Abstract: A method for controlling memory refresh operations in dynamic random access memories. The method includes determining a count of deferred memory refresh operations for a first memory rank. Responsive to the count approaching a high priority threshold, issuing an early high priority refresh notification for the first memory rank, which indicates the pre-determined time for performing a high priority memory refresh operation at the first memory rank. Responsive to the early high priority refresh notification, the behavior of a read reorder queue is dynamically modified to give priority scheduling to at least one read command targeting the first memory rank, and one or more of the at least one read command is executed on the first memory rank according to the priority scheduling. Priority scheduling removes these commands from the re-order queue before the refresh operation is initiated at the first memory rank.

Abstract: Techniques for handling queuing of memory accesses prevent passing excessive requests that implicate a region of memory subject to a high latency memory operation, such as a memory refresh operation, memory scrubbing or an internal bus calibration event, to a re-order queue of a memory controller. The memory controller includes a queue for storing pending memory access requests, a re-order queue for receiving the requests, and a control logic implementing a queue controller that determines if there is a collision between a received request and an ongoing high-latency memory operation. If there is a collision, then transfer of the request to the re-order queue may be rejected outright, or a count of existing queued operations that collide with the high latency operation may be used to determine if queuing the new request will exceed a threshold number of such operations.

Abstract: A method and apparatus for controlling marking store updates in a central electronic complex with a plurality of core processors and eDRAM cache and interconnect bus to a service processor for loading memory controller firmware to dual-channel DDR3 memory controllers with an internal marking store. Loaded firmware of the memory controllers is responsible for tracking of ECC errors using a ECC decoder control whereby said marking store is written by a slow ECC decoder, and read by a fast ECC decoder for every read operation of said memory controllers to provide a blocking mechanism for notifying marking store firmware when the marking store has been updated and which guarantees that marking store firmware cannot write to the marking store until the marking store firmware has seen updates without causing the marking store hardware to time out.

Abstract: A memory system and data processing system for controlling memory refresh operations in dynamic random access memories. The memory controller comprises logic that: tracks a time remaining before a scheduled time for performing a high priority, high latency operation a first memory rank of the memory system; responsive to the time remaining reaching a pre-established early notification time before the schedule time for performing the high priority, high latency operation, biases the re-order queue containing memory access operations targeting the plurality of ranks to prioritize scheduling of any first memory access operations that target the first memory rank. The logic further: schedules the first memory access operations to the first memory rank for early completion relative to other memory access operations in the re-order queue that target other memory ranks; and performs the high priority, high latency operation at the first memory rank at the scheduled time.

Abstract: A technique for memory command throttling in a partitioned memory subsystem includes accepting, by a master memory controller included in multiple memory controllers, a synchronization command. The synchronization command includes command data that includes an associated synchronization indication (e.g., a synchronization bit or bits) for each of the multiple memory controllers and each of the multiple memory controllers controls a respective partition of the partitioned memory subsystem. In response to receiving the synchronization command, the master memory controller forwards the synchronization command to the multiple memory controllers. In response to receiving the forwarded synchronization command each of the multiple memory controllers de-asserts an associated status bit. In response to receiving the forwarded synchronization command, each of the multiple memory controllers determines whether the associated synchronization indication is asserted.

Abstract: Mechanisms are provided for performing a floating point arithmetic operation in a data processing system. A plurality of floating point operands of the floating point arithmetic operation are received and bits in a mantissa of at least one floating point operand of the plurality of floating point operands are shifted. One or more bits of the mantissa that are shifted outside a range of bits of the mantissa of at least one floating point operand are stored and a vector value is generated based on the stored one or more bits of the mantissa that are shifted outside of the range of bits of the mantissa of the at least one floating point operand. A resultant value is generated for the floating point arithmetic operation based on the vector value and the plurality of floating point operands.

Abstract: Mechanisms are provided for performing a floating point arithmetic operation in a data processing system. A plurality of floating point operands of the floating point arithmetic operation are received and bits in a mantissa of at least one floating point operand of the plurality of floating point operands are shifted. One or more bits of the mantissa that are shifted outside a range of bits of the mantissa of at least one floating point operand are stored and a vector value is generated based on the stored one or more bits of the mantissa that are shifted outside of the range of bits of the mantissa of the at least one floating point operand. A resultant value is generated for the floating point arithmetic operation based on the vector value and the plurality of floating point operands.

Abstract: Techniques for handling queuing of memory accesses prevent passing excessive requests that implicate a region of memory subject to a high latency memory operation, such as a memory refresh operation, memory scrubbing or an internal bus calibration event, to a re-order queue of a memory controller. The memory controller includes a queue for storing pending memory access requests, a re-order queue for receiving the requests, and a control logic implementing a queue controller that determines if there is a collision between a received request and an ongoing high-latency memory operation. If there is a collision, then transfer of the request to the re-order queue may be rejected outright, or a count of existing queued operations that collide with the high latency operation may be used to determine if queuing the new request will exceed a threshold number of such operations.

Abstract: A system to improve memory failure management may include memory, and an error control decoder to determine failures in the memory. The system may also include an agent that may monitor failures in the memory. The system may further include a table where the error control decoder may record the failures, and where the agent can read and write to.

Abstract: A system to improve error correction includes a fast decoder to process data packets until the fast decoder finds an uncorrectable error in a data packet at which point a request for at least two data packets is generated. The system also includes a slow decoder to correct the uncorrectable error in a data packet based upon the at least two data packets.

Abstract: A system to improve miscorrection rates in error control code may include an error control decoder with a safe decoding mode that processes at least two data packets. The system may also include a buffer to receive the processed at least two data packets from the error control decoder. The error control decoder may apply a logic OR operation to the uncorrectable error signal related to the processing of the at least two data packets to produce a global uncorrectable error signal. The system may further include a recipient to receive the at least two data packets and the global uncorrectable error signal.

Abstract: A memory buffer, memory system and method for power-on initialization and test for a cascade interconnect memory system. The memory buffer includes a bus interface to links in a high-speed channel for communicating with a memory controller via a direct connection or via a cascade interconnection through an other memory buffer. The interface is operable in a SBC mode and a high-speed mode. The memory buffer also includes a field service interface (FSI) slave for receiving FSI signals from a FSI master. In addition, the memory buffer includes logic for executing a power-on and initialization training sequence initiated by the memory controller.

Abstract: 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.

Abstract: 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.

Abstract: 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.

Abstract: 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.

Abstract: A method and circuit for implementing enhanced memory reliability using memory scrub operations to determine a frequency of intermittent correctable errors, and a design structure on which the subject circuit resides are provided. A memory scrub for intermittent performs at least two reads before moving to a next memory scrub address. A number of intermittent errors is tracked where an intermittent error is identified, responsive to identifying one failing read and one passing read of the at least two reads.

Abstract: 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.

Abstract: A method and apparatus for controlling marking store updates in a central electronic complex with a plurality of core processors and eDRAM cache and interconnect bus to a service processor for loading memory controller firmware to dual-channel DDR3 memory controllers with an internal marking store. Loaded firmware of the memory controllers is responsible for tracking of ECC errors using a ECC decoder control whereby said marking store is written by a slow ECC decoder, and read by a fast ECC decoder for every read operation of said memory controllers to provide a blocking mechanism for notifying marking store firmware when the marking store has been updated and which guarantees that marking store firmware cannot write to the marking store until the marking store firmware has seen updates without causing the marking store hardware to time out.

Abstract: A method and circuit for implementing enhanced memory reliability using memory scrub operations to determine a frequency of intermittent correctable errors, and a design structure on which the subject circuit resides are provided. A memory scrub for intermittent performs at least two reads before moving to a next memory scrub address. A number of intermittent errors is tracked where an intermittent error is identified, responsive to identifying one failing read and one passing read of the at least two reads.