公开(公告)号：US08176304B2

公开(公告)日：2012-05-08

申请号：US12256250

申请日：2008-10-22

申请人： Rahoul Puri ,  Arvind Srinivasan ,  Louise Y. Yeung ,  Marcelino M. Dignum ,  John E. Watkins

发明人： Rahoul Puri ,  Arvind Srinivasan ,  Louise Y. Yeung ,  Marcelino M. Dignum ,  John E. Watkins

IPC分类号： G06F9/00 ,  G06F15/177 ,  G06F1/24 ,  G06F3/00 ,  G06F13/00

CPC分类号： G06F13/385

摘要： An I/O device having function level reset functionality includes a host interface that may include a master reset unit, a plurality of client interfaces, each corresponding to one or more functions, and a plurality of hardware resources. Each hardware resource may be associated with a respective function. In response to receiving a reset request to reset a specific function, the master reset unit may provide to each client interface, a request signal corresponding to the reset request, and a signal identifying the specific function. Each client interface having an association with the specific function may initiate a reset operation of the associated hardware resources, and also provide a client reset done signal for the specific function to the master reset unit in response to completion of the reset operations of the hardware resources. The master reset unit provides a reset done signal for the specific function to the host interface.

摘要翻译： 具有功能级复位功能的I / O设备包括主机接口，其可以包括主复位单元，多个客户端接口，每个对应于一个或多个功能，以及多个硬件资源。 每个硬件资源可以与相应的功能相关联。 响应于接收到重置特定功能的复位请求，主复位单元可以向每个客户端接口提供对应于重置请求的请求信号，以及标识特定功能的信号。 具有与特定功能的关联的每个客户端接口可以启动相关联的硬件资源的复位操作，并且响应于硬件资源的复位操作的完成，向主复位单元提供用于特定功能的客户端重置完成信号 。 主复位单元为主机接口提供特定功能的复位完成信号。

公开(公告)号：US20100100717A1

公开(公告)日：2010-04-22

申请号：US12256250

申请日：2008-10-22

申请人： Rahoul Puri ,  Arvind Srinivasan ,  Louise Y. Yeung ,  Marcelino M. Dignum ,  John E. Watkins

发明人： Rahoul Puri ,  Arvind Srinivasan ,  Louise Y. Yeung ,  Marcelino M. Dignum ,  John E. Watkins

IPC分类号： G06F9/00

CPC分类号： G06F13/385

摘要： An I/O device having function level reset functionality includes a host interface that may include a master reset unit, a plurality of client interfaces, each corresponding to one or more functions, and a plurality of hardware resources. Each hardware resource may be associated with a respective function. In response to receiving a reset request to reset a specific function, the master reset unit may provide to each client interface, a request signal corresponding to the reset request, and a signal identifying the specific function. Each client interface having an association with the specific function may initiate a reset operation of the associated hardware resources, and also provide a client reset done signal for the specific function to the master reset unit in response to completion of the reset operations of the hardware resources. The master reset unit provides a reset done signal for the specific function to the host interface.

摘要翻译： 具有功能级复位功能的I / O设备包括主机接口，其可以包括主复位单元，多个客户端接口，每个对应于一个或多个功能，以及多个硬件资源。 每个硬件资源可以与相应的功能相关联。 响应于接收到重置特定功能的复位请求，主复位单元可以向每个客户端接口提供对应于重置请求的请求信号，以及标识特定功能的信号。 具有与特定功能的关联的每个客户端接口可以启动相关联的硬件资源的复位操作，并且响应于硬件资源的复位操作的完成，向主复位单元提供用于特定功能的客户端重置完成信号 。 主复位单元为主机接口提供特定功能的复位完成信号。

公开(公告)号：US20080228977A1

公开(公告)日：2008-09-18

申请号：US11685516

申请日：2007-03-13

申请人： Marcelino M. Dignum ,  Rahoul Puri

发明人： Marcelino M. Dignum ,  Rahoul Puri

IPC分类号： G06F13/14

CPC分类号： G06F13/1642 ,  G06F13/362 ,  H04L47/623 ,  H04L49/901 ,  H04L49/9078

摘要： A method and apparatus for dynamically arbitrating, in hardware, requests for a resource shared among multiple clients. Multiple data streams or service requests require access to a shared resource, such as memory, communication bandwidth, etc. A hardware arbiter monitors the streams' traffic levels and determines when one or more of their arbitration weights should be adjusted. When a queue used by one of the streams is filled to a threshold level, the hardware reacts by quickly and dynamically modifying that queue's arbitration weight. Therefore, as the queue is filled or emptied to different thresholds, the queue's arbitration weight rapidly changes to accommodate the corresponding client's temporal behavior. The arbiter may also consider other factors, such as the client's type of traffic, a desired quality of service, available credits, available descriptors, etc.

摘要翻译： 一种用于在硬件中动态地仲裁在多个客户端之间共享资源的请求的方法和装置。 多个数据流或服务请求需要访问诸如存储器，通信带宽等共享资源。硬件仲裁器监视流的流量级别，并确定何时应调整他们的仲裁权重中的一个或多个。 当一个流使用的队列被填充到阈值级别时，硬件通过快速且动态地修改该队列的仲裁权重来做出反应。 因此，当队列被填充或排空到不同的阈值时，队列的仲裁权重迅速变化以适应相应的客户端的时间行为。 仲裁者还可以考虑其他因素，例如客户的流量类型，所需的服务质量，可用的信用，可用的描述符等。

公开(公告)号：US07647444B2

公开(公告)日：2010-01-12

申请号：US11685516

申请日：2007-03-13

申请人： Marcelino M. Dignum ,  Rahoul Puri

发明人： Marcelino M. Dignum ,  Rahoul Puri

IPC分类号： G06F12/00

CPC分类号： G06F13/1642 ,  G06F13/362 ,  H04L47/623 ,  H04L49/901 ,  H04L49/9078

摘要： A method and apparatus for dynamically arbitrating, in hardware, requests for a resource shared among multiple clients. Multiple data streams or service requests require access to a shared resource, such as memory, communication bandwidth, etc. A hardware arbiter monitors the streams' traffic levels and determines when one or more of their arbitration weights should be adjusted. When a queue used by one of the streams is filled to a threshold level, the hardware reacts by quickly and dynamically modifying that queue's arbitration weight. Therefore, as the queue is filled or emptied to different thresholds, the queue's arbitration weight rapidly changes to accommodate the corresponding client's temporal behavior. The arbiter may also consider other factors, such as the client's type of traffic, a desired quality of service, available credits, available descriptors, etc.

摘要翻译： 一种用于在硬件中动态地仲裁在多个客户端之间共享资源的请求的方法和装置。 多个数据流或服务请求需要访问诸如存储器，通信带宽等共享资源。硬件仲裁器监视流的流量级别，并确定何时应调整他们的仲裁权重中的一个或多个。 当一个流使用的队列被填充到阈值级别时，硬件通过快速且动态地修改该队列的仲裁权重来做出反应。 因此，当队列被填充或排空到不同的阈值时，队列的仲裁权重迅速变化以适应相应的客户端的时间行为。 仲裁者还可以考虑其他因素，例如客户的流量类型，所需的服务质量，可用的信用，可用的描述符等。

公开(公告)号：US08606975B2

公开(公告)日：2013-12-10

申请号：US12784631

申请日：2010-05-21

申请人： Arvind Srinivasan ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Marcelino M. Dignum

IPC分类号： G06F13/28 ,  G06F13/24 ,  G06F15/167

CPC分类号： G06F13/102 ,  G06F11/0745 ,  G06F11/0763 ,  G06F13/24

摘要： Methods and apparatus are provided for managing interrupts within a virtualizable communication device. Through virtualization, one port of the device may be able to support multiple hosts (e.g., computers) and multiple functions operating on each host. Any number of interrupt resources may be allocated to the supported functions, and may include receive/transmit DMAs, receive/transmit mailboxes, errors, and so on. Resources may migrate from one function to another, such as when a function requests additional resources. Each function's set of allocated resources is isolated from other functions' resources so that their interrupts may be managed and reported in a non-blocking manner. If an interrupt cannot be immediately reported to a destination host/function, the interrupt may be delayed, retried, cancelled or otherwise handled in a way that avoids blocking interrupts to other hosts and functions.

摘要翻译： 提供了用于管理可虚拟化通信设备内的中断的方法和装置。 通过虚拟化，设备的一个端口可能能够支持在每个主机上操作的多个主机（例如，计算机）和多个功能。 可以将任何数量的中断资源分配给所支持的功能，并且可以包括接收/发送DMA，接收/发送邮箱，错误等。 资源可以从一个功能迁移到另一个功能，例如当一个函数请求额外的资源时。 每个功能的分配资源集与其他功能的资源隔离，以便它们的中断可以以非阻塞的方式进行管理和报告。 如果中断无法立即报告给目标主机/功能，则中断可能会以避免中断其他主机和功能的方式延迟，重试，取消或以其他方式处理。

公开(公告)号：US08369325B2

公开(公告)日：2013-02-05

申请号：US12494498

申请日：2009-06-30

申请人： Arvind Srinivasan ,  Michael F. Speer ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Michael F. Speer ,  Marcelino M. Dignum

IPC分类号： H04L12/28 ,  H04L12/56

CPC分类号： H04L47/10 ,  G06F9/45533 ,  H04L47/125 ,  H04L47/2441

摘要： Some embodiments of the present invention provide a system for packet classification and spreading in a virtualized system. The system can use information in a packet's header to determine a destination system-image in the virtualized system, and a packet-spreading policy for the destination system-image. The system can determine a key using the information in a packet's header. Alternatively, the system can hash the information in the packet's header to obtain an index value. Next, the system can use the key or the index value to perform a lookup in a table which associates keys or index values with system images and/or packet-spreading policies. Once the destination system-image and the packet-spreading policy are determined, the system can deliver the packet to a thread on the destination system-image according to the packet-spreading policy.

摘要翻译： 本发明的一些实施例提供了一种用于虚拟化系统中的分组分类和扩展的系统。 系统可以使用分组报头中的信息来确定虚拟化系统中的目标系统映像，以及目的系统映像的分组传播策略。 该系统可以使用分组报头中的信息来确定密钥。 或者，系统可以对分组报头中的信息进行散列以获得索引值。 接下来，系统可以使用密钥或索引值在将密钥或索引值与系统映像和/或分组传播策略相关联的表中执行查找。 一旦确定了目的地系统图像和分组扩展策略，系统可以根据分组传播策略将分组递送到目的地系统图像上的线程。

公开(公告)号：US08634415B2

公开(公告)日：2014-01-21

申请号：US13028650

申请日：2011-02-16

申请人： Arvind Srinivasan ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Marcelino M. Dignum

IPC分类号： H04L12/28

CPC分类号： H04L45/7457 ,  H04L12/56 ,  H04L45/586

摘要： A system and method for routing network traffic for a blade server chassis, involving receiving a packet by a network express manager (NEM); obtaining the L2 address, the L3 address, and the L4 address from the packet; obtaining a first mapping that associates the L2 address with a network function; obtaining a second mapping that associates the network function with a configuration; generating a key based on the network function, the L3 address, and the L4 address; obtaining a third mapping that associates the key with a policy; making a determination to transmit the packet based on the policy; based on the determination, selecting a direct memory access (DMA) engine for transmitting the packet based on the configuration; and based on the determination, transmitting the packet to a blade using the DMA engine and the network function.

摘要翻译： 一种用于路由刀片服务器机箱的网络流量的系统和方法，包括由网络快速管理器（NEM）接收分组; 从分组获取L2地址，L3地址和L4地址; 获得将L2地址与网络功能相关联的第一映射; 获得将网络功能与配置相关联的第二映射; 根据网络功能，L3地址和L4地址生成密钥; 获取将密钥与策略相关联的第三映射; 确定基于策略来传送分组; 基于所述确定，基于所述配置来选择用于发送所述分组的直接存储器访问（DMA）引擎; 并且基于确定，使用DMA引擎和网络功能将分组发送到刀片。

公开(公告)号：US08625448B2

公开(公告)日：2014-01-07

申请号：US13028513

申请日：2011-02-16

申请人： Arvind Srinivasan ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Marcelino M. Dignum

IPC分类号： H04L12/26

CPC分类号： H04L43/50 ,  H04L47/2441

摘要： A system and method for validating network traffic routing within a blade chassis, involving generating a first packet for sending to a first packet receiver by a first route; inserting a first session identifier into a payload of the first packet, where the first session identifier identifies a first session of the first packet receiver; sending the first packet to a packet classifier; sending a first copy packet to a first expect queue, where the first copy packet is a duplicate of the first packet; receiving the first packet by the packet classifier; classifying the first packet by the packet classifier to obtain a first classified packet; extracting the first session identifier from the first classified packet to obtain a first extracted session identifier; and determining whether the first extracted session identifier matches the first session identifier.

摘要翻译： 一种用于验证刀片机箱内的网络流量路由的系统和方法，包括生成用于通过第一路由发送到第一分组接收机的第一分组; 将第一会话标识符插入到所述第一分组的有效载荷中，其中所述第一会话标识符识别所述第一分组接收机的第一会话; 将第一分组发送到分组分类器; 将第一复制分组发送到第一期望队列，其中所述第一复制分组是所述第一分组的副本; 由分组分类器接收第一分组; 通过分组分类器对第一分组进行分类以获得第一分类分组; 从第一分类分组提取第一会话标识符以获得第一提取的会话标识符; 以及确定所述第一提取会话标识符是否与所述第一会话标识符匹配。

公开(公告)号：US20120207156A1

公开(公告)日：2012-08-16

申请号：US13028650

申请日：2011-02-16

申请人： Arvind Srinivasan ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Marcelino M. Dignum

IPC分类号： H04L12/56

CPC分类号： H04L45/7457 ,  H04L12/56 ,  H04L45/586

摘要： A system and method for routing network traffic for a blade server chassis, involving receiving a packet by a network express manager (NEM); obtaining the L2 address, the L3 address, and the L4 address from the packet; obtaining a first mapping that associates the L2 address with a network function; obtaining a second mapping that associates the network function with a configuration; generating a key based on the network function, the L3 address, and the L4 address; obtaining a third mapping that associates the key with a policy; making a determination to transmit the packet based on the policy; based on the determination, selecting a direct memory access (DMA) engine for transmitting the packet based on the configuration; and based on the determination, transmitting the packet to a blade using the DMA engine and the network function.

摘要翻译： 一种用于路由刀片服务器机箱的网络流量的系统和方法，包括由网络快速管理器（NEM）接收分组; 从分组获取L2地址，L3地址和L4地址; 获得将L2地址与网络功能相关联的第一映射; 获得将网络功能与配置相关联的第二映射; 根据网络功能，L3地址和L4地址生成密钥; 获取将密钥与策略相关联的第三映射; 确定基于策略来传送分组; 基于所述确定，基于所述配置来选择用于发送所述分组的直接存储器访问（DMA）引擎; 并且基于确定，使用DMA引擎和网络功能将分组发送到刀片。

公开(公告)号：US20100329253A1

公开(公告)日：2010-12-30

申请号：US12494498

申请日：2009-06-30

申请人： Arvind Srinivasan ,  Michael F. Speer ,  Marcelino M. Dignum

发明人： Arvind Srinivasan ,  Michael F. Speer ,  Marcelino M. Dignum

IPC分类号： H04L12/56

CPC分类号： H04L47/10 ,  G06F9/45533 ,  H04L47/125 ,  H04L47/2441

摘要： Some embodiments of the present invention provide a system for packet classification and spreading in a virtualized system. The system can use information in a packet's header to determine a destination system-image in the virtualized system, and a packet-spreading policy for the destination system-image. The system can determine a key using the information in a packet's header. Alternatively, the system can hash the information in the packet's header to obtain an index value. Next, the system can use the key or the index value to perform a lookup in a table which associates keys or index values with system images and/or packet-spreading policies. Once the destination system-image and the packet-spreading policy are determined, the system can deliver the packet to a thread on the destination system-image according to the packet-spreading policy.

摘要翻译： 本发明的一些实施例提供了一种用于虚拟化系统中的分组分类和扩展的系统。 系统可以使用分组报头中的信息来确定虚拟化系统中的目标系统映像，以及目的系统映像的分组传播策略。 该系统可以使用分组报头中的信息来确定密钥。 或者，系统可以对分组报头中的信息进行散列以获得索引值。 接下来，系统可以使用密钥或索引值在将密钥或索引值与系统映像和/或分组传播策略相关联的表中执行查找。 一旦确定了目的地系统图像和分组扩展策略，系统可以根据分组传播策略将分组递送到目的地系统图像上的线程。