首页> 外国专利> Support for multiple outstanding requests to multiple targets in a pipelined memory system

Support for multiple outstanding requests to multiple targets in a pipelined memory system

机译:支持对流水线内存系统中多个目标的多个未完成请求

摘要

One embodiment of the present invention provides an apparatus (152) that supports multiple outstanding load and/or store requests from an execution engine (106, 108) to multiple sources of data in a computer system. This apparatus includes a load store unit (102, 104) coupled to the execution engine (106, 108), a first data source (114) and a second data source (144, 146, 148, 150). This load store unit (102) includes a load address buffer (216), which contains addresses for multiple outstanding load requests. The load store unit (102) also includes a controller (250) that co-ordinates data flow between the load address buffer (216), a register file (110), the first data source and the second data source (144, 146, 148, 150) so that multiple load requests can simultaneously be outstanding for both the first data source and the second data source. These load requests return in-order for each of the multiple sources of data in the computer system, except for load requests directed to a data cache (114) which can return out-of-order. Load requests may return out-of-order with respect to load requests from other data sources. According to one aspect of the present invention, the load store unit (102) additionally includes a store address buffer (220), that contains addresses for multiple outstanding store requests, and a store data buffer (230) that contains data for the multiple outstanding store requests. The controller (250) is further configured to co-ordinate data flow between the first data source, the second data source, the store address buffer (220) and the store data buffer (230), so that multiple store requests can simultaneously be outstanding for both the first data source and the second data source. IMAGE
机译:本发明的一个实施例提供了一种设备(152),其支持多个未完成的负载和/或存储从执行引擎(106、108)到计算机系统中的多个数据源的请求。该设备包括耦合到执行引擎(106、108)的负载存储单元(102、104),第一数据源(114)和第二数据源(144、146、148、150)。该负载存储单元(102)包括负载地址缓冲器(216),其包含用于多个未完成的负载请求的地址。负载存储单元(102)还包括控制器(250),该控制器协调负载地址缓冲区(216),寄存器文件(110),第一数据源和第二数据源(144、146, 148、150),因此对于第一数据源和第二数据源,可以同时同时处理多个加载请求。这些加载请求针对计算机系统中的多个数据源中的每一个按顺序返回,除了指向可以无序返回的数据高速缓存(114)的加载请求之外。加载请求可能相对于来自其他数据源的加载请求无序返回。根据本发明的一方面,加载存储单元(102)还包括:存储地址缓冲器(220),其包含用于多个未完成的存储请求的地址;以及存储数据缓冲器(230),其包含用于多个未完成的数据的数据。存储请求。控制器(250)还被配置为协调第一数据源,第二数据源,存储地址缓冲器(220)和存储数据缓冲器(230)之间的数据流,从而可以同时处理多个存储请求。对于第一数据源和第二数据源。 <图像>

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号