Advanced Micro Devices Corporation

Other

John M King

Gregory William Smaus

David A Kaplan

Michael Estlick

Matthew Andrew Rafacz

Kevin Michael Lepak

Anthony J Bybell

Michael T Clark

Tarun Nakra

Michael David Achenbach

Erik D Swanson

Kai Troester

Matthew T Sobel

Matthew M Crum

Magiting M Talisayon

Scott Thomas Bingham

Jeff Rupley

Daniel Hopper

Robert Krick

Eben V Fodor

Noah Bamford

Folly F King

Robert Weidner

Kevin Lepaak

A buffer of a processing system allows younger stores to write to a data cache before an older store completes its write operation to the data cache while maintaining the appearance of committing stores in program order. To maintain the appearance that a blocked store completed its write operation to the data cache, the processing system cancels the blocked store while 'locking' the cache line in the data cache in an exclusive state to which the blocked store is attempting to write. The data cache negatively acknowledges any probes to the cache line until the blocked store has completed the write operation. The buffer thus decouples completing the write operation from global observability of the write operation.

Pseudo out-of-order store commit

The disclosed computer-implemented method for tracking miss requests using data cache tags can include generating a data cache miss request associated with data requested in connection with a cacheline and allocating a miss address buffer entry for the miss request. Additionally, the method can include, setting a fill-pending flag associated with the cacheline in response to the data associated with the data cache miss request being absent from a first data cache, and de-allocating the miss address buffer entry. In the event that another load or store operation requests the same data associated with the cacheline while the fill-pending flag is set, the method can include monitoring for a fill response associated with the miss request until the fill response is received. Upon receipt of the fill response, the method can include re-setting the fill-pending flag associated with the cacheline.

Systems and methods for tracking data cache miss requests with data cache tags

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

A processor includes a load/store unit and an execution pipeline to execute an instruction that represents a single-instruction-multiple-data (SIMD) operation, and which references a memory block storing operand data for one or more lanes of a plurality of lanes and a mask vector indicating which lanes of a plurality of lanes are enabled and which are disabled for the operation. The execution pipeline executes an instruction in a first execution mode unless a memory fault is generated during execution of the instruction in the first execution mode. In response to the memory fault, the execution pipeline re-executes the instruction in a second execution mode. In the first execution mode, a single load operation is attempted to access the memory block via the load/store unit. In the second execution mode, a separate load operation is performed by the load/store unit for each enabled lane of the plurality of lanes prior to executing the SIMD operation.

Masked multi-lane instruction memory fault handling using fast and slow execution paths

An apparatus includes a plurality of load buses and a load store unit that includes a plurality of load ports to access the plurality of load buses. The load store unit performs a gather operation to concurrently gather a plurality of subsets of data from a memory via the plurality of load buses in a first mode. The apparatus also includes a register that is partitioned into a plurality of portions to hold the plurality of subsets of data provided by the load store unit. The load store unit ignores exceptions or faults while performing the gather operation in the first mode and transitions to a second mode in response to an exception or fault. Two lanes are dispatched to concurrently perform the gather operation per clock cycle in the first mode and a single lane is dispatched to perform the gather operation per clock cycle in the second mode.

Multi-modal gather operation

A system and method for load fusion fuses small load operations into fewer, larger load operations. The system detects that a pair of adjacent operations are consecutive load operations, where the adjacent micro-operations refers to micro-operations flowing through adjacent dispatch slots and the consecutive load micro-operations refers to both of the adjacent micro-operations being load micro-operations. The consecutive load operations are then reviewed to determine if the data sizes are the same and if the load operation addresses are consecutive. The two load operations are then fused together to form one load micro-operation with twice the data size and one load data micro-operation with no load component.

System and method for load fusion

The techniques described herein improve cache traffic performance in the context of contended lock instructions. More specifically, each core maintains a lock address contention table that stores addresses corresponding to contended lock instructions. The lock address contention table also includes a state value that indicates progress through a series of states meant to track whether a load by the core in a spin-loop associated with semaphore acquisition has obtained the semaphore in an exclusive state. Upon detecting that a load in a spin-loop has obtained the semaphore in an exclusive state, the core responds to incoming requests for access to the semaphore with negative acknowledgments. This allows the core to maintain the semaphore cache line in an exclusive state, which allows it to acquire the semaphore faster and to avoid transmitting that cache line to other cores unnecessarily.

Techniques for handling cache coherency traffic for contended semaphores

A system and method for a lightweight fence is described. In particular, micro-operations including a fencing micro-operation are dispatched to a load queue. The fencing micro-operation allows micro-operations younger than the fencing micro-operation to execute, where the micro-operations are related to a type of fencing micro-operation. The fencing micro-operation is executed if the fencing micro-operation is the oldest memory access micro-operation, where the oldest memory access micro-operation is related to the type of fencing micro-operation. The fencing micro-operation determines whether micro-operations younger than the fencing micro-operation have load ordering violations and if load ordering violations are detected, the fencing micro-operation signals the retire queue that instructions younger than the fencing micro-operation should be flushed. The instructions to be flushed should include all micro-operations with load ordering violations.

System and method for a lightweight fencing operation

A technique for performing store-to-load forwarding is provided. The technique includes determining a virtual address for data to be loaded for the load instruction, identifying a matching store instruction from one or more store instruction memories by comparing a virtual-address-based comparison value for the load instruction to one or more virtual-address-based comparison values of one or more store instructions, determining a physical address for the load instruction, and validating the load instruction based on a comparison between the physical address of the load instruction and a physical address of the matching store instruction.

Techniques for performing store-to-load forwarding

Methods, devices, and systems for determining an address in a physical memory which corresponds to a virtual address using a skewed-associative translation lookaside buffer (TLB) are described. A virtual address and a configuration indication are received using receiver circuitry. A physical address corresponding to the virtual address is output if a TLB hit occurs. A first subset of a plurality of ways of the TLB is configured to hold a first page size. The first subset includes a number of the ways based on the configuration indication. A physical address corresponding to the virtual address is retrieved from a page table if a TLB miss occurs, and at least a portion of the physical address is installed in a least recently used way of a subset of a plurality of ways the TLB, determined according to a replacement policy based on the configuration indication.

Configurable skewed associativity in a translation lookaside buffer

A system and method for load queue (LDQ) and store queue (STQ) entry allocations at address generation time that maintains age-order of instructions is described. In particular, writing LDQ and STQ entries are delayed until address generation time. This allows the load and store operations to dispatch, and younger operations (which may not be store and load operations) to also dispatch and execute their instructions. The address generation of the load or store operation is held at an address generation scheduler queue (AGSQ) until a load or store queue entry is available for the operation. The tracking of load queue entries or store queue entries is effectively being done in the AGSQ instead of at the decode engine. The LDQ and STQ depth is not visible from a decode engine's perspective, and increases the effective processing and queue depth.

System and method for load and store queue allocations at address generation time

A system and method for a virtual load queue is described. Load micro-operations are processed through an instruction pipeline without requiring an entry in a load queue (LDQ). An address generation scheduler queue (AGSQ) entry is allocated to the load micro-operation and a LDQ entry is not allocated to the load micro-operation. The LDQ entries are reserved for the N oldest load micro-operations, where N is the depth of the LDQ. Deallocation of the AGSQ entry is done if the load micro-operation is one of the N oldest load micro-operations, or upon successful completion of the load micro-operation. Deallocation of the AGSQ entry is not done if the load micro-operation gets a bad status and is not one of the N oldest micro-operations. Consequently, the AGSQ acts as a virtual queue for the LDQ and mitigates the limiting effect of the LDQ depth.

System and method for processing a load micro-operation by allocating an address generation scheduler queue entry without allocating a load queue entry

Described herein is a system and method for store fusion that fuses small store operations into fewer, larger store operations. The system detects that a pair of adjacent operations are consecutive store operations, where the adjacent micro-operations refers to micro-operations flowing through adjacent dispatch slots and the consecutive store micro-operations refers to both of the adjacent micro-operations being store micro-operations. The consecutive store operations are then reviewed to determine if the data sizes are the same and if the store operation addresses are consecutive. The two store operations are then fused together to form one store operation with twice the data size and one store data HI operation.

System and method for store fusion

Embodiments herein provide for improved store-to-load-forwarding (STLF) logic and linear aliasing effect reduction logic. In one embodiment, a load instruction to be executed is selected. Whether a first linear address associated with said load instruction matches a linear address of a store instruction of a plurality of store instructions in a queue is determined. Data associated with said store instruction for executing said load instruction is forwarded, in response to determining that the first linear address matches the linear address of the store instruction.

Growing community of inventors

Austin, TX, United States of America

John M King