboring PU of the first PU, and the request comprises a first address and first coherency state of the selected castout cache line. The second PU determines whether the first address matches an address of any cache line in the second PU. The second PU sends a response to the first PU based on a coherency state of each of a plurality of cache lines in the second cache and whether there is an address hit. The first PU determines whether to transmit the castout cache line to the second PU based on the response. And, in the event the first PU determines to transmit the castout cache line to the second PU, the first PU transmits the castout cache line to the second PU.

System and method for optimizing neighboring cache usage in a multiprocessor environment

A system and method for simulating a circuit design using both an unknown Boolean state and a negative unknown Boolean state is provided. When the circuit is simulated, one or more initial simulated logic elements are initialized to the unknown Boolean state. The initialized unknown Boolean states are then fed to one or more simulated logic elements and the simulator simulates the handling of the unknown Boolean state by the simulated logic elements. Examples of simulated logic elements include gates and latches, such as flip-flops, inverters, and basic logic gates. The processing results in at least one negative unknown Boolean state. An example of when a negative unknown Boolean state would result would be when the unknown Boolean state is inverted by an inverter. The resulting negative unknown Boolean state is then fed to other simulated logic elements that generate further simulation results based on processing the negative unknown Boolean state.

Structure for improved logic simulation using a negative unknown boolean state

An access speculation predictor is provided that may be implemented using idle command processing resources, such as registers of idle finite state machines (FSMs) in a memory controller. The access speculation predictor may predict whether to perform speculative retrieval of data for a data request from a main memory of the data processing system based on history information stored for a memory region targeted by the data request. In particular, a first address may be extracted from the data request and compared to memory regions associated with second addresses stored in address registers of a plurality of FSMs of the memory controller. A FSM whose memory region includes the first address may be selected. History information for the memory region may be obtained from the selected FSM. The history information may be used to control whether to speculatively retrieve the data for the data request from a main memory.

Access speculation predictor implemented via idle command processing resources

An access speculation predictor may predict whether to perform speculative retrieval of data for a data request from a main memory based on whether or not a domain indicator in the data request indicates that the cache line corresponding to the data has a special invalid state or not. In particular, a first address and a domain indicator are extracted from first data request. The first address is used to select a finite state machine (FSM) of a memory controller based on memory regions associated with the FSMs of the memory controller. Speculative retrieval of data for the first data request from main memory is controlled based on whether the domain indicator identifies the special invalid state or not and, if the domain indicator identifies that the cache line does not have the special invalid state, based on information stored in registers associated with the selected FSM.

Access speculation predictor with predictions based on a domain indicator of a cache line

An access speculation predictor may predict whether to perform speculative retrieval of data for a data request from a main memory based on whether a scope predictor indicates whether a local or global request is predicted to be necessary to obtain the data for the data request. In particular, a first address and a scope predictor may be extracted from a first data request. A determination may be made as to whether a memory controller receiving the first data request is local to a source of the first data request or not. Speculative retrieval of the data for the first data request from a main memory may be controlled based on whether the memory controller is local to the source of the first data request and whether the scope predictor identifies whether a local or a global request is predicted to be necessary.

Access speculation predictor with predictions based on a scope predictor

An access speculation predictor may predict whether to perform speculative retrieval of data for a data request from a main memory based on whether or not a current requestor tag matches a previous requestor tag. In particular, a first address and a first requester tag may be extracted from a first data request and a finite state machine (FSM) of a memory controller may be selected whose memory region includes the first address. A second requester tag, that identifies a previous requester that attempted to access the memory region association with the selected FSM, may be retrieved from a register associated with the selected FSM and compared to the first requester tag. Speculatively retrieving the data for the first data request from a main memory may be controlled based on results of the comparison of the first requester tag to the second requester tag.

Access speculation predictor with predictions based on memory region prior requestor tag information

Methods, systems, and computer program products are provided for dynamic selective memory mirroring in solid state devices. An amount of memory is reserved. Sections of the memory to select for mirroring in the reserved memory are dynamically determined. The selected sections of the memory contain critical areas. The selected sections of the memory are mirrored in the reserved memory.

Methods, systems, and computer program products for dynamic selective memory mirroring

A design structure for piggybacking multiple data tenures on a single data bus grant to achieve higher bus utilization is disclosed. In one embodiment of the design structure, a method in a computer-aided design system includes a source device sending a request for a bus grant to deliver data to a data bus connecting a source device and a destination device. The device receives the bus grant and logic within the device determines whether the bandwidth of the data bus allocated to the bus grant will be filled by the data. If the bandwidth of the data bus allocated to the bus grant will not be filled by the data, the device appends additional data to the first data and delivers the combined data to the data bus during the bus grant for the first data. When the bandwidth of the data bus allocated to the bus grant will be filled by the first data, the device delivers only the first data to the data bus during the bus grant.

Structure for piggybacking multiple data tenures on a single data bus grant to achieve higher bus utilization

In a method and apparatus associated with a bus controller, a set of mechanisms are selectively added to the bus controller, as well as to slave devices connected to the bus controller. A mechanism is also added to one or more master devices connected to the bus controller, in order to provide the master devices with a transaction ordering capability. The added mechanisms collectively achieve the objective of supporting connection of multiple slave devices to a common controller interface, and at the same time allowing pipelined operation of the slave devices. One embodiment of the invention is directed to a method for use with a bus and an associated bus controller, wherein the bus controller has respective master and slave interfaces for use in selectively interconnecting master devices and slave devices. The method comprises the steps of connecting one or more of the master devices to one of the master interfaces, and connecting each of a plurality of slave devices to the same one of the slave interfaces. The method further comprises operating a connected master device to send multiple commands to a selected one of the connected slave devices in accordance with a command pipelining procedure.

Method and apparatus for attaching multiple slave devices to a single bus controller interface while supporting command pipelining

A method for managing a cache operates in a data processing system with a system memory and a plurality of processing units (PUs). A first PU determines that one of a plurality of cache lines in a first cache of the first PU must be replaced with a first data block, and determines whether the first data block is a victim cache line from another one of the plurality of PUs. In the event the first data block is not a victim cache line from another one of the plurality of PUs, the first cache does not contain a cache line in coherency state invalid, and the first cache contains a cache line in coherency state moved, the first PU selects a cache line in coherency state moved, stores the first data block in the selected cache line and updates the coherency state of the first data block.

Growing community of inventors

Round Rock, TX, United States of America

Richard Nicholas