公开(公告)号：US07882498B2

公开(公告)日：2011-02-01

申请号：US11278329

申请日：2006-03-31

申请人： Guilherme D. Ottoni ,  Xinmin Tian ,  Hong Wang ,  Richard A. Hankins ,  Wei Li ,  John Shen

发明人： Guilherme D. Ottoni ,  Xinmin Tian ,  Hong Wang ,  Richard A. Hankins ,  Wei Li ,  John Shen

IPC分类号： G06F9/45

CPC分类号： G06F8/456 ,  G06F8/314

摘要： Provided are a method, system, and program for parallelizing source code with a compiler. Source code including source code statements is received. The source code statements are processed to determine a dependency of the statements. Multiple groups of statements are determined from the determined dependency of the statements, wherein statements in one group are dependent on one another. At least one directive is inserted in the source code, wherein each directive is associated with one group of statements. Resulting threaded code is generated including the inserted at least one directive. The group of statements to which the directive in the resulting threaded code applies are processed as a separate task. Each group of statements designated by the directive to be processed as a separate task may be processed concurrently with respect to other groups of statements.

摘要翻译： 提供了一种用于将源代码并行化为编译器的方法，系统和程序。 收到包含源代码语句的源代码。 处理源代码语句以确定语句的依赖关系。 根据确定的语句依赖关系确定多组语句，其中一组中的语句彼此依赖。 在源代码中插入至少一个指令，其中每个指令与一组语句相关联。 产生的结果线程代码包括插入的至少一个指令。 生成的线程代码中的指令所适用的语句组被处理为单独的任务。 指定为要作为单独任务处理的指令的每组语句可以与其他语句组并发处理。

公开(公告)号：US08205200B2

公开(公告)日：2012-06-19

申请号：US11289803

申请日：2005-11-29

申请人： Shih-wei Liao ,  Ryan N. Rakvic ,  Richard A. Hankins ,  Hong Wang ,  Gansha Wu ,  Guei-Yuan Lueh ,  Xinmin Tian ,  Paul M. Petersen ,  Sanjiv Shah ,  Trung Diep ,  John Shen ,  Gautham Chinya

发明人： Shih-wei Liao ,  Ryan N. Rakvic ,  Richard A. Hankins ,  Hong Wang ,  Gansha Wu ,  Guei-Yuan Lueh ,  Xinmin Tian ,  Paul M. Petersen ,  Sanjiv Shah ,  Trung Diep ,  John Shen ,  Gautham Chinya

IPC分类号： G06F9/44 ,  G06F9/46

CPC分类号： G06F9/485 ,  G06F9/4881

摘要： Method, apparatus and system embodiments to schedule user-level OS-independent “shreds” without intervention of an operating system. For at least one embodiment, the shred is scheduled for execution by a scheduler routine rather than the operating system. The scheduler routine may receive compiler-generated hints from a compiler. The compiler hints may be generated by the compiler without user-provided pragmas, and may be passed to the scheduler routine via an API-like interface. The interface may include a scheduling hint data structure that is maintained by the compiler. Other embodiments are also described and claimed.

摘要翻译： 方法，装置和系统实施例来调度用户级别的与OS无关的“碎片”，而不需要操作系统的干预。 对于至少一个实施例，碎片被调度为由调度器例程而不是操作系统执行。 调度程序例程可以从编译器接收编译器生成的提示。 编译器提示可能由编译器生成，而不需要用户提供的编译指示，并且可以通过类API接口传递给调度程序。 接口可以包括由编译器维护的调度提示数据结构。 还描述和要求保护其他实施例。

公开(公告)号：US20060070047A1

公开(公告)日：2006-03-30

申请号：US10950693

申请日：2004-09-28

申请人： Satish Narayanasamy ,  Hong Wang ,  John Shen ,  Roni Rosner ,  Yoav Almog ,  Naftali Schwartz ,  Gerolf Hoflehner ,  Daniel LaVery ,  Wei Li ,  Xinmin Tian ,  Milind Girkar ,  Perry Wang

发明人： Satish Narayanasamy ,  Hong Wang ,  John Shen ,  Roni Rosner ,  Yoav Almog ,  Naftali Schwartz ,  Gerolf Hoflehner ,  Daniel LaVery ,  Wei Li ,  Xinmin Tian ,  Milind Girkar ,  Perry Wang

IPC分类号： G06F9/45

CPC分类号： G06F8/443 ,  G06F8/433 ,  G06F8/451

摘要： Embodiments of the present invention provide a method, apparatus and system which may include splitting a dependency chain into a set of reduced-width dependency chains; mapping one or more dependency chains onto one or more clustered dependency chain processors, wherein an issue-width of one or more of the clusters is adapted to accommodate a size of the dependency chains; and/or processing in parallel a plurality of dependency chains of a trace. Other embodiments are described and claimed.

公开(公告)号：US08037465B2

公开(公告)日：2011-10-11

申请号：US11242489

申请日：2005-09-30

申请人： Xinmin Tian ,  Milind Girkar ,  David C. Sehr ,  Richard Grove ,  Wei Li ,  Hong Wang ,  Chris Newburn ,  Perry Wang ,  John Shen

发明人： Xinmin Tian ,  Milind Girkar ,  David C. Sehr ,  Richard Grove ,  Wei Li ,  Hong Wang ,  Chris Newburn ,  Perry Wang ,  John Shen

IPC分类号： G06F9/44 ,  G06F9/45

CPC分类号： G06F8/45

摘要： Thread-data affinity optimization can be performed by a compiler during the compiling of a computer program to be executed on a cache coherent non-uniform memory access (cc-NUMA) platform. In one embodiment, the present invention includes receiving a program to be compiled. The received program is then compiled in a first pass and executed. During execution, the compiler collects profiling data using a profiling tool. Then, in a second pass, the compiler performs thread-data affinity optimization on the program using the collected profiling data.

摘要翻译： 线程数据亲和度优化可以在编译要在高速缓存相干非均匀内存访问（cc-NUMA）平台上执行的计算机程序时由编译器执行。 在一个实施例中，本发明包括接收要编译的程序。 接收的程序然后被编译成第一遍并被执行。 在执行期间，编译器使用分析工具收集分析数据。 然后，在第二遍，编译器使用收集的分析数据对程序执行线程数据关联优化。

公开(公告)号：US20070234276A1

公开(公告)日：2007-10-04

申请号：US11278329

申请日：2006-03-31

申请人： Guilherme Ottoni ,  Xinmin Tian ,  Hong Wang ,  Richard Hankins ,  Wei Li ,  John Shen

发明人： Guilherme Ottoni ,  Xinmin Tian ,  Hong Wang ,  Richard Hankins ,  Wei Li ,  John Shen

IPC分类号： G06F9/44

CPC分类号： G06F8/456 ,  G06F8/314

摘要： Provided are a method, system, and program for parallelizing source code with a compiler. Source code including source code statements is received. The source code statements are processed to determine a dependency of the statements. Multiple groups of statements are determined from the determined dependency of the statements, wherein statements in one group are dependent on one another. At least one directive is inserted in the source code, wherein each directive is associated with one group of statements. Resulting threaded code is generated including the inserted at least one directive. The group of statements to which the directive in the resulting threaded code applies are processed as a separate task. Each group of statements designated by the directive to be processed as a separate task may be processed concurrently with respect to other groups of statements.

摘要翻译： 提供了一种用于将源代码并行化为编译器的方法，系统和程序。 收到包含源代码语句的源代码。 处理源代码语句以确定语句的依赖关系。 根据确定的语句依赖关系确定多组语句，其中一组中的语句彼此依赖。 在源代码中插入至少一个指令，其中每个指令与一组语句相关联。 产生的结果线程代码包括插入的至少一个指令。 生成的线程代码中的指令所适用的语句组被处理为单独的任务。 指定为要作为单独任务处理的指令的每组语句可以与其他语句组并发处理。

公开(公告)号：US07603546B2

公开(公告)日：2009-10-13

申请号：US10950693

申请日：2004-09-28

申请人： Satish Narayanasamy ,  Hong Wang ,  John Shen ,  Roni Rosner ,  Yoav Almog ,  Naftali Schwartz ,  Gerolf Hoflehner ,  Daniel LaVery ,  Wei Li ,  Xinmin Tian ,  Milind Girkar ,  Perry Wang

发明人： Satish Narayanasamy ,  Hong Wang ,  John Shen ,  Roni Rosner ,  Yoav Almog ,  Naftali Schwartz ,  Gerolf Hoflehner ,  Daniel LaVery ,  Wei Li ,  Xinmin Tian ,  Milind Girkar ,  Perry Wang

IPC分类号： G06F9/00 ,  G06F9/24 ,  G06F15/177

CPC分类号： G06F8/443 ,  G06F8/433 ,  G06F8/451

摘要： Embodiments of the present invention provide a method, apparatus and system which may include splitting a dependency chain into a set of reduced-width dependency chains; mapping one or more dependency chains onto one or more clustered dependency chain processors, wherein an issue-width of one or more of the clusters is adapted to accommodate a size of the dependency chains; and/or processing in parallel a plurality of dependency chains of a trace. Other embodiments are described and claimed.

摘要翻译： 本发明的实施例提供了一种方法，装置和系统，其可以包括将依赖链分解成一组缩减宽度的依赖性链; 将一个或多个依赖关系链映射到一个或多个聚类依赖链处理器上，其中一个或多个所述簇的问题宽度适于适应所述依赖链的大小; 和/或并行处理多个跟踪的依赖性链。 描述和要求保护其他实施例。

公开(公告)号：US20070079298A1

公开(公告)日：2007-04-05

申请号：US11242489

申请日：2005-09-30

申请人： Xinmin Tian ,  Milind Girkar ,  David Sehr ,  Richard Grove ,  Wei Li ,  Hong Wang ,  Chris Newburn ,  Perry Wang ,  John Shen

发明人： Xinmin Tian ,  Milind Girkar ,  David Sehr ,  Richard Grove ,  Wei Li ,  Hong Wang ,  Chris Newburn ,  Perry Wang ,  John Shen

IPC分类号： G06F9/45

CPC分类号： G06F8/45

摘要： Thread-data affinity optimization can be performed by a compiler during the compiling of a computer program to be executed on a cache coherent non-uniform memory access (cc-NUMA) platform. In one embodiment, the present invention includes receiving a program to be compiled. The received program is then compiled in a first pass and executed. During execution, the compiler collects profiling data using a profiling tool. Then, in a second pass, the compiler performs thread-data affinity optimization on the program using the collected profiling data.

摘要翻译： 线程数据亲和度优化可以在编译要在高速缓存相干非均匀内存访问（cc-NUMA）平台上执行的计算机程序时由编译器执行。 在一个实施例中，本发明包括接收要编译的程序。 接收的程序然后被编译成第一遍并被执行。 在执行期间，编译器使用分析工具收集分析数据。 然后，在第二遍，编译器使用收集的分析数据对程序执行线程数据关联优化。

公开(公告)号：US20160019067A1

公开(公告)日：2016-01-21

申请号：US14866875

申请日：2015-09-26

申请人： Hong Wang ,  John P. Shen ,  Edward T. Grochowski ,  Richard A. Hankins ,  Gautham N. Chinya ,  Bryant E. Bigbee ,  Shivnandan D. Kaushik ,  Xiang Chris Zou ,  Per Hammarlund ,  Scott Dion Rodgers ,  Xinmin Tian ,  Anil Aggawal ,  Prashant Sethi ,  Baiju V. Patel ,  James P Held

发明人： Hong Wang ,  John P. Shen ,  Edward T. Grochowski ,  Richard A. Hankins ,  Gautham N. Chinya ,  Bryant E. Bigbee ,  Shivnandan D. Kaushik ,  Xiang Chris Zou ,  Per Hammarlund ,  Scott Dion Rodgers ,  Xinmin Tian ,  Anil Aggawal ,  Prashant Sethi ,  Baiju V. Patel ,  James P Held

IPC分类号： G06F9/38 ,  G06F9/30

CPC分类号： G06F9/3867 ,  G06F9/30003 ,  G06F9/30043 ,  G06F9/3005 ,  G06F9/3009 ,  G06F9/30145 ,  G06F9/3017 ,  G06F9/30174 ,  G06F9/3851 ,  G06F9/4843 ,  G06F9/4881

摘要： In an embodiment, a method is provided. The method includes managing user-level threads on a first instruction sequencer in response to executing user-level instructions on a second instruction sequencer that is under control of an application level program. A first user-level thread is run on the second instruction sequencer and contains one or more user level instructions. A first user level instruction has at least 1) a field that makes reference to one or more instruction sequencers or 2) implicitly references with a pointer to code that specifically addresses one or more instruction sequencers when the code is executed.

公开(公告)号：US08607235B2

公开(公告)日：2013-12-10

申请号：US11027445

申请日：2004-12-30

申请人： Richard A. Hankins ,  Hong Wang ,  Gautham N. Chinya ,  Trung A. Diep ,  Shivnandan D. Kaushik ,  Bryant E. Bigbee ,  John P. Shen ,  Asit K. Mallick ,  Baiju V. Patel ,  James Paul Held ,  Milind B. Girkar ,  Prashant Sethi ,  Xinmin Tian

发明人： Richard A. Hankins ,  Hong Wang ,  Gautham N. Chinya ,  Trung A. Diep ,  Shivnandan D. Kaushik ,  Bryant E. Bigbee ,  John P. Shen ,  Asit K. Mallick ,  Baiju V. Patel ,  James Paul Held ,  Milind B. Girkar ,  Prashant Sethi ,  Xinmin Tian

IPC分类号： G06F9/46

CPC分类号： G06F9/4843

摘要： Method, apparatus and system embodiments to schedule OS-independent “shreds” without intervention of an operating system. For at least one embodiment, the shred is scheduled for execution by a scheduler routine rather than the operating system. A scheduler routine may run on each enabled sequencer. The schedulers may retrieve shred descriptors from a queue system. The sequencer associated with the scheduler may then execute the shred described by the descriptor. Other embodiments are also described and claimed.

摘要翻译： 方法，装置和系统实施例来调度不依赖于操作系统的OS无关的“碎片”。 对于至少一个实施例，碎片被调度为由调度器例程而不是操作系统执行。 调度程序例程可以在每个启用的顺控程序上运行。 调度器可以从队列系统检索碎片描述符。 与调度器相关联的定序器然后可以执行由描述符描述的碎片。 还描述和要求保护其他实施例。

公开(公告)号：US20140115594A1

公开(公告)日：2014-04-24

申请号：US14067223

申请日：2013-10-30

申请人： Richard A. Hankins ,  Hong Wang ,  Gautham N. Chinya ,  Trung A. Diep ,  Shivnandan D. Kaushik ,  Bryant E. Bigbee ,  John P. Shen ,  Asit K. Mallick ,  Baiju V. Patel ,  James Paul Held ,  Milind B. Girkar ,  Prashant Sethi ,  Xinmin Tian

发明人： Richard A. Hankins ,  Hong Wang ,  Gautham N. Chinya ,  Trung A. Diep ,  Shivnandan D. Kaushik ,  Bryant E. Bigbee ,  John P. Shen ,  Asit K. Mallick ,  Baiju V. Patel ,  James Paul Held ,  Milind B. Girkar ,  Prashant Sethi ,  Xinmin Tian

IPC分类号： G06F9/48

CPC分类号： G06F9/4843

摘要： Method, apparatus and system embodiments to schedule OS-independent “shreds” without intervention of an operating system. For at least one embodiment, the shred is scheduled for execution by a scheduler routine rather than the operating system. A scheduler routine may run on each enabled sequencer. The schedulers may retrieve shred descriptors from a queue system. The sequencer associated with the scheduler may then execute the shred described by the descriptor. Other embodiments are also described and claimed.

摘要翻译： 方法，装置和系统实施例来调度不依赖于操作系统的OS无关的“碎片”。 对于至少一个实施例，碎片被调度为由调度器例程而不是操作系统执行。 调度程序例程可以在每个启用的顺控程序上运行。 调度器可以从队列系统检索碎片描述符。 与调度器相关联的定序器然后可以执行由描述符描述的碎片。 还描述和要求保护其他实施例。