International Business Machines Corporation

Gary E Strait

Pak-kin Mak

Robert J Sonnelitter, Iii

Matthias Klein

Garrett M Drapala

Michael F Fee

Michael Andrew Blake

Hanno Ulrich

Patrick James Meaney

Craig Raymond Walters

Ekaterina M Ambroladze

Mark Anthony Check

Norbert Hagspiel

Deanna Postles Dunn

Thomas Anthony Gregg

Deanna Postles Dunn Berger

Arthur J O'Neill

Sascha Junghans

Diana Lynn Orf

Christine Comins Jones

Kathryn Marie Jackson

Hong Deng

Joshua D Massover

Mark S Farrell

David F Craddock

Charles Franklin Webb

Gary A Van Huben

Ulrich Mayer

Timothy Carl Bronson

Janet R Easton

Kulwant M Pandey

William Wu Shen

Don Coppersmith

John Cocke

Adrian Eric Seigler

Guy G Tracy

Douglas G Balazich

Carl B Ford, Iii

Poornima P Sulibele

Richard Fairbanks Arnold

Michael D Campbell

Alia Shah

Timothy W Steele

Sajay Mathew Karottukottarathil Baby

Methods, systems, and computer program products for queue management are provided. Aspects include receiving a first queue entry and storing the first queue entry in a queue at a first location, wherein the first queue entry includes a first target destination, receiving a second queue entry and storing the second queue entry in the queue at a second location, wherein the second queue entry includes a second target destination, tracking a relative age for each of the first queue entry and the second queue entry, transmitting the first queue entry to the first target destination based at least in part on the relative age for the first queue entry being greater than the relative age for the second queue entry, and receiving a third queue entry and storing the third queue entry in the queue at the first location.

Queue management for multiway queues

Embodiments are directed to methodologies for efficiently sampling data for analysis by a pipeline analysis algorithm. The amount of sampled data is maximized without increasing sampling overhead by sampling 'non-pipeline activity' data if the subject pipeline is inactive during the sampling time. The non-pipeline activity data is selected to include overall system information that is relevant to the subject pipeline's performance but is not necessarily dependent on whether the subject pipeline is active. In some embodiments, the non-pipeline activity data allows for confirmation of a pipeline performance characteristic that must otherwise be inferred by the subsequent pipeline analysis algorithm from data sampled while the pipeline was active. In some embodiments, the non-pipeline activity data allows the pipeline analysis algorithm to analyze additional performance characteristics that cannot otherwise be inferred from the data sampled while the pipeline was active.

Dual/multi-mode processor pipelines sampling

A method, computer program product, and system for maintaining a proper ordering of a data steam that includes two or more sequentially ordered stores, the data stream being moved to a destination memory device, the two or more sequentially ordered stores including at least a first store and a second store, wherein the first store is rejected by the destination memory device. A computer-implemented method includes sending the first store to the destination memory device. A conditional request is sent to the destination memory device for approval to send the second store to the destination memory device, the conditional request dependent upon successful completion of the first store. The second store is cancelled responsive to receiving a reject response corresponding to the first store.

Increased bandwidth of ordered stores in a non-uniform memory subsystem

A computer implemented method for avoiding false activation of hang avoidance mechanisms of a system is provided. The computer implemented method includes receiving, by a nest of the system, rejects from a processor core of the system. The rejects are issued based on a cache line being locked by the processor core. The computer implemented method includes accumulating the rejects by the nest. The computer implemented method includes determining, by the nest, when an amount of the rejects accumulated by the nest has met or exceeded a programmable threshold. The computer implemented method also includes triggering, by the nest, a global reset to counters of the hang avoidance mechanisms of a system in response to the amount meeting or exceeding the programmable threshold.

Deadlock avoidance in a multi-processor computer system with extended cache line locking

Embodiments are directed to systems and methodologies for efficiently sampling data for analysis by a pipeline analysis algorithm. The amount of sampled data is maximized without increasing sampling overhead by sampling 'non-pipeline activity' data if the subject pipeline is inactive during the sampling time. The non-pipeline activity data is selected to include overall system information that is relevant to the subject pipeline's performance but is not necessarily dependent on whether the subject pipeline is active. In some embodiments, the non-pipeline activity data allows for confirmation of a pipeline performance characteristic that must otherwise be inferred by the subsequent pipeline analysis algorithm from data sampled while the pipeline was active. In some embodiments, the non-pipeline activity data allows the pipeline analysis algorithm to analyze additional performance characteristics that cannot otherwise be inferred from the data sampled while the pipeline was active.

Dual/multi-mode processor pipeline sampling

Maintaining store order with high throughput in a distributed shared memory system. A request is received for a first ordered data store and a coherency check is initiated. A signal is sent that pipelining of a second ordered data store can be initiated. If a delay condition is encountered during the coherency check for the first ordered data store, rejection of the first ordered data store is signaled. If a delay condition is not encountered during the coherency check for the first ordered data store, a signal is sent indicating a readiness to continue pipelining of the second ordered data store.

Maintaining order with parallel access data streams

Embodiments relate to matrix and compression-based error detection. An aspect includes summing, by each of a first plurality of summing modules of a first compressor, a respective row of a matrix, the matrix comprising a plurality of rows and a plurality of columns of output bits of a circuit under test wherein each output bit of the circuit under test comprises an element of the matrix, and is a member of a row of a column that is orthogonal to the row. Another aspect includes summing, by each of a second plurality of summing modules of a second compressor, a respective column of output bits of the matrix. Yet another aspect includes determining a presence of an error in the circuit under test based at least one of an output of the first compressor and an output of the second compressor.

Matrix and compression-based error detection

A technique for error detection is provided. A controller is configured to detect errors by using error correcting code (ECC), and a cache includes independent ECC words for storing data. The controller detects the errors in the ECC words for a wordline that is read. The controller detects a first error in a first ECC word on the wordline and a second error in a second ECC word on the wordline. The controller determines that the wordline is a failing wordline based on detecting the first error in the first ECC word and the second error in the second ECC word.

Bad wordline/array detection in memory

A system and computer implemented method for storing of data in the memory of a computer system in order at a fast rate is provided. The method includes launching a first store to memory. A wait counter is initiated. A second store to memory is speculatively launched when the wait counter expires. The second store to memory is cancelled when the second store achieves coherency prior to the first store to memory.

Memory ordered store system in a multiprocessor computer system

A system and computer implemented method for isolating errors in a computer system is provided. The method includes receiving a direct memory access (DMA) command to access a computer memory, a read response, or an interrupt; associating the DMA command to access the computer memory, the read response, or the interrupt with a stream identified by a stream identification (ID); detecting a memory error caused by the DMA command in the stream, the memory error resulting in stale data in the computer memory; and isolating the memory error in the stream associated with the stream ID from other streams associated with other stream IDs upon detecting the memory error.

Memory error isolation and recovery in a multiprocessor computer system

A method for merging data including receiving a request from an input/output device to merge a data, wherein a merge of the data includes a manipulation of the data, determining that the data exists in a local cache memory that is in local communication with the input/output device, fetching the data to the local cache memory from a remote cache memory or a main memory if the data does not exist in the local cache memory, merging the data according to the request to obtain a merged data, and storing the merged data in the local cache, wherein the merging of the data is performed without using a memory controller within a control flow or a data flow of the merging of the data.

Merging data in an L2 cache memory

Systems, methods and computer program products for data buffers partitioned from a cache array. An exemplary embodiment includes a method in a processor and for providing data buffers partitioned from a cache array, the method including clearing cache directories associated with the processor to an initial state, obtaining a selected directory state from a control register preloaded by the service processor, in response to the control register including the desired cache state, sending load commands with an address and data, loading cache lines and cache line directory entries into the cache and storing the specified data in the corresponding cache line.

Method, system and computer program product for data buffers partitioned from a cache array

Growing community of inventors

Poughkeepsie, NY, United States of America

Gary E Strait