Mechanisms and methods for using data access patterns
    81.
    发明授权
    Mechanisms and methods for using data access patterns 有权
    使用数据访问模式的机制和方法

    公开(公告)号:US07395407B2

    公开(公告)日:2008-07-01

    申请号:US11250288

    申请日:2005-10-14

    IPC分类号: G06F12/00

    摘要: The present invention comprises a data access pattern interface that allows software to specify one or more data access patterns such as stream access patterns, pointer-chasing patterns and producer-consumer patterns. Software detects a data access pattern for a memory region and passes the data access pattern information to hardware via proper data access pattern instructions defined in the data access pattern interface. Hardware maintains the data access pattern information properly when the data access pattern instructions are executed. Hardware can then use the data access pattern information to dynamically detect data access patterns for a memory region throughout the program execution, and voluntarily invoke appropriate memory and cache operations such as pre-fetch, pre-send, acquire-ownership and release-ownership. Further, hardware can provide runtime monitoring information for memory accesses to the memory region, wherein the runtime monitoring information indicates whether the software-provided data access pattern information is accurate.

    摘要翻译: 本发明包括数据访问模式接口,其允许软件指定一个或多个数据访问模式,例如流访问模式,指针追踪模式和生产者 - 消费者模式。 软件检测存储器区域的数据访问模式,并通过数据访问模式界面中定义的适当的数据访问模式指令将数据访问模式信息传递给硬件。 当执行数据访问模式指令时,硬件正确地维护数据访问模式信息。 然后,硬件可以使用数据访问模式信息在整个程序执行期间动态地检测存储器区域的数据访问模式,并且主动地调用适当的存储器和缓存操作,例如预取,预发送,获取所有权和释放所有权。 此外,硬件可以提供用于存储器访问存储器区域的运行时监视信息,其中运行时监视信息指示软件提供的数据访问模式信息是否准确。

    Memory access monitoring
    82.
    发明申请
    Memory access monitoring 失效
    内存访问监控

    公开(公告)号:US20080034355A1

    公开(公告)日:2008-02-07

    申请号:US11478743

    申请日:2006-06-30

    IPC分类号: G06F9/45

    摘要: A computer-implemented method for memory access monitoring, implemented by a managed runtime environment computer system including a controller that monitors application behavior and determines actions to be taken to change a behavior of an application, and a runtime, dynamic compiler that analyzes the application and generates code sequences to access a memory access monitoring (MAM) mechanism, includes determining monitor information of a plurality of fields of a memory block to drive an optimization of the application.

    摘要翻译: 一种用于存储器访问监视的计算机实现的方法,由管理的运行时环境计算机系统实现,所述计算机系统包括监视应用程序行为并确定要改变应用程序行为的控制器的控制器,以及分析应用程序的运行时,动态编译器,以及 生成用于访问存储器访问监视(MAM)机制的代码序列,包括确定存储器块的多个字段的监视器信息以驱动应用的优化。

    Scope-based cache coherence
    84.
    发明申请
    Scope-based cache coherence 失效
    基于范围的缓存一致性

    公开(公告)号:US20060095684A1

    公开(公告)日:2006-05-04

    申请号:US10981370

    申请日:2004-11-04

    申请人: Xiaowei Shen

    发明人: Xiaowei Shen

    IPC分类号: G06F12/00

    CPC分类号: G06F12/0826 G06F12/0831

    摘要: With scope-based cache coherence, a cache can maintain scope information for a memory address. The scope information specifies caches in which data of the address is potentially cached, but not necessarily caches in which data of the address is actually cached. Appropriate scope information can be used as snoop filters to reduce unnecessary coherence messages and snoop operations in SMP systems. If a cache maintains scope information of an address, it can potentially avoid sending cache requests to caches outside the scope in case of a cache miss on the address. Scope information can be adjusted dynamically via a scope calibration operation to reflect changing data access patterns. A calibration prediction mechanism can be employed to predict when a scope calibration needs to be invoked.

    摘要翻译: 使用基于范围的缓存一致性,缓存可以维护存储器地址的范围信息。 范围信息指定缓存,其中地址的数据可能被高速缓存,但不一定是高速缓存,其中地址的数据实际上被缓存。 适当的范围信息可以用作窥探过滤器,以减少SMP系统中不必要的一致性消息和窥探操作。 如果缓存维护地址的范围信息,则在地址上存在高速缓存未命中的情况下,可能会避免向高速缓存发送缓存请求。 范围信息可以通过示波器校准操作动态调整,以反映不断变化的数据访问模式。 校准预测机制可用于预测何时需要调用范围校准。

    Location-aware cache-to-cache transfers
    85.
    发明申请
    Location-aware cache-to-cache transfers 失效
    位置感知缓存到缓存传输

    公开(公告)号:US20050240735A1

    公开(公告)日:2005-10-27

    申请号:US10833197

    申请日:2004-04-27

    IPC分类号: G06F12/00 G06F12/08

    摘要: In shared-memory multiprocessor systems, cache interventions from different sourcing caches can result in different cache intervention costs. With location-aware cache coherence, when a cache receives a data request, the cache can determine whether sourcing the data from the cache will result in less cache intervention cost than sourcing the data from another cache. The decision can be made based on appropriate information maintained in the cache or collected from snoop responses from other caches. If the requested data is found in more than one cache, the cache that has or likely has the lowest cache intervention cost is generally responsible for supplying the data. The intervention cost can be measured by performance metrics that include, but are not limited to, communication latency, bandwidth consumption, load balance, and power consumption.

    摘要翻译: 在共享内存多处理器系统中,来自不同采购缓存的缓存干预可能导致不同的缓存干预成本。 利用位置感知高速缓存一致性,当缓存接收到数据请求时,高速缓存可以确定从高速缓存中提取数据是否会导致比从另一高速缓存提供数据更少的高速缓存干预成本。 该决定可以基于保存在缓存中的适当信息或从其他缓存的窥探响应收集。 如果在多个缓存中找到所请求的数据,则具有或可能具有最低缓存干预成本的高速缓存通常负责提供数据。 干预成本可以通过包括但不限于通信延迟,带宽消耗,负载平衡和功耗的性能指标来衡量。

    Adaptive cache coherence protocols
    86.
    发明授权
    Adaptive cache coherence protocols 失效
    自适应高速缓存一致性协议

    公开(公告)号:US06526481B1

    公开(公告)日:2003-02-25

    申请号:US09561168

    申请日:2000-04-27

    IPC分类号: C06F1200

    摘要: A methodology for designing a distributed shared-memory system, which can incorporate adaptation or selection of cache protocols during operation, guarantees semantically correct processing of memory instructions by the multiple processors. A set of rules includes a first subset of “mandatory” rules and a second subset of “voluntary” rules such that correct operation of the memory system is provided by application of all of the mandatory rules and selective application of the voluntary rules. A policy for enabling voluntary rules specifies a particular coherent cache protocol. The policy can include various types of adaptation and selection of different operating modes for different addresses and at different caches. A particular coherent cache protocol can make use of a limited capacity directory in which some but not necessarily all caches that hold a particular address are identified in the directory. In another coherent cache protocol, various caches hold an address in different modes which, for example, affect communication between a cache and a shared memory in processing particular memory instructions.

    摘要翻译: 用于设计分布式共享存储器系统的方法,其可以在操作期间结合适应或选择高速缓存协议,保证了多处理器对存储器指令的语义上正确的处理。 一组规则包括“强制性”规则的第一子集和“自愿”规则的第二子集,从而通过应用所有强制性规则并选择性地应用自愿规则来提供存储系统的正确操作。 启用自愿规则的策略规定了特定的一致高速缓存协议。 该策略可以包括各种类型的适配和针对不同地址和不同高速缓存的不同操作模式的选择。 特定的一致高速缓存协议可以使用有限容量的目录,其中在目录中识别出一些但不一定所有保存特定地址的高速缓存。 在另一个相干高速缓存协议中,各种高速缓存以不同的模式保存地址,这些地址例如在处理特定的存储器指令时影响高速缓存和共享存储器之间的通信。