Advanced Micro Devices Corporation

Arm Limited

Samsung Electronics Co., Ltd.

Intel Corporation

James David Dundas

Yasuo Ishii

Marvin A Denman

Chang Joo Lee

Rajesh Bhikhubhai Patel

Abhishek Raja

Gerald David Zuraski, Jr

Anthony Jarvis

Michael Brian Schinzler

Madhu Saravana Sibi Govindan

Muhammed Umar Farooq

Shashank Nemawarkar

Adi Yoaz

Joseph Michael Pusdesris

Bradley Gene Burgess

Alexander Cole Shulyak

Muhammad Umar Farooq

Nikhil Gupta

Eric C Quinnell

Karthik Sundaram

Mukesh Patel

Jeff Rupley

Ryan J Hensley

Xiaoxin Fan

Jungsoo Kim

Fuzhou Zou

Monika Tkaczyk

Timothy Russell Snyder

Anoop Ramachandra Iyer

Gerald D Zuraski

Anthony X Jarvis

A data processing apparatus includes control flow prediction circuitry that generates a control flow prediction in respect of a group of one or more instructions. Storage circuitry used by the control flow prediction circuitry stores data in association with groups of instructions used to generate the control flow prediction for each of the groups of instructions. Control flow prediction update circuitry inserts new data into the storage circuitry in association with a new group of one or more instructions in dependence on one or more conditions being met when a miss occurs for the group of one or more instructions in the storage circuitry.

Selective control flow predictor insertion

There is provided a data processing apparatus in which receive circuitry receives a result signal from a lower level cache and a higher level cache in respect of a first instruction block. The lower level cache and the higher level cache are arranged hierarchically and transmit circuitry transmits, to the higher level cache, a query for the result signal. In response to the result signal originating from the higher level cache containing requested data, the transmit circuitry transmits a further query to the higher level cache for a subsequent instruction block at an earlier time than the further query is transmitted to the higher level cache when the result signal containing the requested data originates from the lower level cache.

Early cache querying

An apparatus comprises memory management circuitry to perform a translation table walk for a target address of a memory access request and to signal a fault in response to the translation table walk identifying a fault condition for the target address, prefetch circuitry to generate a prefetch request to request prefetching of information associated with a prefetch target address to a cache; and faulting address prediction circuitry to predict whether the memory management circuitry would identify the fault condition for the prefetch target address if the translation table walk was performed by the memory management circuitry for the prefetch target address. In response to a prediction that the fault condition would be identified for the prefetch target address, the prefetch circuitry suppresses the prefetch request and the memory management circuitry prevents the translation table walk being performed for the prefetch target address of the prefetch request.

Faulting address prediction for prefetch target address

A data processing apparatus and a method are disclosed. The data processing apparatus comprising: a prediction cache to store a plurality of prediction entries, each defining an association between a prediction cache lookup address and a predicted behaviour; prediction circuitry to select a prediction entry based on a prediction cache lookup of the prediction cache based on a given prediction cache lookup address and to determine the predicted behaviour associated with the given prediction cache lookup address based on the selected prediction entry; and a candidate prediction buffer to store a plurality of candidate predictions each indicative of a candidate prediction entry to be selected for inclusion in a subsequent prediction cache lookup, wherein the candidate prediction entry is selected in response to a candidate prediction lookup based on a candidate lookup address different to a candidate prediction cache lookup address indicated as associated with a candidate predicted behaviour in the candidate prediction entry.

Data processing apparatus and method for providing candidate prediction entries

An apparatus comprises first-level and second-level set-associative caches each comprising the same number of sets of cache entries. Indexing circuitry generates, based on a lookup address, a set index identifying which set of the first-level set-associative cache or the second-level set-associative cache is a selected set of cache entries to be looked up for information associated with the lookup address. The indexing circuitry generates the set index using an indexing scheme which maps the lookup address to the same set index for both the first-level set-associative cache and the second-level set-associative cache. This can make migration of cached information between the cache levels more efficient, which can be particularly useful for caches with high access frequency, such as branch target buffers for a branch predictor.

Set indexing for first-level and second-level set-associative cache

Micro-operations (&#x3bc;ops) are allocated into a &#x3bc;op cache by dividing, by a micro branch target buffer (&#x3bc;BTB), instructions into a first basic block in which the instructions are executed by a processing device and the first basic block corresponds to an edge of the instructions being executed by the processing device. The &#x3bc;BTB allocates the first basic block to an inverted basic block queue (IBBQ) and the IBBQ determines that the first basic block fits into the &#x3bc;op cache. The IBBQ allocates the first basic block to the &#x3bc;op cache based on a number of times the edge of the instructions corresponding to the first basic block is repeatedly executed by the processing device.

Using a graph based micro-BTB and inverted basic block queue to efficiently identify program kernels that will fit in a micro-op cache

First and second-level caches are provided. Cache control circuitry performs a first-level cache lookup of the first-level cache based on a lookup address, to determine whether the first-level cache stores valid cached data corresponding to the lookup address. When lookup hint information associated with the lookup address is available, the cache control circuitry determines based on the lookup hint information whether to activate or deactivate a second-level cache lookup of the second-level cache. The lookup hint information is indicative of whether the second-level cache is predicted to store valid cached data associated with the lookup address. When the second-level cache lookup is activated, the second-level cache lookup of the second-level cache is performed based on the lookup address to determine whether the second-level cache stores valid cached data corresponding to the lookup address.

Lookup hint information

According to one general aspect, an apparatus may include an instruction fetch unit circuit configured to retrieve instructions from a memory. The apparatus may include an instruction decode unit configured to convert instructions into one or more micro-operations that are provided to an execution unit circuit. The apparatus may also include a micro-operation cache configured to store micro-operations. The apparatus may further include a branch prediction circuit configured to: determine when a kernel of instructions is repeating, store at least a portion of the kernel within the micro-operation cache, and provide the stored portion of the kernel to the execution unit circuit without the further aid of the instruction decode unit circuit.

Micro-operation cache using predictive allocation

According to one general aspect, an apparatus may include a main-branch target buffer (BTB). The apparatus may include a micro-BTB separate from and smaller than the main-BTB, and configured to produce prediction information associated with a branching instruction. The apparatus may include a micro-BTB confidence counter configured to measure a correctness of the prediction information produced by the micro-BTB. The apparatus may further include a micro-BTB misprediction rate counter configured to measure a rate of mispredictions produced by the micro-BTB. The apparatus may also include a micro-BTB enablement circuit configured to enable a usage of the micro-BTB's prediction information, based, at least in part, upon the values of the micro-BTB confidence counter and the micro-BTB misprediction rate counter.

Method and apparatus to control the use of hierarchical branch predictors based on the effectiveness of their results

Embodiments include a micro BTB, which can predict up to two branches per cycle, every cycle, with zero bubble insertion on either a taken or not taken prediction, thereby significantly improving performance and reducing power consumption of a microprocessor. A front end of a microprocessor can include a main front end logic section having a main BTB, a micro BTB to produce prediction information, and a decoupling queue. The micro BTB can include a graph having multiple entries, and a CAM having multiple items. Each of the entries of the graph can include a link pointer to a next branch in a taken direction, and a link pointer to a next branch in a not-taken direction. The micro BTB can insert a hot branch into the graph as a new seed.

High performance zero bubble conditional branch prediction using micro branch target buffer

A method and apparatus to utilize a fetching scheme for instructions in a processor to limit the expenditure of power caused by the speculative execution of branch instructions is provided. Also provided is a computer readable storage device encoded with data for adapting a manufacturing facility to create an apparatus. The method includes calculating a cumulative confidence measure based on one or more outstanding conditional branch instructions. The method also includes reducing prefetching operations in response to detecting that the cumulative confidence measure is below a first threshold level.

Cumulative confidence fetch throttling

A method and apparatus for branch prediction is disclosed. A pattern history table (PHT) is accessed based on at least one global history value to obtain a prediction value. The prediction value and the at least one global history value used to obtain the prediction value are placed in a queue. If a branch prediction is requested, the queue is accessed to obtain a prediction value. The queue may include any number of entries and the queue maintains the oldest prediction value at the head of the queue. The prediction value at the head of the queue is used when a branch prediction is needed.

Power efficient pattern history table fetch in branch predictor

The present application describes a method and apparatus for prefetching instructions based on predicted branch target addresses. Some embodiments of the method include providing a second cache line to a second cache when a target address for a branch instruction in a first cache line of a first cache is included in the second cache line of the first cache and when the second cache line is not resident in the second cache.

Pre-fetching instructions using predicted branch target addresses

A method and apparatus for eliminating fetch cancels for inclusive caches are presented. Some embodiments of the apparatus include a first cache configurable to issue fetch or prefetch requests to a second cache that is inclusive of said at least one first cache. The first cache is not permitted to cancel issued fetch or prefetch requests to the second cache. Some embodiments of the method include preventing the first cache(s) from canceling issued fetch or prefetch requests to a second cache that is inclusive of the first cache(s).

Eliminating fetch cancel for inclusive caches

A method and apparatus to utilize a branch prediction scheme that limits the expenditure of power and the area consumed caused by branch prediction schemes is provided. The method includes accessing a first entry and a second entry of the data structure, wherein each entry stores a portion of a predicted target address, determining the predicted target address using the portion of the predicted target address stored in the first entry and a portion of a branch address of a fetched branch instruction for a fetched branch instruction of a first type, and determining the predicted target address using the portion of the predicted target address stored in the first entry and the portion of the predicted target address stored in the second entry for a fetched branch instruction of a second type.

Generating predicted branch target address from two entries storing portions of target address based on static/dynamic indicator of branch instruction type

Growing community of inventors

Austin, TX, United States of America

James David Dundas