Arm Limited

Alexander Alfred Hornung

Ian Michael Caulfield

Jose Gonzalez-Gonzalez

Richard Roy Grisenthwaite

Simon John Craske

Andrew Brookfield Swaine

Bruce James Mathewson

Abhishek Raja

Phanindra Kumar Mannava

Alex James Waugh

Klas Magnus Bruce

Michael Andrew Campbell

Max John Batley

Kauser Yakub Johar

Gregory Andrew Chadwick

Adrian Viorel Popescu

Lei Ma

Kenny Ju Min Yeoh

Roberto Gattuso

Prefetch circuitry generates, based on stream prefetch state information, prefetch requests for prefetching data to at least one cache. Cache control circuitry controls, based on cache policy information associated with cache entries in a given level of cache, at least one of cache entry replacement in the given level of cache, and allocation of data evicted from the given level of cache to a further level of cache. The stream prefetch state information specifies, for at least one stream of addresses, information representing an address access pattern for generating addresses to be specified by a corresponding series of prefetch requests. Cache policy information for at least one prefetched cache entry of the given level of cache (to which data is prefetched for a given stream of addresses) is set to a value dependent on at least one stream property associated with the given stream of addresses.

Setting cache policy information for prefetched cache entry

An apparatus supports decoding and execution of a bulk memory instruction specifying a block size parameter. The apparatus comprises control circuitry to determine whether the block size corresponding to the block size parameter exceeds a predetermined threshold, and performs a micro-architectural control action to influence the handling of at least one bulk memory operation by memory operation processing circuitry. The micro-architectural control action varies depending on whether the block size exceeds the predetermined threshold, and further depending on the states of other components and operations within or coupled with the apparatus. The micro-architectural control action could include an alignment correction action, cache allocation control action, or processing circuitry selection action.

Control of bulk memory instructions

An apparatus comprising a cache comprising a plurality of cache entries, cache access circuitry responsive to a cache access request to perform, based on a target memory address associated with the cache access request, a cache lookup operation, tracking circuitry to track pending requests to modify cache entries of the cache, and prediction circuitry responsive to the cache access request to make a prediction of whether the pending requests tracked by the tracking circuitry include a pending request to modify a cache entry associated with the target memory address, wherein the cache access circuitry is responsive to the cache access request to determine, based on the prediction, whether to perform an additional lookup of the tracking circuitry. A method and a non-transitory computer-readable medium to store computer-readable code for fabrication of the apparatus are also provided.

Determining whether to perform an additional lookup of tracking circuitry

Requester circuitryissues an access request specifying a target physical address (PA) and a target physical address space (PAS) identifier identifying a target PAS. Prior to a point of physical aliasing (PoPA), a pre-PoPA memory system componenttreats aliasing PAs from different PASs which actually correspond to the same memory system resource as if they correspond to different memory system resources. A post-PoPA memory system componenttreats the aliasing PAs as referring to the same memory system resource. When the target PA and target PAS of a read-if-hit-pre-PoPA request hit in a pre-PoPA cache, a data response is returned to the requester circuitry. If the read-if-hit-pre-PoPA request misses in the pre-PoPA cache, a no-data response is returned. The read-if-hit-pre-PoPA request is safe to issue speculatively while waiting for security checks to be performed, improving performance.

Read-if-hit-pre-PoPA request

An apparatus comprises instruction fetch circuitry to retrieve instructions from storage and branch target storage to store entries comprising source and target addresses for branch instructions. A confidence value is stored with each entry and when a current address matches a source address in an entry, and the confidence value exceeds a confidence threshold, instruction fetch circuitry retrieves a predicted next instruction from a target address in the entry. Branch confidence update circuitry increases the confidence value of the entry on receipt of a confirmation of the target address and decreases the confidence value on receipt of a non-confirmation of the target address. When the confidence value meets a confidence lock threshold below the confidence threshold and non-confirmation of the target address is received, a locking mechanism with respect to the entry is triggered. A corresponding method is also provided.

Branch prediction in a data processing apparatus

An apparatus comprises processing circuitry to issue demand memory access requests to access data stored in a memory system. Stride pattern detection circuitry detects whether a sequence of demand target addresses specified by the demand memory access requests includes two or more constant stride sequences of addresses interleaved within the sequence of demand target addresses. Each constant stride sequence comprises addresses separated by intervals of a constant stride value. Prefetch control circuitry controls issuing of prefetch load requests to prefetch data from the memory system. The prefetch load requests specify prefetch target addresses predicted based on the constant stride sequences detected by the stride pattern detection circuitry.

Prefetching based on detection of interleaved constant stride sequences of addresses within a sequence of demand target addresses

Aspects of the present disclosure relate to an apparatus comprising decode circuitry to receive an instruction and identify the received instruction as a load instruction, and prediction circuitry to predict, based on a prediction scheme, a target address of the load instruction, and trigger a speculative memory access in respect of the predicted target address.

Methods and apparatus for handling processor load instructions

A requester issues a request specifying a target address indicating an addressed location in a memory system. A completer responds to the request. Tag error checking circuitry performs a tag error checking operation when the request issued by the requester is a tag-error-checking request specifying an address tag. The tag error checking operation comprises determining whether the address tag matches an allocation tag stored in the memory system associated with a block of one or more addresses comprising the target address specified by the tag-error-checking request. The requester and the completer communicate via a memory interface having at least one data signal path to exchange read data or write data between the requester and the completer; and at least one tag signal path, provided in parallel with the at least one data signal path, to exchange address tags or allocation tags between the requester and the completer.

Memory interface having data signal path and tag signal path

An apparatus comprises a cache memory to store data as a plurality of cache lines each having a data size and an associated physical address in a memory, access circuitry to access the data stored in the cache memory, detection circuitry to detect, for at least a set of sub-units of the cache lines stored in the cache memory, whether a number of accesses by the access circuitry to a given sub-unit exceeds a predetermined threshold, in which each sub-unit has a data size that is smaller than the data size of a cache line, prediction circuitry to generate a prediction, for a given region of a plurality of regions of physical address space, of whether data stored in that region comprises streaming data in which each of one or more portions of the given cache line is predicted to be subject to a maximum of one read operation or multiple access data in which each of the one or more portions of the given cache line is predicted to be subject to more than one read operation, the prediction circuitry being configured to generate the prediction in response to a detection by the detection circuitry of whether the number of accesses to a sub-unit of a cache line having an associated physical address in the given region exceeds the predetermined threshold, and allocation circuitry to selectively allocate a next cache line to the cache memory in dependence upon the prediction applicable to the region of physical address space containing that next cache line.

Cache access detection and prediction

A processor fetches instructions from a plurality of threads. Each entry in a branch information storage (BIS) stores a virtual address ID for a branch, information about the branch, and thread ID information. The BIS is accessed using a virtual address of an instruction to be fetched for a thread to determine whether a hit exists, and if so, to obtain the branch information stored in the entry that gave rise to the hit. The virtual address is converted into a physical address, and an address translation regime is specified for each thread. When allocating an entry into the BIS, allocation circuitry determines, for a branch instruction for a current thread, whether the address translation regime is shared by plural threads. If so, the allocation circuitry identifies both the current thread and any other thread for which the address translation regime is shared.

Apparatus and method for sharing branch information storage entries between threads that share an address translation regime

Apparatus for data processing and a method of data processing are provided, according to which the processing circuitry of the apparatus can access a memory system and execute data processing instructions in one context of multiple contexts which it supports. When the processing circuitry executes a barrier instruction, the resulting access ordering constraint may be limited to being enforced for accesses which have been initiated by the processing circuitry when operating in an identified context, which may for example be the context in which the barrier instruction has been executed. This provides a separation between the operation of the processing circuitry in its multiple possible contexts and in particular avoids delays in the completion of the access ordering constraint, for example relating to accesses to high latency regions of memory, from affecting the timing sensitivities of other contexts.

Growing community of inventors

Cambridge, United Kingdom

Alexander Alfred Hornung