International Business Machines Corporation

Barry Watson Krumm

Charles Franklin Webb

Chung-Lung K Shum

Mark S Farrell

Aaron Tsai

John Stephen Liptay

Jennifer A Navarro

Christian Jacobi

Timothy John Slegel

Brian David Barrick

Pak-kin Mak

Michael H Wood

Mark Anthony Check

Bradley Nelson

Karl J Duvalsaint

James R Mitchell

Donald W McCauley

Jennifer Serena Navarro

Brian Robert Prasky

Martin Recktenwald

Dean G Bair

Joshua M Weinberg

Steven QiHong Ying

Patrick James Meaney

Philip George Emma

Michael F Fee

Yuen H Chan

Mary Prilotski Kusko

Gregory William Alexander

Norman Jay Rohrer

Erwin F Pfeffer

Peter Anthony Sandon

Paul David Kartschoke

Hans-Werner Tast

Jane Helen Bartik

Moon Ju Kim

Jin Song Ji

Bryan J Robbins

Peter H Gum

Ching-Farn Eric Wu

Steven T Comfort

Allen Prescott Haar

John Fenton Scanlon

Debbie A St Onge

David Man Wong

Steven B Risch

Data operand fetching control includes a computer processor that includes a control unit for determining memory access operations. The control unit is configured to perform a method. The method includes calculating a summation weight value for each instruction in a pipeline, the summation weight value calculated as a function of branch uncertainty and a pendency in which the instruction resides in the pipeline relative to other instructions in the pipeline. The method also includes mapping the summation weight value of a selected instruction that is attempting to access system memory to a memory access control, each memory access control specifying a manner of handling data fetching operations. The method further includes performing a memory access operation for the selected instruction based upon the mapping.

Operand fetching control as a function of branch confidence

Data operand fetching control includes calculating a summation weight value for each instruction in a pipeline, the summation weight value calculated as a function of branch uncertainty and a pendency in which the instruction resides in the pipeline relative to other instructions in the pipeline. The data operand fetching control also includes mapping the summation weight value of a selected instruction that is attempting to access system memory to a memory access control. Each memory access control specifies a manner of handling data fetching operations. The data operand fetching control further includes performing a memory access operation for the selected instruction based upon the mapping.

Embodiments relate to loading data in a pipelined microprocessor. An aspect includes issuing a load request that comprises a load address requiring at least one block of data the same size as a largest contiguous granularity of data returned from a cache. Another aspect includes determining that the load address matches at least one block address. Another aspect includes, based on determining that there is an address match, reading a data block from a buffer register and sending the data to satisfy the load request; comparing a unique set id of the data block to the set id of the matching address after sending the data block; based on determining that there is a set id match, continuing the load request, or, based on determining that there is not a set id match, setting a store-forwarding state of the matching address to no store-forwarding and rejecting the load request.

Store data forwarding with no memory model restrictions

A pipelined microprocessor includes circuitry for store forwarding by performing: for each store request, and while a write to one of a cache and a memory is pending; obtaining the most recent value for at least one complete block of data; merging store data from the store request with the complete block of data thus updating the block of data and forming a new most recent value and an updated complete block of data; and buffering the updated complete block of data into a store data queue; for each load request, where the load request may require at least one updated completed block of data: determining if store forwarding is appropriate for the load request on a block-by-block basis; if store forwarding is appropriate, selecting an appropriate block of data from the store data queue on a block-by-block basis; and forwarding the selected block of data to the load request.

A pipelined processor includes circuitry adapted for store forwarding, including: for each store request, and while a write to one of a cache and a memory is pending; obtaining the most recent value for at least one block of data; merging store data from the store request with the block of data thus updating the block of data and forming a new most recent value and an updated complete block of data; and buffering the updated block of data into a store data queue; for each additional store request, where the additional store request requires at least one updated block of data: determining if store forwarding is appropriate for the additional store request on a block-by-block basis; if store forwarding is appropriate, selecting an appropriate block of data from the store data queue on a block-by-block basis; and forwarding the selected block of data to the additional store request.

Microprocessor and method for deferred store data forwarding for store background data in a system with no memory model restrictions

A test method and apparatus for randomly testing logic structures. The method includes identifying and analyzing a functional behavior of a logic structure to be covered during the random testing, modifying the logic structure such that the logic structure behaves in a functional manner during random testing, and generating patterns to exercise the modified logic structure.

Method and apparatus for improving random pattern testing of logic structures

Method and system for a multi-level virtual/real cache system with synonym resolution. An exemplary embodiment includes a multi-level cache hierarchy, including a set of L1 caches associated with one or more processor cores and a set of L2 caches, wherein the set of L1 caches are a subset of the set of L2 caches, wherein the set of L1 caches underneath a given L2 cache are associated with one or more of the processor cores.

Method and system for a multi-level virtual/real cache system with synonym resolution

A method for handling cache coherency includes allocating a tag when a cache line is not exclusive in a data cache for a store operation, and sending the tag and an exclusive fetch for the line to coherency logic. An invalidation request is sent within a minimum amount of time to an I-cache, preferably only if it has fetched to the line and has not been invalidated since, which request includes an address to be invalidated, the tag, and an indicator specifying the line is for a PSC operation. The method further includes comparing the request address against stored addresses of prefetched instructions, and in response to a match, sending a match indicator and the tag to an LSU, within a maximum amount of time. The match indicator is timed, relative to exclusive data return, such that the LSU can discard prefetched instructions following execution of the store operation that stores to a line subject to an exclusive data return, and for which the match is indicated.

Method and system for handling cache coherency for self-modifying code

There is provided a method, system and computer program product for generating trace data related to a data processing system event. The method includes: receiving an instruction relating to the system event from a location in the system; determining a minimum number of trace segment records required to record instruction information; and creating a trace segment table including the number of trace segment records, the number of trace segment records including at least one instruction record.

Method, system and computer program product for generating trace data

A computer system having a plurality of processors where each of the processors is dynamically assigned for execution of tasks based upon either performance or reliability.

Method and apparatus for dynamically allocating processors

A central processor uses virtual addresses to access data via cache logic including a DAT and ART, and the cache logic accesses data in the hierarchical storage subsystem using absolute addresses to access data, a part of the first level of the cache memory includes a translator for virtual or real addresses to absolute addresses. When requests are sent for a data fetch and the requested data are not resident in the first level of cache the request for data is delayed and may be forwarded to a lower level of said hierarchical memory, and a delayed request may result in cancellation of any process during a delayed request that has the ability to send back an exception. A delayed request may be rescinded if the central processor has reached an interruptible stage in its pipeline logic at which point, a false exception is forced clearing all I the wait states while the central processor ignores the false exception. Forcing of an exception occurs during dynamic address translation (DAT) or during access register translation (ART). A request for data signal to the storage subsystem cancellation is settable by the first hierarchical level of cache logic. A false exception signal to the first level cache is settable by the storage subsystem logic.

False exception for cancelled delayed requests

A pipelined multiprocessor system for ESA/390 operations which executes a simple instruction set in a hardware controlled execution unit and executes a complex instruction set in a milli-mode architected state with a millicode sequence of simple instructions in the hardware controlled execution unit, comprising a plurality of CPU processors each of which is part of said multiprocessing system and capable of generating and responding to a quiesce request, and controls for system operations which allow the CPUs in the ESA/390 system to process the local buffer update portion of IPTE and SSKE operations without waiting for all other processors to reach an interruptible point, and then to continue program execution with minor temporary restrictions on operations until the IPTE or SSKE operation is globally completed. In addition, Licensed Internal Code (LIC) sequences are defined which allow these IPTE and SSKE operations to co-exist with other operations which require conventional system quiescing (i.e. all processors must pause together), and to allow for CPU retry actions on any of the CPUs in the system at any point in the operation.

System serialization with early release of individual processor

SMP computers systems can add to the first level cache a fill mode latch and achieve straightforward, high-performance loading of a writable cache code array that is part of a hierarchical cache structure.A new code array's write control elements include a control latch called 'fill mode' for the BCE controls which when fill mode is active, then a disable is also active, since reads of the code array may not give accurate data when the array is not yet filled-up/fully valid. New mode follows the sequential steps which process code by:

System speed loading of a writable cache code array

Multiprocessor serialization with early release of processors

A computer for executing programs and having a structure for fetching instructions and/or operands along a path which may not be taken by a process being executed by a computer processor having a hierarchical memory structure with data being loaded into cache lines of a cache in the structure, and having block line fetch signal selection logic and computational logic with hedge selection logic for generating line fetch block signals for control of hedging by fetching instructions and/or operands along a path which may not be taken by a process being executed and making selected hedge fetches sensitive to whether the data is in the cache so as to gain the best performance advantage with a selected hedge fetch signal which accompanies each fetch request to the cache to identify whether a line should be loaded if it misses the cache to indicate a selected hedge fetch when this signal is ON, and rejecting a fetch request in the event the selected hedge fetch signal is turned ON if the data is not in the cache, the cache will reject the fetch, and thereafter repeating the fetch request after a fetch request has been rejected when the selected hedge fetch signal was turned ON the data was not in the cache to repeat the fetch request at a later time when it is more certain that the process being executed wants the data, or never repeating the request upon determination that the process being executed does not need the data to he fetched.

Computer with optimizing hardware for conditional hedge fetching into

Growing community of inventors

Poughkeepsie, NY, United States of America

Barry Watson Krumm