Advanced Micro Devices Corporation

Arm Limited

Matthew Andrew Rafacz

Gregory William Smaus

Matthew M Crum

Michael David Achenbach

Yasuo Ishii

Michael Filippo

Emil Talpes

Miles Robert Dooley

John M King

Francesco Spadini

Huzefa Moiz Sanjeliwala

Christopher J Burke

Kevin Michael Lepak

Abhishek Raja

Guillaume Bolbenes

Houdhaifa Bouzguarrou

Michael T Clark

Klas Magnus Bruce

Michael E Tuuk

Leigang Kou

Noah Bamford

Kevin Lepaak

A data processing apparatus is provided. Cache circuitry caches data, the data being indexed according to execution contexts of processing circuitry. Receive circuitry receives invalidation requests each referencing a specific execution context in the execution contexts. Invalidation circuitry invalidates at least some of the data in the cache circuitry and filter circuitry filters the invalidation requests based on at least one condition and, when the condition is met, causes the invalidation circuitry to invalidate the data in the cache circuitry.

Filtering invalidation requests

Techniques for selectively executing a lock instruction speculatively or non-speculatively based on lock address prediction and/or temporal lock prediction. including methods an devices for locking an entry in a memory device. In some techniques, a lock instruction executed by a thread for a particular memory entry of a memory device is detected. Whether contention occurred for the particular memory entry during an earlier speculative lock is detected on a condition that the lock instruction comprises a speculative lock instruction. The lock is executed non-speculatively if contention occurred for the particular memory entry during an earlier speculative lock. The lock is executed speculatively if contention did not occur for the particular memory entry during an earlier speculative lock.

Lock address contention predictor

An arithmetic unit performs store-to-load forwarding based on predicted dependencies between store instructions and load instructions. In some embodiments, the arithmetic unit maintains a table of store instructions that are awaiting movement to a load/store unit of the instruction pipeline. In response to receiving a load instruction that is predicted to be dependent on a store instruction stored at the table, the arithmetic unit causes the data associated with the store instruction to be placed into the physical register targeted by the load instruction. In some embodiments, the arithmetic unit performs the forwarding by mapping the physical register targeted by the load instruction to the physical register where the data associated with the store instruction is located.

Store-to-load forwarding

A data processing apparatus is provided, which includes a cache to store operations produced by decoding instructions fetched from memory. The cache is indexed by virtual addresses of the instructions in the memory. Receiving circuitry receives an incoming invalidation request that references a physical address in the memory. Invalidation circuitry invalidates entries in the cache where the virtual address corresponds with the physical address. Coherency is thereby achieved when using a cache that is indexed using virtual addresses.

Instruction cache coherence

A data processing apparatus is provided that includes a plurality of storage elements. Receiving circuitry receives a plurality of incoming data beats from cache circuitry and stores the incoming data beats in the storage elements. At least one existing data beat in the storage elements is replaced by an equal number of the incoming data beats belonging to a different cache line of the cache circuitry. The existing data beats stored in said plurality of storage elements form an incomplete cache line.

Cache storage

A data processing apparatus is provided including queue circuitry to respond to control signals each associated with a memory access instruction, and to queue a plurality of requests for data, each associated with a reference to a storage location. Resolution circuitry acquires a request for data, and issues the request for data, the resolution circuitry having a resolution circuitry limit. When a current capacity of the resolution circuitry is below the resolution circuitry limit, the resolution circuitry acquires the request for data by receiving the request for data from the queue circuitry, stores the request for data in association with the storage location, issues the request for data, and causes a result of issuing the request for data to be provided to said storage location. When the current capacity of the resolution circuitry meets or exceeds the resolution circuitry limit, the resolution circuitry acquires the request for data by examining a next request for data in the queue circuitry and issues a further request for the data based on the request for data.

Queuing memory access requests

Techniques for improving execution of a lock instruction are provided herein. A lock instruction and younger instructions are allowed to speculatively retire prior to the store portion of the lock instruction committing its value to memory. These instructions thus do not have to wait for the lock instruction to complete before retiring. In the event that the processor detects a violation of the atomic or fencing properties of the lock instruction prior to committing the value of the lock instruction, the processor rolls back state and executes the lock instruction in a slow mode in which younger instructions are not allowed to retire until the stored value of the lock instruction is committed. Speculative retirement of these instructions results in increased processing speed, as instructions no longer need to wait to retire after execution of a lock instruction.

Speculative retirement of post-lock instructions

A processor includes a store queue that stores information representing store instructions. In response to retirement of a store instruction, the processor invalidates the corresponding entry in the store queue, thereby indicating that the entry is available to store a subsequent store instruction. The store address is not removed from the queue until the subsequent store instruction is stored. Accordingly, the store address is available for comparison to a dependent load address.

Confirming store-to-load forwards

A system and method for copying and initializing a block of memory. To copy several data entities from a source region of memory to a destination region of memory, an instruction may copy each data entity one at a time. If an aggregate condition is determined to be satisfied, multiple data entities may be copied simultaneously. The aggregate condition may rely on an aggregate data size, the size of the data entities to be copied, and the alignment of the source and destination addresses.

Growing community of inventors

Austin, TX, United States of America

Matthew Andrew Rafacz