摘要:
Methods and apparatus to insert prefetch instructions based on garbage collector analysis and compiler analysis are disclosed. In an example method, one or more batches of samples associated with cache misses from a performance monitoring unit in a processor system are received. One or more samples from the one or more batches of samples based on delinquent information are selected. A performance impact indicator associated with the one or more samples is generated. Based on the performance indicator, at least one of a garbage collector analysis and a compiler analysis is initiated to identify one or more delinquent paths. Based on the at least one of the garbage collector analysis and the compiler analysis, one or more prefetch points to insert prefetch instructions are identified.
摘要:
Methods and apparatus to dynamically insert prefetch instructions are disclosed. In an example method, one or more samples associated with cache misses are identified from a performance monitoring unit in a processor system. Based on sample information associated with the one or more samples, delinquent information is generated. To dynamically insert one or more prefetch instructions, a prefetch point is identified based on the delinquent information.
摘要:
A method for managing a transaction includes determining that an optimistically immutable field in the transaction is written to. Invaliding a method in response to determining that the method in the transaction reads is the optimistically immutable field.Other embodiments are disclosed and claimed.
摘要:
Embodiments of a system and method for transactional memory (TM) with automatic object versioning are described. Embodiments described herein include a TM system and method that facilitates the execution of object-oriented application programs in a transactional environment, including automatically versioning objects to enhance efficiency. Embodiments of the TM automatically designate versions of objects using pointers, accurately identifying usable and unusable versions. Object versioning as described herein allows the garbage collector to easily and efficiently determine which objects may be moved, freeing memory space and reducing the number of objects traversed by a transaction before finding a useable version of an object. Other embodiments are described and claimed.
摘要:
Methods, systems, and mediums are described for scheduling data parallel tasks onto multiple thread execution units of processing system. Embodiments of a lock-free queue structure and methods of operation are described to implement a method for scheduling fine-grained data-parallel tasks for execution in a computing system. The work of one of a plurality of worker threads is wait-free with respect to the other worker threads. Each node of the queue holds a reference to a task that may be concurrently performed by multiple thread execution units, but each on a different subset of data. Various embodiments relate to software-based scheduling of data-parallel tasks on a multi-threaded computing platform that does not perform such scheduling in hardware. Other embodiments are also described and claimed.
摘要:
Memory reclamation with optimistic concurrency is described. In one example an allocated memory object is tentatively freed in a software transactional memory, the object having pointers into it from at least one transaction. A time when all transactions that are outstanding at the time an object is tentatively freed have ended is detected, and the object is actually freed based on the detection.
摘要:
Memory reclamation with optimistic concurrency is described. In one example an allocated memory object is tentatively freed in a software transactional memory, the object having pointers into it from at least one transaction. A time when all transactions that are outstanding at the time an object is tentatively freed have ended is detected, and the object is actually freed based on the detection.
摘要:
Provided is a method, system, and program for coordinating access to memory locations for hardware transactional memory transactions and software transactional memory transactions. A hardware transaction executing in hardware transactional memory initiates a request to access a memory location. A fault is returned to the hardware transaction request in response to an operation by one software transaction executing in a software transactional memory.
摘要:
Methods, systems, and mediums are described for scheduling data parallel tasks onto multiple thread execution units of processing system. Embodiments of a lock-free queue structure and methods of operation are described to implement a method for scheduling fine-grained data-parallel tasks for execution in a computing system. The work of one of a plurality of worker threads is wait-free with respect to the other worker threads. Each node of the queue holds a reference to a task that may be concurrently performed by multiple thread execution units, but each on a different subset of data. Various embodiments relate to software-based scheduling of data-parallel tasks on a multi-threaded computing platform that does not perform such scheduling in hardware. Other embodiments are also described and claimed.
摘要:
Embodiments of a system and method for transactional memory (TM) with automatic object versioning are described. Embodiments described herein include a TM system and method that facilitates the execution of object-oriented application programs in a transactional environment, including automatically versioning objects to enhance efficiency. Embodiments of the TM automatically designate versions of objects using pointers, accurately identifying usable and unusable versions. Object versioning as described herein allows the garbage collector to easily and efficiently determine which objects may be moved, freeing memory space and reducing the number of objects traversed by a transaction before finding a useable version of an object. Other embodiments are described and claimed.