公开(公告)号：US20080270708A1

公开(公告)日：2008-10-30

申请号：US11741858

申请日：2007-04-30

申请人： Craig Warner ,  Bryan Hornung ,  Chris Michael Brueggen ,  Ryan L. Akkerman ,  Michael K. Dugan ,  Gary Gostin ,  Harvey Ray ,  Dan Robinson ,  Christopher Greer

发明人： Craig Warner ,  Bryan Hornung ,  Chris Michael Brueggen ,  Ryan L. Akkerman ,  Michael K. Dugan ,  Gary Gostin ,  Harvey Ray ,  Dan Robinson ,  Christopher Greer

IPC分类号： G06F12/08

CPC分类号： G06F12/082 ,  G06F12/0831 ,  G06F2212/1048

摘要： A system and method are disclosed for achieving cache coherency in a multiprocessor computer system having a plurality of sockets with processing devices and memory controllers and a plurality of memory blocks. In at least some embodiments, the system includes a plurality of node controllers capable of being respectively coupled to the respective sockets of the multiprocessor computer, a plurality of caching devices respectively coupled to the respective node controllers, and a fabric coupling the respective node controllers, by which cache line request signals can be communicated between the respective node controllers. Cache coherency is achieved notwithstanding the cache line request signals communicated between the respective node controllers due at least in part to communications between the node controllers and the respective caching devices to which the node controllers are coupled. In at least some embodiments, the caching devices track remote cache line ownership for processor and/or input/output hub caches.

摘要翻译： 公开了一种用于在具有多个具有处理设备和存储器控制器以及多个存储器块的套接字的多处理器计算机系统中实现高速缓存一致性的系统和方法。 在至少一些实施例中，系统包括能够分别耦合到多处理器计算机的相应插槽的多个节点控制器，分别耦合到相应节点控制器的多个高速缓存设备，以及耦合各个节点控制器的结构， 由此可以在相应的节点控制器之间传送高速缓存行请求信号。 尽管至少部分地由于节点控制器与节点控制器所耦合的相应的高速缓存设备之间的通信而在各个节点控制器之间传送高速缓存行请求信号，仍然实现了高速缓存一致性。 在至少一些实施例中，缓存设备跟踪用于处理器和/或输入/输出集线器高速缓存的远程高速缓存行所有权。

公开(公告)号：US07818508B2

公开(公告)日：2010-10-19

申请号：US11741453

申请日：2007-04-27

申请人： Bryan Hornung ,  Erin A. Handgen ,  Gary Gostin ,  Craig Warner

发明人： Bryan Hornung ,  Erin A. Handgen ,  Gary Gostin ,  Craig Warner

IPC分类号： G06F13/00

CPC分类号： G06F12/0607 ,  G06F12/0817 ,  G06F13/1657

摘要： A computer system, related components such as a processor agent, and related method are disclosed. In at least one embodiment, the computer system includes a first core, at least one memory device including a first memory segment, and a first memory controller coupled to the first memory segment. Further, the computer system includes a fabric and at least one processor agent coupled at least indirectly to the first core and the first memory segment, and also coupled to the fabric. A first memory request of the first core in relation to a first memory location within the first memory segment proceeds to the first memory controller by way of the at least one processor agent and the fabric.

摘要翻译： 公开了一种计算机系统，诸如处理器代理的相关组件和相关方法。 在至少一个实施例中，计算机系统包括第一核心，包括第一存储器段的至少一个存储器设备和耦合到第一存储器段的第一存储器控制器。 此外，计算机系统包括织物和至少一个处理器代理，至少间接耦合到第一核心和第一存储器段，并且还耦合到织物。 相对于第一存储器段内的第一存储器位置的第一存储器的第一存储器请求通过至少一个处理器代理和结构进行到第一存储器控制器。

公开(公告)号：US20080270743A1

公开(公告)日：2008-10-30

申请号：US11741453

申请日：2007-04-27

申请人： Bryan Hornung ,  Erin A. Handgen ,  Gary Gostin ,  Craig Warner

发明人： Bryan Hornung ,  Erin A. Handgen ,  Gary Gostin ,  Craig Warner

IPC分类号： G06F9/26

CPC分类号： G06F12/0607 ,  G06F12/0817 ,  G06F13/1657

摘要： A computer system, related components such as a processor agent, and related method are disclosed. In at least one embodiment, the computer system includes a first core, at least one memory device including a first memory segment, and a first memory controller coupled to the first memory segment. Further, the computer system includes a fabric and at least one processor agent coupled at least indirectly to the first core and the first memory segment, and also coupled to the fabric. A first memory request of the first core in relation to a first memory location within the first memory segment proceeds to the first memory controller by way of the at least one processor agent and the fabric.

摘要翻译： 公开了一种计算机系统，诸如处理器代理的相关组件和相关方法。 在至少一个实施例中，计算机系统包括第一核心，包括第一存储器段的至少一个存储器设备和耦合到第一存储器段的第一存储器控制器。 此外，计算机系统包括织物和至少一个处理器代理，至少间接耦合到第一核心和第一存储器段，并且还耦合到织物。 相对于第一存储器段内的第一存储器位置的第一存储器的第一存储器请求通过至少一个处理器代理和结构进行到第一存储器控制器。

公开(公告)号：US20080147888A1

公开(公告)日：2008-06-19

申请号：US11588903

申请日：2006-10-26

申请人： Chris M. Giles ,  Bryan Hornung ,  Michael J. Phelps ,  Joseph F. Orth

发明人： Chris M. Giles ,  Bryan Hornung ,  Michael J. Phelps ,  Joseph F. Orth

IPC分类号： G06F3/00

CPC分类号： G06F12/06

摘要： One exemplary method of assigning addresses in two or more address spaces with address fields of different lengths comprises defining address types, assigning a value to first bits at the high ends of the address fields to identify a first said address type, assigning second bits at the low ends of the address fields to identify addresses of the first said address type, and inserting different numbers of additional bits between the first bits and the second bits in the two or more address spaces.

摘要翻译： 在具有不同长度的地址字段的两个或多个地址空间中分配地址的一种示例性方法包括定义地址类型，将值分配给地址字段的高端的第一位以识别第一所述地址类型，在第 地址字段的低端以标识第一所述地址类型的地址，以及在两个或多个地址空间中的第一位和第二位之间插入不同数量的附加位。

公开(公告)号：US07117313B2

公开(公告)日：2006-10-03

申请号：US10803289

申请日：2004-03-18

申请人： Bryan Hornung ,  Bryan Marietta ,  Robert K. King

发明人： Bryan Hornung ,  Bryan Marietta ,  Robert K. King

IPC分类号： G06F12/00

CPC分类号： G06F13/1657

摘要： A SCI controller manages responses and requests between SCI interconnection rings and memory access controllers. The SCI controller includes a request activation queue that stores information about the requests until the SCI rings have the resources to handle the requests. The controller also has a response activation queue that stores information about the responses until the memory access controller is accessible. The queues do not store the request and response packets, but rather store information that is used to construct the request and response packets. The SCI controller also has a contents addressable memory or CAM that checks for an address match between the current requests and responses and previous requests and responses. A table stores more specific information about the previous requests.

摘要翻译： SCI控制器管理SCI互连环和存储器访问控制器之间的响应和请求。 SCI控制器包括一个请求激活队列，用于存储关于请求的信息，直到SCI环有资源来处理请求。 控制器还具有响应激活队列，其存储关于响应的信息，直到存储器访问控制器可访问。 队列不存储请求和响应数据包，而是存储用于构造请求和响应数据包的信息。 SCI控制器还具有内容可寻址存储器或CAM，用于检查当前请求与响应之间的地址匹配以及以前的请求和响应。 表格存储有关先前请求的更多具体信息。

公开(公告)号：US06684381B1

公开(公告)日：2004-01-27

申请号：US09675725

申请日：2000-09-29

申请人： Lionel Bening ,  Bryan Hornung ,  Robert Pflederer

发明人： Lionel Bening ,  Bryan Hornung ,  Robert Pflederer

IPC分类号： G06F1750

CPC分类号： G06F17/5045

摘要： A method of providing hardware description language-embedded regular expression support for module iteration and interconnection. Regular expressions such as those used in the Perl programming language are used in a preprocessing process to generate instances and interconnections in a hardware description to automate the generation of repetitive code for a Hardware Description Language (HDL). This is accomplished by generating HDL code with embedded regular expressions, analyzing the code to identify the regular expressions and checking to see that the code complies with the HDL grammar rules. A data structure is generated for each module or submodule and these data structures are then elaborated to expand them into the instances and interconnections. A text generator traverses the elaborated data structures and generates HDL compliant text.

摘要翻译： 一种为模块迭代和互连提供硬件描述语言嵌入式正则表达式支持的方法。 用于Perl编程语言的正则表达式用于预处理过程，以在硬件描述中生成实例和互连，以自动生成硬件描述语言（HDL）的重复代码。 这是通过使用嵌入式正则表达式生成HDL代码来实现的，分析代码以识别正则表达式，并检查代码是否符合HDL语法规则。 为每个模块或子模块生成数据结构，然后详细阐述这些数据结构，将其扩展到实例和互连中。 文本生成器遍历详细的数据结构并生成符合HDL的文本。

公开(公告)号：US06665830B2

公开(公告)日：2003-12-16

申请号：US09773124

申请日：2001-01-31

申请人： Bryan Hornung ,  Gregory S Palmer ,  Paul F. Vogel

发明人： Bryan Hornung ,  Gregory S Palmer ,  Paul F. Vogel

IPC分类号： G11C2900

CPC分类号： G06F11/1004

摘要： A system for building checksums efficiently builds a checksum of various data values that are stored in different memory units of a computer system. During the checksum build process, data stores to the memory locations storing the various data values are enabled, thereby enabling the checksum to be built without significantly impacting the performance of the computer system.

公开(公告)号：US06473845B1

公开(公告)日：2002-10-29

申请号：US09675021

申请日：2000-09-28

申请人： Bryan Hornung ,  Michael L. Ziegler ,  Michael K. Traynor ,  Gregory S. Palmer

发明人： Bryan Hornung ,  Michael L. Ziegler ,  Michael K. Traynor ,  Gregory S. Palmer

IPC分类号： G06F1300

CPC分类号： G06F12/0292 ,  G06F13/1663

摘要： In general, a system and method is provided for dynamically reallocating computer memory. A mapper receives requests to access data. The requests include bus addresses, and the mapper maps the bus addresses to memory unit addresses based on a plurality of mappings maintained by the mapper. The memory unit addresses identify a plurality of memory locations including a destination memory location and a source memory location. Data requested by the requests received by the mapper is accessed based on the memory unit addresses mapped from the bus addresses included in the requests. When desired, a data value from the source memory location is dynamically moved to the destination memory location, and the mappings are updated such that a bus address mapped to a memory unit address identifying the source memory location is instead mapped to a memory unit address identifying the destination memory location.

摘要翻译： 通常，提供用于动态地重新分配计算机存储器的系统和方法。 映射器接收访问数据的请求。 请求包括总线地址，并且映射器基于由映射器维护的多个映射将总线地址映射到存储器单元地址。 存储器单元地址识别包括目的地存储器位置和源存储器位置的多个存储器位置。 基于从请求中包括的总线地址映射的存储器单元地址来访问由映射器接收到的请求所请求的数据。 当需要时，来自源存储器位置的数据值被动态地移动到目的地存储器位置，并且映射被更新，使得映射到标识源存储器位置的存储器单元地址的总线地址被映射到存储器单元地址识别 目的地记忆位置。

公开(公告)号：US06317857B1

公开(公告)日：2001-11-13

申请号：US09759853

申请日：2001-01-12

申请人： Bryan Hornung ,  Gregory S Palmer ,  Paul F. Vogel

发明人： Bryan Hornung ,  Gregory S Palmer ,  Paul F. Vogel

IPC分类号： G11C2900

CPC分类号： G06F11/1088 ,  G06F11/1004

摘要： A system for storing and retrieving data utilizes a plurality of memory units having memory locations for storing data values. A checksum of a plurality of the data values in a particular checksum set is maintained in one of the memory locations. One of the plurality of data values can be recovered by combining each of the remaining plurality of data values with the checksum. After retrieving one of the plurality of data values during the data recovery process, steps are taken to ensure that any further attempts to access the location of the retrieved data value do not cause an update to the checksum. Therefore, the locations storing the data values of the checksum set may be accessed (e.g., read from or written to) during the data recovery process without causing errors to the data recovery process.

摘要翻译： 用于存储和检索数据的系统利用具有用于存储数据值的存储器位置的多个存储器单元。 在特定校验和集合中的多个数据值的校验和被保持在存储器位置之一中。 可以通过将剩余的多个数据值中的每一个与校验和组合来恢复多个数据值中的一个。 在数据恢复过程中检索多个数据值中的一个数据值之后，采取步骤以确保进一步尝试访问检索到的数据值的位置不会对校验和进行更新。 因此，可以在数据恢复处理期间存储存储校验和集合的数据值的位置（例如，从其读取或写入），而不会对数据恢复处理造成错误。

公开(公告)号：US6052761A

公开(公告)日：2000-04-18

申请号：US792310

申请日：1997-01-31

申请人： Bryan Hornung ,  Bryan Marietta

发明人： Bryan Hornung ,  Bryan Marietta

IPC分类号： G06F15/16 ,  G06F12/08 ,  G06F15/177 ,  G06F12/00

CPC分类号： G06F12/0824

摘要： The system and method for performing an update operation on the sharing list in a multi-node SCI-based system involves issuing an update command from a first node in the list to the remaining nodes in the sharing list, however the first node maintains its position at a head position in the sharing list. The first node issues an increment command to each successive node in the list until the end of the list is reached. Each node receives the increment command and updates its cache and then sends a reply back to the first node, listing the state of the node and a pointer to the next node. The method and system also prevents a node from rolling out of the sharing list before that node receives the update command. Each node has a phase value that is toggled as the node is incremented. Thus, a rollout request will be nullified unless the phase value of the rollout request matches the phase value of the node. The system and method also interlocks the rollout requests to inhibit a node that is being rolled out from being incremented.

摘要翻译： 用于在基于多节点SCI的系统中对共享列表执行更新操作的系统和方法包括从列表中的第一节点向共享列表中的其余节点发出更新命令，然而第一节点保持其位置 在共享列表中的头位。 第一个节点向列表中的每个连续节点发出增量命令，直到达到列表的末尾。 每个节点接收增量命令并更新其缓存，然后将回复发送回第一个节点，列出节点的状态和指向下一个节点的指针。 该方法和系统还防止节点在该节点接收到更新命令之前滚出共享列表。 每个节点具有在节点递增时切换的相位值。 因此，除非推出请求的相位值与节点的相位值相匹配，否则推出请求将被取消。 系统和方法还互锁了推出请求，以阻止正在推出的节点递增。