International Business Machines Corporation

Moshe Klausner

Gad Haber

Vadim Eisenberg

Nitzan Peleg

Yaakov Yaari

Aviad Zlotnick

Orna Raz

Rachel Tzoref-Brill

Onn Menahem Shehory

Yousef Shajrawi

Yoram Simha Adler

Daniel Citron

Aharon Kupershtok

Grigori Chtrasberg

Itzhack Goldberg

Shmuel Ben-Yehuda

Ronen Itshak Kat

Marcel Zalmanovici

Rachel Tzoref

Maxim Gurevich

An aspect includes performance profiling of an application. A processor executes an instruction stream of the application including instructions that are dynamically grouped at run-time. The processor monitors for an event associated with sampled instructions. A sampled instruction is associated with other events that include instruction grouping information. A number of the instructions in a group that includes the sampled instruction is determined as a group size. The monitored event is tracked as separate events with respect to each of the sampled instruction and one or more other instructions of the group. Subsequent monitored events are tracked as the separate events for each of the instructions from additional groups having various group sizes formed from a sequence of the instructions. An execution count for the sequence of the instructions is generated based on accumulating the separate events over a period of time.

Basic block profiling using grouping events

A method comprising: counting each occurrence of a hardware event by a Performance Monitoring Counter of a hardware processor during the execution of a target program code; orderly and continuously storing in a buffer of a Taken Branch Trace (TBT) Facility of said hardware processor a predefined TBT size of last taken branches of said target program code during its execution; every time said counting equals a sampling rate, triggering sampling of said buffer, to receive a TBT comprising current said predefined TBT size of last taken branches; constructing a full branch trace for each said TBT based on said target program code; extracting a predefined Chopped Branch Trace (CBT) size of last branches from each said full branch trace, to receive a chopped branch trace for said each TBT; and incrementally storing each said chopped branch trace to generate an edge profile of said target program code.

Hardware-based edge profiling

Batch processing machines, systems and methods are provided. The method comprises copying target data from a target location to a first data storage medium, wherein the target data stored in the first data storage medium is utilized by a batch process; executing the batch process; checkpointing state information for the batch process at one or more time intervals while the batch process is executing; determining whether a data record in the target data stored in the target location was updated while the batch process was executing; and reinstating the batch process from a point in time prior to an earliest point in the batch process when a representation of an updated data record in the target data in the target location was accessed by the batch process.

Restarting a batch process from an execution point

Test coverage is enhanced by measuring various types of coupling between coverage tasks. The coupling measurements may be implicit coupling measurements, explicit coupling measurements, coding coupling measurements, performance coupling measurements, resource coupling measurements or the like. Coupling scores are calculated for coverage tasks and based thereon ranking of the coverage tasks or groups of coverage tasks may be determined The ranking may be utilized in selecting for which uncovered coverage task a test should be designed. The ranking may be utilized in computing a coverage measurement of a test suite. The ranking may be utilized to rank tests, based on the coverage tasks each test covers. Ranking of tests may be utilized for various purposes such as performing test selection.

Measuring coupling between coverage tasks and use thereof

Test coverage is enhanced by measuring various types of coupling between coverage tasks. The coupling measurements may be implicit coupling measurements, explicit coupling measurements, coding coupling measurements, performance coupling measurements, resource coupling measurements or the like. Based on the coupling measurements, different coverage tasks may be grouped together. For example, closely coupled coverage tasks may be grouped together. The groups may also be determined based on an initial distribution of groups, by combining groups having closely coupled member coverage tasks. The groups may be ordered and prioritized, such as based on the size of the groups and the number of uncovered tasks in each group. The groups may also be ordered, such as based on coupling score which aggregate the coupling measurements of the member coverage tasks.

Detecting optimization opportunities is enabled by utilizing a trace of a target concurrent computer program and determining a relation between data objects accessed during the tracked execution. The relation may be stored in a Temporal Relation Graph (TRG), in an extended-TRG or another data structure. The relation may be affected by temporally-adjacent accesses to data objects. The relation may further be affected by accesses to data objects performed during critical sections of the target program.

Performance optimization based on data accesses during critical sections

Systems and methods for cache optimization are provided. The method comprises tracing objects instantiated during execution of a program code under test according to type of access by one or more threads running in parallel, wherein said tracing provides information about order in which different instances of one or more objects are accessed by said one or more threads and whether the type of access is a read operation or a write operation; and utilizing tracing information to build a temporal relationship graph (TRG) for the accessed objects, wherein the objects are represented by nodes in the TRG and at least two types of edges for connecting the nodes are defined.

Detecting and optimizing false sharing

A method, system, and computer program product for operating system (OS) aided code coverage are provided. The method includes reading context information associated with a software process in response to a context switching event in an OS, the OS initiating the reading of the context information and controlling scheduling of the software process. The method further includes determining coverage information for code implementing the software process as a function of the context information in response to the OS reading the context information, and storing the coverage information as coverage data.

Operating system aided code coverage

A novel and useful apparatus for and method of associating a dedicated coverage bit to each instruction in a software system. Coverage bits are set every time the software application runs, enabling a more comprehensive and on-going code coverage analysis. The code coverage bit mechanism enables code coverage analysis for all installations of a software application, not just software in development mode or at a specific installation. Code coverage bits are implemented in either the instruction set architecture (ISA) of the central processing unit, the executable file of a software application, a companion file to the executable file or a code coverage table residing in memory of the computer system.

Apparatus for and method of life-time test coverage for executable code

A method for improving program performance including reordering a global data area of a program and for each load instruction referencing global variables within range of the immediate part of an add immediate instruction from a TOC anchor, replacing the load instruction with an add immediate instruction. The method may further include placing a TOC at the top, or within a predetermined distance from the top, of the global data area. The method may also include placing the global variables after the TOC, wherein more frequently referenced global variable are closer to the TOC than less frequently referenced global variables. Also, the method may further include placing in run-time order, groups of the global variables that frequently follow each other in run-time.

Reducing load instructions via global data reordering

A method for post-link code optimization by identifying cold caller functions of a hot callee function, and percolating store and restore instructions with respect to non-volatile registers from the callee function to the caller function.

Percolating hot function store/restores to colder calling functions

A post-link optimization method for renaming non-volatile registers with unused volatile registers, then removing the non-volatile register store/restore instructions from a function prolog/epilog. The volatile registers are saved before each function call and restored immediately afterwards. The optimization method is worthwhile if all of the call instructions are rarely invoked relative to the function prolog/epilog. The method requires that the hot function be disassembled, but does not require the full control flow graph.

Eliminating store/restores within hot function prolog/epilogs using volatile registers

A post-link optimization method for removing non-volatile register store/restore instructions from a hot function prolog/epilog, when the non-volatile register is referenced only in cold sections of code within the hot function. The method requires that the hot function be disassembled, but does not require the full control flow graph.

Eliminating cold register store/restores within hot function prolog/epilogs

A method for code optimization includes disassembling object code that has been compiled and linked, and identifying a function in the disassembled code, the function including store and restore instructions with respect to a register. The disassembled code is analyzed to locate a call to the function followed by an instruction to kill the register. The code is modified so as to eliminate the store instruction from the function when the function is invoked by the located call.

Growing community of inventors

Ramat Yishay, Israel

Moshe Klausner