公开(公告)号：EP2352090A4

公开(公告)日：2015-05-06

申请号：EP09819051

申请日：2009-08-13

Applicant: IBM

Inventor： HORII HIROSHI ,  OZAWA YOHSUKE ,  ENOKI MIKI ,  ONODERA TAMIYA

IPC: G06F9/52 ,  G06F17/30

CPC classification number: H04L67/1097 ,  G06F9/52 ,  G06F17/30171 ,  G06F17/30362

Abstract: Problem To cause a system to operate efficiently regardless of the time of day. Solving Means There is provided a system including multiple application servers for accessing shared data and a centralized control unit for centrally controlling a lock applied to the shared data by each of the application servers. Each of the application servers includes a distributed control unit for controlling a lock applied to the shared data by the application server and a selection unit for selecting any one of distributed mode in which a lock is acquired from the distributed control unit and centralized mode in which a lock is acquired from the centralized control unit.

公开(公告)号：JP2001188685A

公开(公告)日：2001-07-10

申请号：JP37173099

申请日：1999-12-27

Applicant: IBM

Inventor： ONODERA TAMIYA

IPC: G06F9/46 ,  G06F9/52

Abstract: PROBLEM TO BE SOLVED: To provide a new composite locking method not to lower the processing speed of a high frequency path. SOLUTION: When no thread to lock an object exists 1, zero is stored in both of a field for lock and a competitive bit. After that, the object is locked (light weight lock) by a certain thread, an identifier of the thread is stored in the field for lock 2. If no lock is attempted by other threads before the lock is released by the thread with the thread identifier, SPECIAL is stored in the field for lock 5 and the processing is returned to 1. When the lock is attempted by other threads before the lock is released, competition in the light weight lock is generated, therefore, the competitive bit is established to record the competition 3. After that, the competitive bit is cleared when the lock is transferred to heavy weight lock 4 and the processing of 4 is transferred to 1, if possible.

公开(公告)号：JPH0934711A

公开(公告)日：1997-02-07

申请号：JP18377695

申请日：1995-07-20

Applicant: IBM

Inventor： NAKAMURA HIROAKI ,  ONODERA TAMIYA ,  TAKEUCHI MIKIO

IPC: G06F9/44 ,  G06F9/445 ,  G06F9/54

Abstract: PROBLEM TO BE SOLVED: To execute the start of a program at high speed by loading an object from the route object of plural objects into a main memory and removing the specified object from a load object. SOLUTION: When loading an object orientated program into the main memory, a CPU 1 generates a main memory managing list by loading the route object from an HDD 5 to a main memory 3. By referring to this list, it is discriminated whether the object to be loaded is an object to be accessed from only one other object or not. In this case, when the object is discriminated as the object to be accessed only from one other object, the object to be accessed from only one other object and the object to be accessed from the object to be accessed only from one other object are removed from the load object and the main memory managing list is updated.

公开(公告)号：JP2012088812A

公开(公告)日：2012-05-10

申请号：JP2010233085

申请日：2010-10-15

Applicant: Internatl Business Mach Corp ,  インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Maschines Corporation

Inventor： KAWACHIYA KIYOKUNI ,  TACHIBORI MICHIAKI ,  ONODERA TAMIYA ,  INGLIS DEREK

IPC: G06F11/34

CPC classification number: G06F11/3644 ,  G06F8/4443

Abstract: PROBLEM TO BE SOLVED: To provide a technique for acquiring information on call context comprehensively at a low cost.SOLUTION: Suitably, an 8-bit ID is allocated to each call site, and call site IDs which are most recent and in four stages as preferable number are held as a call history of 32-bit data for each thread and are used as context information. For calling a method, a call history in a current frame is shifted by 8 bits to the left, and the call site ID of the call side is put in the low-order 8 bits and is stacked in a new frame. For returning from the method, the frame is put back to obtain the value of the original call history. For virtual calling, information of the call history is recorded in a profile table at a calling point in addition to determined jump destination information and the number of times of calling. This profile table is used in JIT compilation to specify a method to be inlined.

Abstract translation: 要解决的问题：提供以低成本全面地获取关于呼叫上下文的信息的技术。 解决方案：适当地，将8位ID分配给每个呼叫站点，并且呼叫最近和四个阶段的呼叫站点ID作为优选号码被保持为每个线程的32位数据的呼叫历史记录，并且是 用作上下文信息。 为了调用方法，当前帧中的呼叫历史向左移位8位，呼叫侧的呼叫位置ID被置于低位8位，并且堆叠成新的帧。 为了从该方法返回，将帧返回以获得原始呼叫历史的值。 对于虚拟呼叫，除了确定的跳转目的地信息和呼叫次数之外，呼叫历史的信息也记录在呼叫点的简档表中。 此配置文件表用于JIT编译，以指定要内联的方法。 版权所有（C）2012，JPO＆INPIT

公开(公告)号：JP2007183905A

公开(公告)日：2007-07-19

申请号：JP2006115275

申请日：2006-04-19

Applicant: Internatl Business Mach Corp ,  インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Maschines Corporation

Inventor： KOCHIYA KIYOKUNI ,  OGATA KAZUNORI ,  ONODERA TAMIYA

IPC: G06F9/44 ,  G06F12/00

Abstract: PROBLEM TO BE SOLVED: To provide a character string operation method for reducing the storage capacity of a storage part, a program, and an information processor.
SOLUTION: The information processor 1 for operating character strings is provided with: a storage part 30 for storing a character string group including a plurality of character strings each of which is constituted of a plurality of character data; a reference part 41 for referring to a specific character string from the character string group; a character data detection part 422 for detecting character data which are not referred to as the specific character string from the character string group; and a character data deletion part 423 for deleting the character data detected from the character string group.
COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：提供一种用于降低存储部件，程序和信息处理器的存储容量的字符串操作方法。 解决方案：用于操作字符串的信息处理器1设置有：存储部件30，用于存储包括由多个字符数据构成的多个字符串的字符串组; 用于从字符串组引用特定字符串的参考部分41; 字符数据检测部分422，用于从字符串组中检测不被称为特定字符串的字符数据; 以及字符数据删除部423，用于删除从字符串组检测到的字符数据。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2004259007A

公开(公告)日：2004-09-16

申请号：JP2003049414

申请日：2003-02-26

Applicant: Internatl Business Mach Corp ,  インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Maschines Corporation

Inventor： TAKEUCHI MIKIO ,  ONODERA TAMIYA ,  OGASAWARA TAKESHI

IPC: G06F9/45 ,  G06F9/44

CPC classification number: G06F8/443

Abstract: PROBLEM TO BE SOLVED: To optimize a program for handling character strings.
SOLUTION: This compiler device for optimizing the program for handling character strings comprises an addition instruction detection part for detecting an addition instruction for adding a character string to a character string variable storing a character string in the program; a storage code generation part for generating, in stead of each of a plurality of addition instructions for adding a character string to the same character string variable, which is detected by the addition instruction detection part, a storage code for storing data of an additional character string to be added according to the addition instruction concerned; and an addition code generation part for generating, in a position to be executed prior to an instruction referring to the character string variable in the program, an additional code for adding each of a plurality of additional character strings to the character string variable.
COPYRIGHT: (C)2004,JPO&NCIPI

公开(公告)号：JP2004152185A

公开(公告)日：2004-05-27

申请号：JP2002319006

申请日：2002-10-31

Applicant: Internatl Business Mach Corp ,  インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Maschines Corporation

Inventor： KOCHIYA KIYOKUNI ,  KOSEKI SATOSHI ,  ONODERA TAMIYA

IPC: G06F9/50 ,  G06F9/46 ,  G06F17/00

CPC classification number: G06F9/526

Abstract: PROBLEM TO BE SOLVED: To speed up acquisition of competitive resources while securing the exclusivity of the acquisition of the competitive resources when only a specific information processing part frequently acquires and releases the competitive resources. SOLUTION: The exclusive control unit 10 for allowing an information processing part to acquire the competitive resources exclusively of other information processing parts is equipped with a plurality of non-priority information processing parts 160 for acquiring exclusively one another non-priority exclusive rights which indicate a candidate of acquiring the competitive resources 50 by first processing and a priority information processing part 150 for acquiring the competitive resources 50 exclusively between itself and the non-priority information processing part 160 acquiring the non-priority exclusivity by second processing shorter in processing time than the first processing. COPYRIGHT: (C)2004,JPO

Abstract translation: 要解决的问题：只有在特定的信息处理部门经常获取和释放竞争资源时，才能加快竞争资源的获取，同时确保获得竞争资源的排他性。 解决方案：用于允许信息处理部分专门获取其他信息处理部分的竞争资源的排他控制单元10配备有多个非优先级信息处理部分160，用于仅获取另一个非优先权专有权 其指示通过第一处理获取竞争资源50的候选者和优先级信息处理部分150，用于通过在处理中较短的第二处理来获取独占于非优先级信息处理部分160的非优先级信息处理部分160之间的竞争资源50 时间比第一次处理。 版权所有（C）2004，JPO

公开(公告)号：JP2001273141A

公开(公告)日：2001-10-05

申请号：JP2000074223

申请日：2000-03-16

Applicant: IBM

Inventor： ONODERA TAMIYA

IPC: G06F9/44 ,  G06F9/445

Abstract: PROBLEM TO BE SOLVED: To provide an object state representing method which is fast and whose influence to the performance efficiency of a processing system is very small. SOLUTION: This computer system on which an object oriented programming language is mounted is provided with an object 10, a model information block 20 and a class structure 30 as object run-time representation in the mounted programming language, the object 10 is provided with a header 11 including a pointer to the block 20, the block 20 is provided with a class pointer indicating the structure 30, a plurality of class pointers of the block 20 are provided in accordance with the number of states of the object, an individual class pointer is made to correspond to the individual state of the object 10, and the pointer included in the header 11 of the object 10 represents the state of the object by individually indicating the plurality of class pointers by using several low- order bits.

公开(公告)号：JPH09282167A

公开(公告)日：1997-10-31

申请号：JP8785796

申请日：1996-04-10

Applicant: IBM

Inventor： ONODERA TAMIYA ,  NAKAMURA HIROAKI

IPC: G06F9/44 ,  G06F9/42

Abstract: PROBLEM TO BE SOLVED: To improve memory efficiency and to transmit a message at a high speed. SOLUTION: By a specified receiver object, the class object C is obtained 110. An instance variable for indicating the dispatch chart of the class is provided in the class object C and the dispatch chart D is obtained 120. From a code which is the argument of a vmicall instruction, the method M of an address stored in the order of the code of the dispatch chart is taken out 130. A card which is the other argument of the vmicall instruction and a tag number stored in the method M taken out in a step 130 are compared 140. When the card and the tag number do not match, a selector non-matching handier is called and a correct method to be called is searched 150. Then, the method M is activated 160.

公开(公告)号：JP2013161186A

公开(公告)日：2013-08-19

申请号：JP2012021187

申请日：2012-02-02

Applicant: Internatl Business Mach Corp ,  インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Machines Corporation

Inventor： KAWACHIYA KIYOKUNI ,  ONODERA TAMIYA ,  TAKEUCHI MIKIO ,  SALIKH ZAKIROV

IPC: G06F9/44 ,  G06F12/00

CPC classification number: G06F12/0276 ,  G06F9/4428 ,  G06F12/0253 ,  G06F17/30097

Abstract: PROBLEM TO BE SOLVED: To increase variation in object hash codes in order to reduce conflicts between object hash codes, and enable object hash codes to be added to pinned objects.SOLUTION: A per-heap seed is provided in each heap area in which an object is placed in order to generate an object hash code, and at least an object address and the per-heap seed are combined to generate the object hash code. Further, it is determined whether an object hash code can be added to an object in a "01" state during GC. If it can be added, the object hash code is calculated and added to the object, and the state is changed to "10". If it cannot be added, the object hash code is calculated, placed outside the heap, and associated with the object, and the state is changed to "11", which represents a pinned object.

Abstract translation: 要解决的问题：增加对象哈希码的变化，以减少对象哈希码之间的冲突，并使对象哈希码被添加到固定对象。解决方案：每个堆区域中提供每堆种子，其中 放置对象以生成对象散列码，并且至少将对象地址和每堆种子组合以生成对象散列码。 此外，在GC期间确定是否可以将对象散列码添加到处于“01”状态的对象。 如果可以添加，则对象哈希码进行计算并添加到对象中，状态变为“10”。 如果不能添加，则对象散列码进行计算，放置在堆外部并与对象相关联，状态更改为“11”，表示固定对象。