会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 62. 发明授权
    • Control apparatus
    • 控制装置
    • US08589633B2
    • 2013-11-19
    • US12606242
    • 2009-10-27
    • Takeshi Ogasawara
    • Takeshi Ogasawara
    • G06F12/00
    • G06F9/526
    • A control apparatus, control method and computer readable article of manufacture for controlling data. The control apparatus includes a data storage unit; a plurality of entry storage units, and a plurality of registration units. The data storage unit stores data. Each of the entry storage units stores an entry for registering a pointer to data. If each of the registration units receives an instruction for registering data, then each registration unit (i) searches the entry storage units for an entry storage unit having an empty entry, (ii) registers a pointer to the data to be registered in the retrieved entry storage unit and (iii) stores the data to be registered and identification information of the retrieved entry storage unit in the data storage unit in such a manner that the data to be registered and the identification information is associated with each other.
    • 用于控制数据的控制装置,控制方法和计算机可读制造品。 控制装置包括数据存储单元; 多个入口存储单元和多个登记单元。 数据存储单元存储数据。 每个条目存储单元存储用于注册指向数据的条目。 如果每个注册单元接收到用于登记数据的指令,则每个注册单元(i)在入口存储单元中搜索具有空条目的条目存储单元,(ii)将注册的数据注册到所检索的数据中 入口存储单元,以及(iii)将要登记的数据和所检索的条目存储单元的识别信息以要登记的数据和识别信息相互关联的方式存储在数据存储单元中。
    • 65. 发明授权
    • Writing to file by multiple application threads in parallel
    • 通过多个应用程序线程并行写入文件
    • US08301857B2
    • 2012-10-30
    • US11241326
    • 2005-10-01
    • Takeshi Ogasawara
    • Takeshi Ogasawara
    • G06F11/30
    • G06F9/544G06F9/52
    • Multiple application threads write to the same file in parallel. The file is written to on a file cluster-by-file cluster basis. For each file cluster of a number of file clusters of the file, a block of a memory cluster is allocated to an application thread, where the memory cluster corresponds to the file cluster, and the block is written to by the application thread until the block is finished, until all the blocks of the memory cluster are finished. A block may be finished where the application thread has completely written to the block, or the application thread has no further data to write to the block. Thereafter, the memory cluster is registered within a queue, from which it is read by a file writer thread, which writes the memory cluster to the file cluster. Each application thread that is to write to the file performs this process.
    • 多个应用程序线程并行写入同一个文件。 文件以逐个文件集群为基础写入文件。 对于文件的多个文件簇的每个文件簇,将内存集群的块分配给应用程序线程,其中内存集群对应于文件集群,并且该块由应用程序线程写入,直到块 完成后,直到内存集群的所有块完成。 可以在应用程序线程完全写入块的情况下完成块,或者应用程序线程没有进一步的数据写入块。 此后,将内存集群注册到队列中,由文件写入程序线程从该队列中读取内存集,该文件写入程序将内存集群写入文件集群。 要写入文件的每个应用程序线程执行此过程。
    • 66. 发明授权
    • Managing timeout in a multithreaded system by instantiating a timer object having scheduled expiration time and set of timeout handling information
    • 通过实例化具有预定到期时间的定时器对象和一组超时处理信息来管理多线程系统中的超时
    • US08239871B2
    • 2012-08-07
    • US12144780
    • 2008-06-24
    • Takeshi Ogasawara
    • Takeshi Ogasawara
    • G06F9/46G06F11/00
    • G06F9/4825G06F9/542G06F2209/543
    • A method of managing a multithreaded computer system comprises instantiating, in response to each transaction initiated by a first thread of a plurality of threads, a timer object including a scheduled expiration time and a set of timeout handling information for the transaction in storage local to the first thread; registering, in response to each passing of a fixed time interval, each timer object in the storage local to the first thread for which the scheduled expiration time is earlier than the fixed time interval added to a current time in a timer processing component by adding a pointer referencing the timer object to a data structure managed by the timer processing component; and managing each timer object corresponding to a transaction initiated by the first thread that is not registered in the timer processing component in the storage local to the first thread. The timer processing component regularly processes each timer object referenced by the data structure for which the scheduled expiration time value is not earlier than the current time in accordance with the set of timeout handling information of the timer object.
    • 一种管理多线程计算机系统的方法包括:响应于由多个线程的第一线程发起的每个事务来实例化定时器对象,其包括调度的到期时间和一组超时处理信息,用于存储在本地 第一线 响应于固定时间间隔的每次通过,将定时处理组件中添加到当前时间的固定时间间隔早于第一线程的本地存储器中的每个定时器对象注册到计时器处理组件中的当前时间的固定时间间隔, 将定时器对象引用到由定时器处理组件管理的数据结构; 以及管理与由第一线程发起的事务相对应的每个定时器对象,该事务未被注册到第一线程的本地存储器中的定时器处理组件中。 定时器处理组件根据定时器对象的超时处理信息的集合,定期处理由调度的到期时间值不早于当前时间的数据结构引用的每个定时器对象。
    • 67. 发明授权
    • Server device operating in response to received request
    • 服务器设备响应于接收到的请求而工作
    • US08078674B2
    • 2011-12-13
    • US12114877
    • 2008-05-05
    • Takeshi Ogasawara
    • Takeshi Ogasawara
    • G06F15/16
    • G06F9/5027G06F2209/5018
    • A server device, methods and a computer program product operating in response to a request received from a client. A request storage device in system memory space with a request storage region stores a request received from a client in association with identification information. An identification information storage region contains a request that is waiting to be processed. An identification information storage unit retrieves at least one request waiting to be processed from the request storage region if the number of identification information pieces stored in the identification information storage region has declined to a reference number or below, and then stores retrieved request identification information in the identification information storage region Multiple request processors operate in parallel, each acquiring identification information from the identification information storage region, a request identified by the acquired identification information from the request storage region, and performing processing in response to the acquired request.
    • 响应于从客户端接收的请求而操作的服务器设备,方法和计算机程序产品。 具有请求存储区域的系统存储空间中的请求存储装置与识别信息相关联地存储从客户端接收到的请求。 识别信息存储区域包含正在等待处理的请求。 如果存储在识别信息存储区域中的识别信息的数量已经下降到参考号码或更低的值,则识别信息存储单元从请求存储区域检索等待处理的至少一个请求,然后将检索到的请求标识信息存储在 识别信息存储区域多请求处理器并行操作,每个从识别信息存储区域获取识别信息,从所述请求存储区域获取由所获取的识别信息识别的请求,并响应于所获取的请求执行处理。
    • 68. 发明申请
    • MEMORY CONTROL APPARATUS, PROGRAM, AND METHOD
    • 记忆控制装置,程序和方法
    • US20100122041A1
    • 2010-05-13
    • US12616236
    • 2009-11-11
    • Takuya NakaikeRei OdairaTakeshi Ogasawara
    • Takuya NakaikeRei OdairaTakeshi Ogasawara
    • G06F12/00
    • G06F9/466
    • A memory control apparatus which controls access to a shared memory for each transaction. The apparatus includes a management unit that stores versions of data stored in the shared memory, a log storage unit that stores an update entry including a version of data subjected to an update operation in response to execution of an update operation on the shared memory in processing each transaction, and a control unit that writes a result of processing corresponding to execution of a relevant update operation to the shared memory when a request to commit a transaction has been given, and a relevant update entry version matches a corresponding version stored in the management unit, or re-executes the update operation and writes a result of re-execution to the shared memory when the update entry version does not match the corresponding version in the management unit.
    • 一种用于每次交易控制对共享存储器的访问的存储器控​​制装置。 该装置包括管理单元,存储存储在共享存储器中的数据的版本;日志存储单元,用于响应于在处理中对共享存储器执行更新操作而存储包含更新操作的数据的版本的更新条目 每个交易,以及控制单元,当已经给出了提交交易的请求时,将与执行相关更新操作相对应的处理结果写入共享存储器,并且相关更新条目版本与存储在管理中的对应版本相匹配 单元,或者当更新入口版本与管理单元中的相应版本不匹配时,重新执行更新操作并将重新执行的结果写入共享存储器。
    • 69. 发明授权
    • Exception handling compiler apparatus, program, recording medium, and compiling method
    • 异常处理编译器,程序,记录介质和编译方法
    • US07526761B2
    • 2009-04-28
    • US10801879
    • 2004-03-16
    • Takeshi Ogasawara
    • Takeshi Ogasawara
    • G06F9/45
    • G06F9/4843G06F2209/481
    • Compiler apparatus, methods and systems for optimizing exception handling in a program. An example apparatus comprises an exception handler detection section for detecting, from exception handlers that catch exceptions thrown in the program, a multiple-catching exception handler that catches a plurality of different exceptions and rethrow the caught exceptions; an exception selection section for selecting a set of exceptions that are to be shifted to common processing through rethrow of the exception by the multiple-catching exception handler from among the plurality of exceptions caught by the detected multiple-catching exception handler; and an exception handler throw section for throwing a clone exception handler that catches the set of exceptions selected by the exception selection section instead of the multiple-catching exception handler and shifting it to common processing.
    • 用于优化程序中的异常处理的编译器装置,方法和系统。 示例性装置包括异常处理程序检测部分,用于从捕获程序中抛出的异常的异常处理程序检测捕获多个不同异常并重新捕获捕获的异常的多重捕获异常处理程序; 异常选择部分,用于通过由所检测的多重捕获异常处理程序捕获的多个异常中的多重捕获异常处理程序重新抛出异常来选择要转换为公共处理的一组异常; 以及一个异常处理程序抛出部分,用于抛出一个克隆异常处理程序,该异常处理程序捕获异常选择部分选择的异常集,而不是多次捕获的异常处理程序并将其转换为公共处理。