公开(公告)号：US06327621B1

公开(公告)日：2001-12-04

申请号：US09139684

申请日：1998-08-25

申请人： Chin Lee ,  Bruce H. Ratcliff ,  Stephen R. Valley

发明人： Chin Lee ,  Bruce H. Ratcliff ,  Stephen R. Valley

IPC分类号： G06F1516

CPC分类号： H04L12/185 ,  H04L12/1836 ,  H04L12/66 ,  H04L45/04 ,  H04L45/16

摘要： A method in a computing network environment for host partition computers to share a multicast communication adapter connected to a local area network. Each host partition sends information identifying the host application as well as a multicast addressing protocol to the shared multicast communication adapter. The communication adapter registers the information and directs inbound multicast packets from the LAN to the appropriate host partition application. In a preferred embodiment, the communication adapter registers the information according to the IP multicast address. In a preferred embodiment, a counter is employed in the communication adapter to maintain a count of the number of host applications that are registered. The invention further provides a method for un-registering a host partition application.

摘要翻译： 用于主机分区计算机的计算网络环境中共享连接到局域网的多播通信适配器的方法。 每个主机分区向共享的多播通信适配器发送标识主机应用的信息以及组播寻址协议。 通信适配器注册信息，并将来自LAN的入站组播数据包引导到相应的主机分区应用程序。 在优选实施例中，通信适配器根据IP多播地址登记该信息。 在优选实施例中，在通信适配器中采用计数器来维护登记的主机应用的数量的计数。 本发明还提供了一种用于取消注册主机分区应用程序的方法。

公开(公告)号：US6003088A

公开(公告)日：1999-12-14

申请号：US924007

申请日：1997-08-29

申请人： John S. Houston ,  Bruce H. Ratcliff ,  Anthony R. Sager ,  Donald W. Schmidt ,  Arthur J. Stagg ,  Stephen R. Valley

发明人： John S. Houston ,  Bruce H. Ratcliff ,  Anthony R. Sager ,  Donald W. Schmidt ,  Arthur J. Stagg ,  Stephen R. Valley

IPC分类号： H04L29/06 ,  H04L29/08 ,  G06F13/00

CPC分类号： H04L29/06 ,  H04L69/14 ,  H04L69/32

摘要： In a computing network system environment having a gateway device that is electronically connected from one side to a plurality of initiating hosts and on another side to at least one local area network, in turn connecting a plurality of hosts are connected to the gateway device, a method and apparatus for eliminating any need for building a separate and special protocol data unit element for each header. The computer network environment uses a Multi-path channel communication protocol as well as protocol data units to point to various portions of data. An interface layer is provided between a plurality of protocol stacks and the multi-path channel protocol. The interface layer has a timer which will wait for data from the protocol stack. A list of all buffers are assembled as received comprising of one entry for each data buffer. This buffer list is then transferred to the multi-path channel protocol layer upon expiration of the timer, which is in turn sent to any channel attached processor as one block. A deblocker interface is also provided on all channel attached processors so that any length fields provided in the protocol headers can be used to determine offset of next data element in said block and thus eliminating need for a special header at next data element used by the communication protocol.

摘要翻译： 在具有网络设备的计算网络系统环境中，网关设备从一侧电连接到多个发起主机，另一方面连接到至少一个局域网，依次连接多个主机连接到网关设备， 方法和装置，用于消除为每个报头构建独立和特殊的协议数据单元元素的任何需要。 计算机网络环境使用多路径信道通信协议以及协议数据单元来指向数据的各个部分。 在多个协议栈和多路径信道协议之间提供接口层。 接口层有一个定时器，它将等待来自协议栈的数据。 所有缓冲器的列表被组合成接收，包括每个数据缓冲器的一个条目。 然后，在定时器到期时，该缓冲器列表被传送到多路径信道协议层，定时器依次发送到任何通道连接的处理器作为一个块。 在所有通道连接的处理器上还提供解块器接口，使得在协议报头中提供的任何长度字段可用于确定所述块中的下一个数据元素的偏移量，从而消除了对通信使用的下一个数据元素的特殊报头的需要 协议。

公开(公告)号：US08869118B2

公开(公告)日：2014-10-21

申请号：US13150600

申请日：2011-06-01

申请人： Raymond M. Higgs ,  George P. Kuch ,  Bruce H. Ratcliff

发明人： Raymond M. Higgs ,  George P. Kuch ,  Bruce H. Ratcliff

IPC分类号： G06F9/44 ,  G06F11/34

CPC分类号： G06F11/3466 ,  G06F11/3419 ,  G06F2201/865 ,  G06F2201/88

摘要： System aware performance counters including a processor for performing a method that includes executing a predefined code segment of an application, the executing on a processor. The executing includes executing an instrumented thread included in the predefined code segment. The method includes performing a first action associated with executing the instrumented thread in the predefined code segment, and registering the instrumented thread. A performance counter associated with the predefined code segment is started. Also, the execution of the predefined code segment is paused and the performance counter is paused. The method further includes performing a second action associated with pausing the executing of the predefined code segment. The executing of the predefined code segment is resumed and the performance counter is resumed responsive to resuming the executing of the predefined code segment. Also, a third action associated with resuming executing the predefined code segment is performed.

摘要翻译： 系统感知性能计数器，包括用于执行包括执行应用程序的预定代码段的方法的处理器，在处理器上执行。 执行包括执行包含在预定义代码段中的被检测线程。 该方法包括执行与在预定代码段中执行被检测线程相关联的第一动作，以及登记被检测的线程。 开始与预定义代码段相关联的性能计数器。 此外，预定义代码段的执行被暂停，并且性能计数器被暂停。 该方法还包括执行与暂停预定义代码段的执行相关联的第二动作。 恢复预定义代码段的执行，并且响应于恢复执行预定义代码段而恢复性能计数器。 此外，执行与恢复执行预定义代码段相关联的第三动作。

公开(公告)号：US08402190B2

公开(公告)日：2013-03-19

申请号：US12326468

申请日：2008-12-02

申请人： Maurice Isrel, Jr. ,  Bruce H. Ratcliff ,  Jerry W. Stevens ,  Edward Zebrowski, Jr.

发明人： Maurice Isrel, Jr. ,  Bruce H. Ratcliff ,  Jerry W. Stevens ,  Edward Zebrowski, Jr.

IPC分类号： G06F13/24 ,  G06F13/32

CPC分类号： G06F13/24 ,  H04L47/56 ,  H04L49/90

摘要： A method and system are disclosed for network adaptor optimization and interrupt reduction. The method may generate an immediate I/O interrupt notification to a host device driver in response to receiving data from a network. The method may also update an inbound buffer list based on the received data in the inbound buffer queue. Furthermore, the method may set the buffer state from an empty state to a primed state to indicate that the received data is available for processing. The method may also build an outbound buffer list based on outgoing data and add the outgoing data to an outbound buffer queue. Furthermore, the method may set a buffer state from an empty state to a primed state to indicate that the outgoing data is prepared for transmitting and signal a network adaptor with a notification signal.

摘要翻译： 公开了一种用于网络适配器优化和中断减少的方法和系统。 响应于从网络接收数据，该方法可以产生对主机设备驱动器的即时I / O中断通知。 该方法还可以基于入站缓冲器队列中接收到的数据来更新入站缓冲器列表。 此外，该方法可以将缓冲器状态从空状态设置为初始化状态，以指示所接收的数据可用于处理。 该方法还可以基于输出数据构建出站缓冲器列表，并将输出数据添加到出站缓冲器队列。 此外，该方法可以将缓冲状态从空状态设置为初始化状态，以指示输出数据准备好用于发送信号并通知具有通知信号的网络适配器。

公开(公告)号：US20110106990A1

公开(公告)日：2011-05-05

申请号：US12609110

申请日：2009-10-30

申请人： Raymond Higgs ,  George P. Kuch ,  Bruce H. Ratcliff ,  Gustav E. Sittmann, III ,  Jerry W. Stevens

发明人： Raymond Higgs ,  George P. Kuch ,  Bruce H. Ratcliff ,  Gustav E. Sittmann, III ,  Jerry W. Stevens

IPC分类号： G06F13/38

CPC分类号： G06F13/385

摘要： Computer program products and methods for efficient handling of queued-direct input/output (QDIO) requests and completions at an adapter in communication with an I/O device are provided. A method includes accessing a queue with one or more storage block address lists (SBALs), where each SBAL includes a plurality of storage block address list entries (SBALEs) and is associated with an SLSB. The method further includes reading an SBAL count in one of the SBALEs, where the SBAL count indicates a number of the SBALs forming an I/O request to the I/O device. In response to determining that the SBAL count is greater than one, a number of the SBALs from the queue and associated SLSBs equivalent to the SBAL count are prefetched without waiting for a notification of completion of each of the SBALs forming the I/O request, and states of the associated SLSBs transition from adapter-owned to program-owned.

摘要翻译： 提供了用于在与I / O设备通信的适配器处有效处理排队直接输入/输出（QDIO）请求和完成的计算机程序产品和方法。 一种方法包括使用一个或多个存储块地址列表（SBAL）访问队列，其中每个SBAL包括多个存储块地址列表条目（SBALE）并且与SLSB相关联。 该方法还包括读取SBALE之一的SBAL计数，其中SBAL计数指示形成I / O设备的I / O请求的SBAL的数目。 响应于确定SBAL计数大于1，来自队列的SBAL和与SBAL计数相关联的相关联的SLSB的数量被预取，而不等待形成I / O请求的每个SBAL的完成通知， 以及相关SLSB从适配器所有权转为程序拥有的状态。

公开(公告)号：US06772221B1

公开(公告)日：2004-08-03

申请号：US09505909

申请日：2000-02-17

申请人： Bruce H. Ratcliff ,  Christopher B. Larsson

发明人： Bruce H. Ratcliff ,  Christopher B. Larsson

IPC分类号： G06F1516

CPC分类号： H04L43/00 ,  H04L29/06 ,  H04L41/0803 ,  H04L43/0811 ,  H04L45/02 ,  H04L69/18

摘要： In a computing network environment having a gateway device electronically connected to a plurality of hosts potentially connected to one or more routers and local area networks, a method and apparatus is provided for dynamically configuring, deconfiguring and monitoring all connected hosts. Special IPX commands are provided to add or remove information about hosts in a routing table. The command provides information about the connected hosts' IPX sessions, router information and IPX packets and their final destinations, and connection routes. In another embodiment of the present invention the IPX command is also used to receive information and pass information about new IPX stations in the network environment. By providing a counter and incrementing it each time an IPX packet is received from the network, the most used entries are identified. A second special IPX host command is also provided for deconfiguring and removing entries from the routing table in a manner similar to the one described above.

摘要翻译： 在具有电子连接到潜在地连接到一个或多个路由器和局域网的多个主机的网关设备的计算网络环境中，提供了用于动态配置，解除配置和监视所有连接的主机的方法和装置。 提供特殊的IPX命令来添加或删除有关路由表中主机的信息。 该命令提供有关连接的主机IPX会话，路由器信息和IPX数据包及其最终目的地以及连接路由的信息。 在本发明的另一个实施例中，IPX命令还用于在网络环境中接收关于新IPX站的信息并传递关于新IPX站的信息。 通过提供计数器并在每次从网络接收到IPX分组时递增计数器，则识别出最多使用的条目。还提供了第二特殊IPX主机命令，用于以类似于该方式的方式从路由表中解除配置和删除条目 如上所述。

公开(公告)号：US06600743B1

公开(公告)日：2003-07-29

申请号：US09140180

申请日：1998-08-25

申请人： Chin Lee ,  Bruce H. Ratcliff ,  Stephen R. Valley

发明人： Chin Lee ,  Bruce H. Ratcliff ,  Stephen R. Valley

IPC分类号： H04L1228

CPC分类号： H04L12/1836 ,  H04L12/185 ,  H04L45/16

摘要： A method of multicasting messages in a computing network environment having at least one host and a plurality of clients connected and in processing communication with one another via a gateway device. Host(s) and client(s) requesting registration will be registered and their information will be entered in a routing table, one entry dedicated to each host or client. A connectivity platform controls all communication between the host and the client(s) by receiving all information packets. The Routing table is then searched and the packets are distributed accordingly to their owner once a matching entry is found. If a matching entry is not found the packet will be dropped.

摘要翻译： 一种在具有至少一个主机和多个客户端的计算网络环境中组播消息的方法，所述客户端经由网关设备彼此连接和处理通信。 请求注册的主机和客户端将被注册，并将他们的信息输入到路由表中，一个专用于每个主机或客户端的条目。 连接平台通过接收所有信息包来控制主机和客户端之间的所有通信。 然后搜索路由表，并且一旦找到匹配的条目，则相应地将分组分发给其所有者。 如果没有找到匹配的条目，则数据包将被丢弃。

公开(公告)号：US06345324B1

公开(公告)日：2002-02-05

申请号：US09252552

申请日：1999-02-19

申请人： Michael E. Baskey ,  Frank W. Brice, Jr. ,  Steven G. Glassen ,  Eugene P. Hefferon ,  Bruce H. Ratcliff ,  Arthur J. Stagg ,  Stephen R. Valley ,  Anthony R. Sager

发明人： Michael E. Baskey ,  Frank W. Brice, Jr. ,  Steven G. Glassen ,  Eugene P. Hefferon ,  Bruce H. Ratcliff ,  Arthur J. Stagg ,  Stephen R. Valley ,  Anthony R. Sager

IPC分类号： G06F1300

CPC分类号： H04L25/05

摘要： An apparatus for transferring data in a network computing environment having a controlling program and a main storage in processing communication with an interface element including one or more adapters. A queuing mechanism is first established in the main storage for asynchronous transmittal of data between the program and said interface element. At least one queue-set is dedicated to input data and another queue-set to output data in the queuing mechanism. Queuing priorities both for said input and output queue-sets are then determined according to importance of data to be processed or transferred or other internal dependencies and a queue-description record is established by passing a QDIO queue command. From this the priority and number of the output queue are provided by order in which the order specified in the queue descriptor record relative to any other output queues described in that record. Data is then transferred and received according to the queuing priority by interrogating multiple queues in the mechanism simultaneously and determining appropriate location to which data has to be stored to or transferred from.

摘要翻译： 一种用于在具有控制程序和主存储器的网络计算环境中传送数据的装置，用于处理与包括一个或多个适配器的接口元件的通信。 首先在主存储器中建立排队机制，以在程序和所述接口元件之间异步传送数据。 至少一个队列集专用于输入数据，另一个队列集用于在排队机制中输出数据。 然后根据要处理或传输的数据的重要性或其他内部依赖关系确定所述输入和输出队列集的排队优先级，并且通过传递QDIO队列命令建立队列描述记录。 从此，输出队列的优先级和编号由队列描述符记录中指定的顺序相对于该记录中描述的任何其他输出队列的顺序提供。 然后，通过同时询问机构中的多个队列，根据排队优先级传送和接收数据，并确定数据必须存储到哪个或从其传送的适当位置。

公开(公告)号：US06339803B1

公开(公告)日：2002-01-15

申请号：US09252727

申请日：1999-02-19

申请人： Steven G. Glassen ,  Eugene P. Hefferon ,  Bruce H. Ratcliff ,  Arthur J. Stagg ,  Stephen R. Valley

发明人： Steven G. Glassen ,  Eugene P. Hefferon ,  Bruce H. Ratcliff ,  Arthur J. Stagg ,  Stephen R. Valley

IPC分类号： G06F1300

CPC分类号： H04L49/901 ,  H04L49/90 ,  H04L67/42

摘要： A computer program product in a network environment having a main storage in processing communication with an interface element having adapters for storing data in a queuing mechanism in the main storage. The program product comprises a data storage device including a computer usable medium having computer readable program means for ensuring proper and timely processing of data and computer readable code means for providing a plurality of queue sets; for dedicating one or more queues in each set as input or output queues; for providing the ability to said queues to share access to a plurality of devices in said network across a plurality of communication stacks; for mapping various resources to said queues to facilitate resource allocation and dynamic configuration by providing initialization of a plurality of configuration parameters; and for providing dynamic expanding and contracting of the number of queues in each set as dictated by traffic patterns and feedback indicators.

摘要翻译： 一种网络环境中的计算机程序产品，具有处理与具有用于在主存储器中的排队机构中存储数据的适配器的接口元件进行通信的主存储器。 该程序产品包括一个数据存储装置，它包括具有计算机可读程序装置的计算机可用介质，该计算机可用介质用于确保数据的适当和及时处理以及用于提供多个队列的计算机可读代码装置 用于将每个集合中的一个或多个队列用作输入或输出队列; 用于提供所述队列在多个通信栈中共享对所述网络中的多个设备的访问的能力; 用于通过提供多个配置参数的初始化来将各种资源映射到所述队列以便于资源分配和动态配置; 并且用于根据流量模式和反馈指示符来提供每个集合中的队列的数量的动态扩展和约束。

公开(公告)号：US06185218B2

公开(公告)日：2001-02-06

申请号：US08920533

申请日：1997-08-29

申请人： Bruce H. Ratcliff ,  Stephen R. Valley

发明人： Bruce H. Ratcliff ,  Stephen R. Valley

IPC分类号： H04L1266

CPC分类号： H04L29/12018 ,  H04L12/2856 ,  H04L12/2872 ,  H04L29/06 ,  H04L29/12009 ,  H04L61/10 ,  H04L67/14 ,  H04L67/28 ,  H04L67/2842 ,  H04L67/34 ,  H04L69/16 ,  H04L69/163 ,  H04L69/329

摘要： A method and apparatus for establishing communication between a first initiating host and a second communicating host in a computing environment having a gateway device. The gateway device is electronically connected from one side to a plurality of initiating hosts and on another side to at least one local area network, further connecting a plurality of receiving hosts to said gateway device. During an initialization step, all connected hosts and the local area network(s) is identified and their addresses and information obtained and stored in a port-sharing table in a memory location accessible to the gateway device. Upon receipt of a special command for establishing communication between a first initiating host and a second receiving host, all information required for establishing of a successful communication between the two hosts is then obtained from the port-sharing table. In an alternate embodiment of the present invention the computing network environment uses Multi-path channel communication protocol.

摘要翻译： 一种用于在具有网关设备的计算环境中建立第一发起主机和第二通信主机之间的通信的方法和装置。 网关设备从一侧电连接到多个发起主机，另一方面电连接到至少一个局域网，还将多个接收主机连接到所述网关设备。 在初始化步骤期间，识别所有连接的主机和局域网，并将其地址和信息获取并存储在网关设备可访问的存储器位置中的端口共享表中。 在接收到用于在第一发起主机和第二接收主机之间建立通信的特殊命令时，从端口共享表获得在两个主机之间建立成功通信所需的所有信息。 在本发明的替代实施例中，计算网络环境使用多路径信道通信协议。