专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

US5442758A Apparatus and method for achieving reduced overhead mutual exclusion and maintaining coherency in a multiprocessor system utilizing execution history and thread monitoring 失效
标题翻译：利用执行历史和线程监视在多处理器系统中实现减少的开销互斥和维持一致性的装置和方法
公开(公告)号：US5442758A
公开(公告)日：1995-08-15
申请号：US094629
申请日：1993-07-19
申请人： John D. Slingwine , Paul E. McKenney
发明人： John D. Slingwine , Paul E. McKenney
IPC分类号： G06F9/46 , G06F13/00
CPC分类号： G06F9/526 , G06F9/52 , Y10S707/956 , Y10S707/99938
摘要： A substantially zero overhead mutual-exclusion apparatus and method (90, 120) is provided that allows concurrent reading and updating data while maintaining data coherency. That is, a data reading process executes the same sequence of instructions that would be executed if the data were never updated. Rather than depending exclusively on overhead-imposing locks, this mutual-exclusion mechanism tracks an execution history (138) of a thread (16, 112) to determine safe times for processing a current generation (108, 130, 131) of data updates while a next generation (110, 132, 133) of data updates is concurrently being saved. A thread is any locus of control, such as a processor. A summary of thread activity (106, 122) tracks which threads have passed through a quiescent state after the current generation of updates was started. When the last thread related to the current generation passes through a quiescent state, the summary of thread activity signals a callback processor (104, 124) that it is safe to end the current generation of updates. The callback processor then processes and erases all updates in the current generation. The next generation of updates then becomes the current generation of updates. The callback processor restarts the summary of thread activity and initiates a new next generation of updates. All data-updating threads pass through a quiescent state between the time they attempt to update data and the time the data are actually updated.
摘要翻译：提供了基本为零的开销互斥装置和方法（90,120），其允许在保持数据一致性的同时读取和更新数据。也就是说，如果数据从未被更新，则数据读取处理执行将执行的相同的指令序列。这种相互排除机制不是完全依赖于高效的锁，而是跟踪线程（16,112）的执行历史（138）以确定用于处理数据更新的当前生成（108,130,131）的安全时间，数据更新的下一代（110,132,133）同时被保存。线程是任何控制的轨迹，例如处理器。线程活动（106,122）的摘要跟踪当前生成更新后线程已经过静态状态。当与当前一代有关的最后一个线程通过静态时，线程活动的摘要将回叫处理器（104,124）发信号通知可以安全地结束当前的更新生成。然后，回调处理器处理和删除当前一代中的所有更新。下一代更新将成为当前的一代更新。回调处理器重新启动线程活动的摘要，并启动新的下一代更新。所有数据更新线程在尝试更新数据的时间和数据实际更新的时间之间经过静态。

2. 发明授权

US08621276B2 File system resiliency management 有权
标题翻译：文件系统弹性管理
公开(公告)号：US08621276B2
公开(公告)日：2013-12-31
申请号：US12970985
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig A. Barkhouse , Lane Haury , Kiran Kumar G. Bangalore , Thiago Sigrist
发明人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig A. Barkhouse , Lane Haury , Kiran Kumar G. Bangalore , Thiago Sigrist
IPC分类号： G06F11/00
CPC分类号： G06F11/0793 , G06F11/0727 , G06F11/0772 , G06F11/0775
摘要： Perceived corruptions encountered on file system volumes, and which cannot be initially remedied online, are processed to verify whether they are true, existing volume data structure corruptions or, alternatively, false positives. Upon the verification of one or more of a volume's corruptions, error scanning is performed to check for, and attempt to remedy online, all the existing corruptions on the volume. Subsequent to error scanning processing, if one or more verified corruptions continue to exist on a file system volume, at file system boot up time spot corruption correction is performed to attempt to remedy the existing, verified corruptions on the volume. Spot corruption correction is performed to attempt to correct verified data structure corruptions on a volume of the file system while the volume is maintained offline for the time necessary to attempt to correct its prior identified corruptions.
摘要翻译：在文件系统卷上遇到的感知到的损坏，最初不能在线补救，将被处理，以验证它们是真实的，现有的卷数据结构损坏还是错误的。在验证一个或多个卷的损坏后，执行错误扫描以检查并尝试在线补救卷上的所有现有损坏。在错误扫描处理之后，如果在文件系统卷上继续存在一个或多个已验证的损坏，则在文件系统启动时执行斑点损坏修正，以尝试补救现有的已验证的卷上的损坏。执行点腐败纠正，以在卷保持脱机的同时尝试更正其先前识别的损坏所需的时间，尝试纠正文件系统卷上的已验证数据结构损坏。

3. 发明授权

US08667323B2 Proactive error scan and isolated error correction 有权
标题翻译：主动错误扫描和隔离纠错
公开(公告)号：US08667323B2
公开(公告)日：2014-03-04
申请号：US12970954
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig A. Barkhouse
发明人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig A. Barkhouse
IPC分类号： G06F11/00
CPC分类号： G06F11/1435 , G06F11/0727 , G06F11/0793 , G06F2201/84
摘要： Processing for file system volume error detection and processing for resultant error correction are separated to support system availability and user satisfaction. File system volumes for storing data structures are proactively scanned while the volumes remain online to search for errors or corruptions thereon. Found errors are scheduled to be corrected, i.e., spot corrected, dependent on the severity of the identified errors, error correction scheduling and/or at the determination of a file system administrator and/or user, to assist in maintaining minimal user and file system impact. When spot correction is initialized, one file system volume at a time is taken offline for correction. Spot correction verifies prior logged corruptions for the offline volume, and if independently verified, attempts to correct the prior noted corruptions. Volumes are retained offline only for the time necessary to verify and attempt to correct prior noted volume corruptions.
摘要翻译：用于文件系统卷错检测和处理的结果错误纠正的处理被分离以支持系统可用性和用户满意度。主动扫描用于存储数据结构的文件系统卷，同时卷保持联机，以搜索其上的错误或损坏。找到的错误被安排被校正，即，根据所识别的错误的严重性，错误校正调度和/或文件系统管理员和/或用户的确定来校正，以协助维护最小的用户和文件系统影响。当点校正被初始化时，一次一个文件系统卷被脱机以进行校正。斑点校正验证离线卷的先前记录的损坏情况，如果经过独立验证，则尝试更正先前发现的损坏。卷仅在验证并尝试纠正先前引发的卷损坏所需的时间内保留脱机。

4. 发明申请

US20120159256A1 File System Resiliency Management 有权
标题翻译：文件系统弹性管理
公开(公告)号：US20120159256A1
公开(公告)日：2012-06-21
申请号：US12970985
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig Barkhouse , Lane Haury , Kiran Kumar G. Bangalore , Thiago M. Sigrist
发明人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Daniel Chan , Craig Barkhouse , Lane Haury , Kiran Kumar G. Bangalore , Thiago M. Sigrist
IPC分类号： G06F11/07
CPC分类号： G06F11/0793 , G06F11/0727 , G06F11/0772 , G06F11/0775
摘要： Perceived corruptions encountered on file system volumes, and which cannot be initially remedied online, are processed to verify whether they are true, existing volume data structure corruptions or, alternatively, false positives. Upon the verification of one or more of a volume's corruptions, error scanning is performed to check for, and attempt to remedy online, all the existing corruptions on the volume. Subsequent to error scanning processing, if one or more verified corruptions continue to exist on a file system volume, at file system boot up time spot corruption correction is performed to attempt to remedy the existing, verified corruptions on the volume. Spot corruption correction is performed to attempt to correct verified data structure corruptions on a volume of the file system while the volume is maintained offline for the time necessary to attempt to correct its prior identified corruptions.
摘要翻译：在文件系统卷上遇到的感知到的损坏，最初不能在线补救，将被处理，以验证它们是真实的，现有的卷数据结构损坏还是错误的。在验证一个或多个卷的损坏后，执行错误扫描以检查并尝试在线补救卷上的所有现有损坏。在错误扫描处理之后，如果在文件系统卷上继续存在一个或多个已验证的损坏，则在文件系统启动时执行斑点损坏修正，以尝试补救现有的已验证的卷上的损坏。执行点腐败纠正，以在卷保持脱机的同时尝试更正其先前识别的损坏所需的时间，尝试纠正文件系统卷上的已验证数据结构损坏。

5. 发明申请

US20120159243A1 Proactive Error Scan And Isolated Error Correction 有权
标题翻译：主动误差扫描和隔离误差校正
公开(公告)号：US20120159243A1
公开(公告)日：2012-06-21
申请号：US12970954
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Chrlstiansen , John D. Slingwine , Daniel Chen , Craig A. Barkhouse
发明人： Sarosh C. Havewala , Neal R. Chrlstiansen , John D. Slingwine , Daniel Chen , Craig A. Barkhouse
IPC分类号： G06F11/14
CPC分类号： G06F11/1435 , G06F11/0727 , G06F11/0793 , G06F2201/84
摘要： Processing for file system volume error detection and processing for resultant error correction are separated to support system availability and user satisfaction. File system volumes for storing data structures are proactively scanned while the volumes remain online to search for errors or corruptions thereon. Found errors are scheduled to be corrected, i.e., spot corrected, dependent on the severity of the identified errors, error correction scheduling and/or at the determination of a file system administrator and/or user, to assist in maintaining minimal user and file system impact. When spot correction is initialized, one file system volume at a time is taken offline for correction. Spot correction verifies prior logged corruptions for the offline volume, and if independently verified, attempts to correct the prior noted corruptions. Volumes are retained offline only for the time necessary to verify and attempt to correct prior noted volume corruptions.
摘要翻译：用于文件系统卷错检测和处理的结果错误纠正的处理被分离以支持系统可用性和用户满意度。主动扫描用于存储数据结构的文件系统卷，同时卷保持联机，以搜索其上的错误或损坏。找到的错误被安排被校正，即，根据所识别的错误的严重性，错误校正调度和/或文件系统管理员和/或用户的确定来校正，以协助维护最小的用户和文件系统影响。当点校正被初始化时，一次一个文件系统卷被脱机以进行校正。斑点校正验证离线卷的先前记录的损坏情况，如果经过独立验证，则尝试更正先前发现的损坏。卷仅在验证并尝试纠正先前引发的卷损坏所需的时间内保留脱机。

6. 发明授权

US5727209A Apparatus and method for achieving reduced overhead mutual-exclusion and maintaining coherency in a multiprocessor system utilizing execution history and thread monitoring 失效
标题翻译：利用执行历史和线程监视在多处理器系统中实现减少的开销互斥和维持一致性的装置和方法
公开(公告)号：US5727209A
公开(公告)日：1998-03-10
申请号：US742613
申请日：1996-11-01
申请人： John D. Slingwine , Paul E. McKenney
发明人： John D. Slingwine , Paul E. McKenney
IPC分类号： G06F9/46 , G06F9/30 , G06F12/12
CPC分类号： G06F9/526 , G06F9/52 , Y10S707/956 , Y10S707/99938
摘要： A substantially zero overhead mutual-exclusion apparatus and method (90, 120) is provided that allows concurrent reading and updating data while maintaining data coherency. That is, a data reading process executes the same sequence of instructions that would be executed if the data were never updated. Rather than depending exclusively on overhead-imposing locks, this mutual-exclusion mechanism tracks an execution history (138) of a thread (16, 112) to determine safe times for processing a current generation (108, 130, 131) of data updates while a next generation (110, 132, 133) of data updates is concurrently being saved. A thread is any locus of control, such as a processor. A summary of thread activity (106, 122) tracks which threads have passed through a quiescent state after the current generation of updates was started. When the last thread related to the current generation passes through a quiescent state, the summary of thread activity signals a callback processor (104, 124) that it is safe to end the current generation of updates. The callback processor then processes and erases all updates in the current generation. The next generation of updates then becomes the current generation of updates. The callback processor restarts the summary of thread activity and initiates a new next generation of updates. All data-updating threads pass through a quiescent state between the time they attempt to update data and the time the data are actually updated.
摘要翻译：提供了基本为零的开销互斥装置和方法（90,120），其允许在保持数据一致性的同时读取和更新数据。也就是说，如果数据从未被更新，则数据读取处理执行将执行的相同的指令序列。这种相互排除机制不是完全依赖于高效的锁，而是跟踪线程（16,112）的执行历史（138）以确定用于处理数据更新的当前生成（108,130,131）的安全时间，数据更新的下一代（110,132,133）同时被保存。线程是任何控制的轨迹，例如处理器。线程活动（106,122）的摘要跟踪当前生成更新后线程已经过静态状态。当与当前一代有关的最后一个线程通过静态时，线程活动的摘要将回叫处理器（104,124）发信号通知可以安全地结束当前的更新生成。然后，回调处理器处理和删除当前一代中的所有更新。下一代更新将成为当前的一代更新。回调处理器重新启动线程活动的摘要，并启动新的下一代更新。所有数据更新线程在尝试更新数据的时间和数据实际更新的时间之间经过静态。

7. 发明授权

US08607099B2 Online fault verification in a file system 有权
标题翻译：文件系统中的在线故障验证
公开(公告)号：US08607099B2
公开(公告)日：2013-12-10
申请号：US12970948
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Craig A. Barkhouse , Daniel Chan
发明人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Craig A. Barkhouse , Daniel Chan
IPC分类号： G06F11/00
CPC分类号： G06F11/1435 , G06F11/0727 , G06F11/0793 , G06F2201/84
摘要： Data structure errors, or corruptions, identified during, e.g., normal computing device system processing, file system processing or user access processing, are verified prior to the file system identifying the error for offline correction or notifying the user or system administrator a data structure error exists. Identified data structure corruptions are verified while the file system volumes are maintained online and otherwise accessible to other processing tasks and user access. Verified data structure corruptions are logged for further corrective processing. Data structure corruptions that cannot be verified, i.e., false positives, are not further processed and are not identified to file system administrators or users as corruptions, freeing the file system to concentrate on normal processing and true, verifiable errors.
摘要翻译：在文件系统识别用于离线校正的错误之前，验证在例如正常的计算设备系统处理，文件系统处理或用户访问处理期间识别的数据结构错误或损坏，或者通知用户或系统管理员数据结构错误存在在文件系统卷保持在线状态以及其他处理任务和用户访问权限的情况下，会验证已识别的数据结构损坏。记录已验证的数据结构损坏以进一步纠正处理。无法验证的数据结构损坏（即假阳性）不会进一步处理，并且不会将文件系统管理员或用户识别为损坏，从而释放文件系统以专注于正常处理和真实可验证的错误。

8. 发明申请

US20120159255A1 Online Fault Verification In A File System 有权
标题翻译：文件系统中的在线故障验证
公开(公告)号：US20120159255A1
公开(公告)日：2012-06-21
申请号：US12970948
申请日：2010-12-17
申请人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Craig A. Barkhouse , Daniel Chan
发明人： Sarosh C. Havewala , Neal R. Christiansen , John D. Slingwine , Craig A. Barkhouse , Daniel Chan
IPC分类号： G06F11/07
CPC分类号： G06F11/1435 , G06F11/0727 , G06F11/0793 , G06F2201/84
摘要： Data structure errors, or corruptions, identified during, e.g., normal computing device system processing, file system processing or user access processing, are verified prior to the file system identifying the error for offline correction or notifying the user or system administrator a data structure error exists. Identified data structure corruptions are verified while the file system volumes are maintained online and otherwise accessible to other processing tasks and user access. Verified data structure corruptions are logged for further corrective processing. Data structure corruptions that cannot be verified, i.e., false positives, are not further processed and are not identified to file system administrators or users as corruptions, freeing the file system to concentrate on normal processing and true, verifiable errors.
摘要翻译：在文件系统识别用于离线校正的错误之前，验证在例如正常的计算设备系统处理，文件系统处理或用户访问处理期间识别的数据结构错误或损坏，或者通知用户或系统管理员数据结构错误存在在文件系统卷保持在线状态以及其他处理任务和用户访问权限的情况下，会验证已识别的数据结构损坏。记录已验证的数据结构损坏以进一步纠正处理。无法验证的数据结构损坏（即假阳性）不会进一步处理，并且不会将文件系统管理员或用户识别为损坏，从而释放文件系统以专注于正常处理和真实可验证的错误。

9. 发明授权

US06219690B1 Apparatus and method for achieving reduced overhead mutual exclusion and maintaining coherency in a multiprocessor system utilizing execution history and thread monitoring 失效
标题翻译：利用执行历史和线程监视在多处理器系统中实现减少的开销互斥和维持一致性的装置和方法
公开(公告)号：US06219690B1
公开(公告)日：2001-04-17
申请号：US08742987
申请日：1996-11-01
申请人： John D. Slingwine , Paul E. McKenney
发明人： John D. Slingwine , Paul E. McKenney
IPC分类号： G06F1200
CPC分类号： G06F9/526 , G06F9/52 , Y10S707/956 , Y10S707/99938
摘要： A substantially zero overhead mutual-exclusion apparatus and method (90, 120) is provided that allows concurrent reading and updating data while maintaining data coherency. That is, a data reading process executes the same sequence of instructions that would be executed if the data were never updated. Rather than depending exclusively on overhead-imposing locks, this mutual-exclusion mechanism tracks an execution history (138) of a thread (16, 112) to determine safe times for processing a current generation (108, 130, 131) of data updates while a next generation (110, 132, 133) of data updates is concurrently being saved. A thread is any locus of control, such as a processor. A summary of thread activity (106, 122) tracks which threads have passed through a quiescent state after the current generation of updates was started. When the last thread related to the current generation passes through a quiescent state, the summary of thread activity signals a callback processor (104, 124) that it is safe to end the current generation of updates. The callback processor then processes and erases all updates in the current generation. The next generation of updates then becomes the current generation of updates. The callback processor restarts the summary of thread activity and initiates a new next generation of updates. All data-updating threads pass through a quiescent state between the time they attempt to update data and the time the data are actually updated.
摘要翻译：提供了基本为零的开销互斥装置和方法（90,120），其允许在保持数据一致性的同时读取和更新数据。也就是说，如果数据从未被更新，则数据读取处理执行将执行的相同的指令序列。这种相互排除机制不是完全依赖于高效的锁，而是跟踪线程（16,112）的执行历史（138）以确定用于处理数据更新的当前生成（108,130,131）的安全时间，数据更新的下一代（110,132,133）同时被保存。线程是任何控制的轨迹，例如处理器。线程活动（106,122）的摘要跟踪当前生成更新后线程已经过静态状态。当与当前一代有关的最后一个线程通过静态时，线程活动的摘要将回叫处理器（104,124）发信号通知可以安全地结束当前的更新生成。然后，回调处理器处理和删除当前一代中的所有更新。下一代更新将成为当前的一代更新。回调处理器重新启动线程活动的摘要，并启动新的下一代更新。所有数据更新线程在尝试更新数据的时间和数据实际更新的时间之间经过静态。

10. 发明授权

US5608893A Method for maintaining data coherency using thread activity summaries in a multicomputer system 失效
标题翻译：使用多计算机系统中的线程活动摘要来维护数据一致性的方法
公开(公告)号：US5608893A
公开(公告)日：1997-03-04
申请号：US480627
申请日：1995-06-07
申请人： John D. Slingwine , Paul E. McKenney
发明人： John D. Slingwine , Paul E. McKenney
IPC分类号： G06F9/46 , G06F12/12 , G06F13/00
CPC分类号： G06F9/526 , G06F9/52 , Y10S707/956 , Y10S707/99938
摘要： A substantially zero overhead mutual-exclusion apparatus and method (90, 120) is provided that allows concurrent reading and updating data while maintaining data coherency. That is, a data reading process executes the same sequence of instructions that would be executed if the data were never updated. Rather than depending exclusively on overhead-imposing locks, this mutual-exclusion mechanism tracks an execution history (138) of a thread (16, 112) to determine safe times for processing a current generation (108, 130, 131) of data updates while a next generation (110, 132, 133) of data updates is concurrently being saved. A thread is any locus of control, such as a processor. A summary of thread activity (106, 122) tracks which threads have passed through a quiescent state after the current generation of updates was started. When the last thread related to the current generation passes through a quiescent state, the summary of thread activity signals a callback processor (104, 124) that it is safe to end the current generation of updates. The callback processor then processes and erases all updates in the current generation. The next generation of updates then becomes the current generation of updates. The callback processor restarts the summary of thread activity and initiates a new next generation of updates. All data-updating threads pass through a quiescent state between the time they attempt to update data and the time the data are actually updated.
摘要翻译：提供了基本为零的开销互斥装置和方法（90,120），其允许在保持数据一致性的同时读取和更新数据。也就是说，如果数据从未被更新，则数据读取处理执行将执行的相同的指令序列。这种相互排除机制不是完全依赖于高效的锁，而是跟踪线程（16,112）的执行历史（138）以确定用于处理数据更新的当前生成（108,130,131）的安全时间，数据更新的下一代（110,132,133）同时被保存。线程是任何控制的轨迹，例如处理器。线程活动（106,122）的摘要跟踪当前生成更新后线程已经过静态状态。当与当前一代有关的最后一个线程通过静态时，线程活动的摘要将回叫处理器（104,124）发信号通知可以安全地结束当前的更新生成。然后，回调处理器处理和删除当前一代中的所有更新。下一代更新将成为当前的一代更新。回调处理器重新启动线程活动的摘要，并启动新的下一代更新。所有数据更新线程在尝试更新数据的时间和数据实际更新的时间之间经过静态。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式