公开(公告)号：US20160291984A1

公开(公告)日：2016-10-06

申请号：US15175742

申请日：2016-06-07

Applicant: HUAWEI TECHNOLOGIES CO.,LTD.

Inventor： Gang Lu ,  Jianfeng Zhan ,  Yunwei Gao ,  Chongkang Tan ,  Dongliang Xue

IPC: G06F9/44 ,  G06F13/24

CPC classification number: G06F9/4406 ,  G06F9/4403 ,  G06F9/4405 ,  G06F9/4408 ,  G06F9/45533 ,  G06F9/5083 ,  G06F13/24

Abstract: A resource processing method, an operating system, and a device are provided. The method is applied to a multi-core operating system, where the multi-core operating system includes a management operating system and multiple load operating systems that run on a host machine and includes a physical resource pool. According to the method, after selecting, in processor cores allocated to a first load operating system, a startup processor core that starts up the first load operating system, the management operating system instructs the startup processor core to read a mapping relationship that is from a virtual memory address to a physical memory address and that is required for executing a startup mirror of the first load operating system. Then, the management operating system instructs the startup processor core to execute the startup mirror pre-constructed for the first load operating system to start up the first load operating system.

Abstract translation: 提供资源处理方法，操作系统和设备。 该方法应用于多核操作系统，其中多核操作系统包括管理操作系统和在主机上运行并包括物理资源池的多个加载操作系统。 根据该方法，在分配给第一负载操作系统的处理器核心中选择启动第一负载操作系统的启动处理器核心之后，管理操作系统指示启动处理器核心读取来自 虚拟内存地址到物理内存地址，这是执行第一个加载操作系统的启动镜像所必需的。 然后，管理操作系统指示启动处理器内核执行为第一加载操作系统预先构建的启动镜，以启动第一加载操作系统。

公开(公告)号：US20240314204A1

公开(公告)日：2024-09-19

申请号：US18676795

申请日：2024-05-29

Applicant: HUAWEI TECHNOLOGIES CO., LTD.

Inventor： Gang Lu ,  Kun Tan ,  Ying Li ,  Junying Li ,  Chihyung WANG ,  Binzhang Fu

IPC: H04L67/1097

CPC classification number: H04L67/1097

Abstract: This application discloses a transaction processing method and a device that may be used in the field of network technologies. The method includes: A transmitting device determines a first order flag for a generated first transaction based on service logic of an application, and obtains order auxiliary information of the first transaction. The order flag indicates a dependency relationship (for example, forward dependency and backward blocking) between different transactions. This application does not depend on an operation type of a transaction. A transaction order between transactions may be set, and the receiving device ensures an execution order of each transaction based on an order flag, provided that a transaction relationship is fully expressed. The transaction order is flexibly defined, to meet the requirements of different scenarios and improve processing performance.

公开(公告)号：US10180843B2

公开(公告)日：2019-01-15

申请号：US15175742

申请日：2016-06-07

Applicant: HUAWEI TECHNOLOGIES CO.,LTD.

Inventor： Gang Lu ,  Jianfeng Zhan ,  Yunwei Gao ,  Chongkang Tan ,  Dongliang Xue

IPC: G06F9/4401 ,  G06F9/50 ,  G06F9/455 ,  G06F13/24 ,  G06F9/46

Abstract: A resource processing method, an operating system, and a device are provided. The method is applied to a multi-core operating system, where the multi-core operating system includes a management operating system and multiple load operating systems that run on a host machine and includes a physical resource pool. According to the method, after selecting, in processor cores allocated to a first load operating system, a startup processor core that starts up the first load operating system, the management operating system instructs the startup processor core to read a mapping relationship that is from a virtual memory address to a physical memory address and that is required for executing a startup mirror of the first load operating system. Then, the management operating system instructs the startup processor core to execute the startup mirror pre-constructed for the first load operating system to start up the first load operating system.

公开(公告)号：US09652150B2

公开(公告)日：2017-05-16

申请号：US14739497

申请日：2015-06-15

Applicant: Huawei Technologies Co., Ltd.

Inventor： Chongkang Tan ,  Gang Lu ,  Jianfeng Zhan ,  Lixin Zhang

IPC: G06F12/10 ,  G06F3/06 ,  G06F9/50 ,  G06F12/08

CPC classification number: G06F3/0604 ,  G06F3/0631 ,  G06F3/0683 ,  G06F9/50 ,  G06F9/5016 ,  G06F12/08

Abstract: A global memory sharing method includes counting, by the sub-operating system, a page replacement rate in a task scheduling period of a predetermined quantity of times, and a memory residence time ratio in a page replacement period; calculating, by the sub-operating system, a memory pressure index according to the page replacement rate and the memory residence time ratio; and if the memory pressure index is greater than a memory pressure threshold, sending, by the sub-operating system, an application to a global memory management service module. According to the forgoing method, in an architecture of multiple operating systems, each sub-operating system can complete much adaptive work, which reduces complexity of the global memory management service module and improves system performance. Meanwhile, a utilization rate of global memory may be improved, and the architecture of the multiple operating systems may have better extensibility.

公开(公告)号：US20150277772A1

公开(公告)日：2015-10-01

申请号：US14739497

申请日：2015-06-15

Applicant: Huawei Technologies Co., Ltd.

Inventor： Chongkang Tan ,  Gang Lu ,  Jianfeng Zhan ,  Lixin Zhang

IPC: G06F3/06

CPC classification number: G06F3/0604 ,  G06F3/0631 ,  G06F3/0683 ,  G06F9/50 ,  G06F9/5016 ,  G06F12/08

Abstract: A global memory sharing method includes counting, by the sub-operating system, a page replacement rate in a task scheduling period of a predetermined quantity of times, and a memory residence time ratio in a page replacement period; calculating, by the sub-operating system, a memory pressure index according to the page replacement rate and the memory residence time ratio; and if the memory pressure index is greater than a memory pressure threshold, sending, by the sub-operating system, an application to a global memory management service module. According to the forgoing method, in an architecture of multiple operating systems, each sub-operating system can complete much adaptive work, which reduces complexity of the global memory management service module and improves system performance. Meanwhile, a utilization rate of global memory may be improved, and the architecture of the multiple operating systems may have better extensibility.

Abstract translation: 一种全局存储器共享方法，包括：在所述子操作系统中，在预定次数的任务调度期间中对所述页更换率进行计数，以及在页替换期间计数存储器停留时间比; 由子操作系统根据页更换率和存储器停留时间比计算存储器压力指数; 并且如果存储器压力指数大于存储器压力阈值，则由子操作系统将应用发送到全局存储器管理服务模块。 根据前述方法，在多操作系统的架构中，每个子操作系统可以完成大量的自适应工作，降低了全局内存管理服务模块的复杂性，提高了系统性能。 同时，可以提高全局存储器的利用率，并且多个操作系统的架构可以具有更好的可扩展性。