International Business Machines Corporation

Zhi Da Luo

Yao Dong Qi

Wei Liu

Yang Che

Wei Li

Rui Han

Xiao Jun Dai

Zhi Hua Pan

Li Jing Mu

Yong Sen Zheng

Ye Xin Wang

Apparatus, computer-readable storage medium and a method executed by a computer for tracing the memory accesses of an object-oriented program comprises assigning a unique identification to each class object created from at least one class in the object-oriented program by modifying a class definition of the at least one class. In response to an unloading of one of the class objects, obtaining class unloading related information from a runtime environment of the object-oriented program and obtaining the unique identification of the unloaded class object according to the returned information. Then releasing memory space assigned to the unloaded class object for storing the memory access information of the unloaded class object.

Method and apparatus for tracing memory accesses

A computer-implemented method for monitoring an object-oriented application is disclosed. A transaction-representative-digraph is built based upon a real-time running of the application, and the digraph contains nodes of a plurality of replaceable classes. A key node is selected based upon invocation relationships between the nodes. A monitor program is instrumented at the selected key node.

Self-monitoring object-oriented applications

Global heap allocation technologies in a multi-thread environment, and particularly a method for dynamically adjusting global heap allocation in the multi-thread environment, and more particularly to a method and system for dynamically adjusting global heap allocation by monitoring conflict parameters of the global heap allocation method. Specifically, a method of dynamically adjusting global heap allocation in multi-thread environment, comprising: identifying a global heap allocation method in an application program; judging whether the global heap allocation method is a multi-thread conflict hot point; and using a local stack to allocate memory space requested by the global heap allocation method in response to a 'yes' judging result. The method is adapted to purposefully dynamically adjust the intrinsic global heap allocation method in the program according to a real-time running state, reduce the lock contention on the global heap, and effectively improve a resource allocating efficiency and a resource utilization rate.

Dynamically adjusting global heap allocation in multi-thread environment

Global heap allocation technologies in a multi-thread environment, and particularly to a system for dynamically adjusting global heap allocation in the multi-thread environment, and more particularly to a method and system for dynamically adjusting global heap allocation by monitoring conflict parameters of the global heap allocation method. The present invention provides a system for dynamically adjusting global heap allocation in multi-thread environment, comprising: identifying a global heap allocation method in an application program; judging whether the global heap allocation method is a multi-thread conflict hot point; and using a local stack to allocate memory space requested by the global heap allocation method in response to a 'yes' judging result. The system according to the present invention is adapted to purposefully dynamically adjust the intrinsic global heap allocation method in the program according to a real-time running state, reduce the lock contention on the global heap, and effectively improve a resource allocating efficiency and a resource utilization rate.

The present invention relates to a technology for deadlock detection in a program, and more particularly relates to a technology for detecting deadlock in a program through lock graph analysis. The present invention provides a method for detecting deadlock, comprising: obtaining lock information related to locking operation in a program; generating a first lock graph based on the obtained lock information, wherein each node in the first lock graph comprises a set of locks comprising at least one lock and a set of program locations comprising at least one lock location; extracting a strongly connected sub graph in the first lock graph; unfolding the strongly connected sub graph in the first lock graph to generate a second lock graph, wherein each node in the second lock graph comprises a single lock; and extracting a strongly connected sub graph in the second lock graph, the strongly connected sub graph in the second lock graph indicating a deadlock in the program.

Method and system for detecting program deadlock

The present invention relates to application program logs. In particular, the invention relates to a method and a system in which an original log method of an application program is optimized and newly injected to the application program. According to one aspect of the present invention, there is provided a method for optimizing an application program log record, comprising: identifying an original log method of an application program; separating a character string parameter in the original log method into a constant part and a variable part; establishing a correspondence flag between the constant part and the variable part; constituting an optimized log method by the constant part, the variable part and the correspondence flag; and replacing the original log method with the optimized log method through a code injection. With the above method, the hard disk storage amount of the log record and the IO operation frequency can be reduced.

Program log record optimization

A deadlock detection method and computer system for parallel programs. A determination is made that a lock of the parallel programs is no longer used in a running procedure of the parallel programs. A node corresponding to the lock that is no longer used, and edges relating to the lock that is no longer used, are deleted from a lock graph corresponding to the running procedure of the parallel programs in order to acquire an updated lock graph. The lock graph is constructed according to a lock operation of the parallel programs. Deadlock detection is then performed on the updated lock graph.

Deadlock detection for parallel programs

A method and system for facilitating runtime memory analysis. The method includes: assigning a unique ID for each task in a running program; recording memory access events occurring during the running program, including the IDs of the task performing the memory accesses; issuing a task termination notification in response to a task terminating, the task termination notification including the ID of the terminating task; and releasing all the memory access events having the ID of the terminating task in the memory, in response to the task termination notification. This method and system can ensure that the memory access events stored in the memory will not increase unlimitedly, so that the memory overhead is reduced remarkably and dynamic memory analysis can be faster and more efficient.

Facilitating memory analysis

A method and/or system for detecting deadlock, comprising: obtaining lock information related to locking operation in a program; generating a first lock graph based on the obtained lock information, wherein each node in the first lock graph comprises a set of locks comprising at least one lock and a set of program locations comprising at least one lock location; extracting a strongly connected sub graph in the first lock graph; unfolding the strongly connected sub graph in the first lock graph to generate a second lock graph, wherein each node in the second lock graph comprises a single lock; and extracting a strongly connected sub graph in the second lock graph, the strongly connected sub graph in the second lock graph indicating a deadlock in the program.

Growing community of inventors

Beijing, China

Zhi Da Luo