Eclipse* TPTP Project Apache* Harmony Project A Harmonious Combo : Experiences Profiling with Eclipse* TPTP * Chris Elford -- Intel Corporation Sergey Kuksenko -- Intel Corporation Confidential | Date | Other Information, if necessary Harmony logo source: http://harmony.apache.org/ © 2008 Intel Corporation; made available under the EPL v1.0 | March 2008 © 2002 IBM Corporation Eclipse* TPTP Project Apache* Harmony Project Outline Overall project goals Introduction to Apache* Harmony The Eclipse* TPTP profiler First experiences using TPTP Analysis of advanced workload with TPTP Looking to the future 2 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Goals and Methodology of our Experiments Goals Identify Harmony performance opportunities using TPTP Integrate TPTP into our performance methodology Approach Start simple… Verify that we can recognize known performance issues w/ TPTP Understand how opportunities manifest in TPTP Understand current limitations of TPTP and Harmony JVMTI Fix critical issues along the way Grow… Move forward looking for new opportunities to optimize Harmony Use TPTP more regularly Suggest TPTP & Harmony JVMTI enhancements for our use case Most of what we show today is in the “Start Simple” category 3 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Apache* Harmony Overview Primary Goal of Harmony “a large and healthy community of those interested in runtime platforms” Creation of “A compatible, independent implementation of the Java SE 5 JDK under the Apache License v2” “A community-developed modular runtime (VM and class library) architecture.” Why? Enables innovation/research and adoption w/o license restrictions that limit consumption 4 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Source: http://harmony.apache.org/ Eclipse* TPTP Project Apache* Harmony Project Apache* Harmony Overview (continued) Why J2SE 5? “Starting with Java SE 5, as that is the first version of Java SE for which the licensing allows an open source implementation” “Continue with Java SE 6 and any subsequent versions that follow.” Who all is using Apache* Harmony technology? Since its an open license, we don’t know for sure… but to name a few http://mail-archives.apache.org/mod_mbox/harmonydev/200712.mbox/%3c4759E2CD.5010308@gmail.com%3e http://mail-archives.apache.org/mod_mbox/harmonydev/200711.mbox/%3c8084e12c0711201546m7e3cc028kf5df2763df708fd3@ mail.gmail.com%3e Harmony is a clean room implementation of J2SE spec distributed under Apache* V2 license 5 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP Source: http://harmony.apache.org/faq © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project A Brief History of Apache* Harmony May 2005: Accepted by the ASF and started in Apache* Incubator May 2006: Harmony has VM and classlib functional to run Eclipse* 3.0 May 2006: JavaOne - demonstrated functional classlib with Swing/AWT July 2006: First JRE Snapshot Published Oct 2006: Became ASF Top Level Project May 2007: First Milestone Build Released (M1) May 2007: JavaOne – “In Harmony with Eclipse” bundle CDs distributed … Dec 2007: Fourth Milestone Build Released (M4) Feb 2008: Fifth Milestone Build Released (M5) Note: We plan to run Eclipse on Harmony in the demo later in talk 6 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Apache* Harmony: What’s Inside? ~2.3 million LOC ~1.6 million Java* code, ~ 0.7 million C/C++ Serious testing and engineering infrastructure Components API: ~99% JDK5, ~90% JDK6 Project VMs: DRLVM (and others) Tools: javac, javah, javap, jarsigner, keytool, appletviewer Platform Support Systems with x86 processors supporting Intel® SSE or greater Systems using processors compatible with Intel® EM64T Intel® Itanium®-based systems Operating System Support Several versions of Windows* and Linux 7 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Apache* Harmony and Eclipse*: Synergy The Eclipse* platform was the first large application enabled on Harmony back in ’06 Harmony test base includes >20 Eclipse* scenarios Harmony performance has been tuned for Eclipse* platform Harmony adopted EUT for daily execution Harmony JVMTI has proven to work with TPTP Harmony adopted key TPTP JVMTI tests for daily execution “In Harmony with Eclipse” bundles updated regularly http://sourceforge.net/projects/eclipse-harmony/ 8 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Eclipse* TPTP Profiling Tool Overview Broadly useful for performance analysis and for gaining a deeper understanding of a Java* program Consists of the Profiling and Logging Perspective and a number of graphical and tabular views Visualize program execution and threading behavior; Pinpoints operations taking most resources; Explore patterns of program behavior Enables you to test your application's performance early in the programming development cycle for improvements Option to run application with agent at near full speed (enabled mode) and attach later at certain application phase. Assorted filtering functionality which can help to localize problem and reduce overhead for large long run applications 9 A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Obligatory Eclipse* TPTP Architecture Diagram Presentation System Target System Distributed Control Framework Agent Control Interface Artifact Management Test Creation and Execution Trace Analysis And Profiling Runtime Monitor / log Agent Control Interface Eclipse TPTP GUI Reference Perspectives And Workflow EMF Data Models A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 10 © 2008 Intel Corporation; made available under the EPL v1.0 Trace Collection Log Collection Execution Environment JVMPI Monitor Correlation Eclipse Platform Application Injection Distributed Data Collection Framework Data Collection Interface Data Loader Statistical Log Test XMI Trace Real Time Export Testability Interface Data Collection Standard Widgets and Core Plug-ins Test Engine System Performance Monitor JVMTI Monitor Eclipse* TPTP Project Apache* Harmony Project The New Java* Profiler (JVMTI) vs. JVMPI Java* 5.0 introduced new standards for profiling Prior to Java* 5.0 (Java* 1.4-) the standards and the interfaces to support profiling (JVMPI) were experimental Java* J2SE 6 eliminated JVMPI interface The new standard (JVMTI): an innovative solution to profiling; enables precise control over application parts to profile Eclipse* TPTP has embraced this innovation and the new Java profiler is based on this new standard In Eclipse* TPTP old Java profiler (JVMPI) still exists Only for backward compatibility with Java 1.4 Because of JVMPI/JVMTI interface differences there are some differences in behavior between old and new profiler Eclipse* TPTP will eventually remove Old Java* profiler A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 11 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Eclipse* TPTP Profiler Updates for Ganymede Improve overall usability via assorted fixes Using feedback from users Reduce usability diffs between JVMPI and JVMTI… e.g., SSL for secure communications e.g., Reduce overuse of dynamic attach/detach Enhancements e.g., Better multithread analysis Improved visualization Contended lock analysis Track more thread states (e.g., join points) e.g., Efficient binary transfer format e.g., Better Java* 6 support A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 12 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Motivation of using TPTP on Harmony When you have 1.6 million lines of Java* code… Using TPTP for Harmony performance analysis: startup optimizations. Team already uses profiling tools extensively Intel® VTune™ tools heavily used Correlation of architectural events to assembly code VM and JIT code quality analysis and great optimization Mainly use batch mode (execute then perform offline analysis) Want to also do higher level Java* language centered analysis Rich Java* centered analysis of call graph Package and Class and Method summarization Java* Thread and Heap analysis Ideally continue with batch mode analysis model Why TPTP Profiler? Leverage open source to create open source Maximize collaboration potential w/ engineer next door A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 13 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Harmony analysis with TPTP: Startup performance Consider simplest workloads Hello World Startup phase of a GUI application Goal Understand detailed end to end behavior Find inefficiencies in class libraries for optimization Approach Disable default filters in TPTP Collect trace Analyze Discoveries from our first use of TPTP Even for small apps, data volume can be huge w/o filters Several TPTP usability bugs; fixes targeted for Ganymede Xml4profiling lacks command line tool for tsv/csv dumps A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 14 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project A few words on filtering in TPTP Collection Time Filtering Pro: Reduce data volume and instrumentation overhead (critical) Con: Data for excluded classes/method can never be seen Model Import Filtering Pro: Reduce workbench memory footprint (critical on 32bit OSes) Con: Can only see data that is in the model View Level Filtering Pro: Remove items from being visualized in the workbench Con: Does not reduce data volume or model footprint A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 15 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Harmony performance analysis w/ TPTP: HWA startup. Sort by number of calls write(int) is at the top Callgraph: Too much time is spent when reading the manifestSort ofby signed base jar file... java.util.jar.InitManifest:.read()->java.io.ByteArrayOutputStream:.write(int) java.util.jar.IninManifest:.read() at the top time 5% boost expected from moving key signed jar to bootclasspath. A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP http://issues.apache.org/jira/browse/HARMONY-5277 16 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Harmony perf analysis w/ TPTP: EIOffice startup. See http://sourceforge.net/projects/eio-harmony/ Java.lang.reflect.Method.getName() is at the top Sort by number of calls further investigation shows that Method:.getName() is used in linear search for finding methods of class from java.lang.Class.getMethod(String) 4% boost expected by replacing linear search w/ hashed search in lookup A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP http://issues.apache.org/jira/browse/HARMONY-5284 17 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project More advanced study: SPECjbb* 2005 Analyze more complex, long running, workload Snapshot performance during workload steady-state We use SPECjbb* 2005 workload for some of our performance tests Goal Understand application behavior Find inefficiencies in class libraries for optimization Ideal Approach Disable default filters in TPTP, Collect trace, Analyze Would prefer to collect trcxml from our batch scripts then do offline analysis Discoveries from our use of TPTP Can’t get sufficient detail with filtering; w/o filtering; too much data Pause/resume helps… But doesn’t help allow batch mode execution Hope for richer control with Application mode in Ganymede Avoid attach mode Instrumentation during run perturbs performance A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 18 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 Overall Execution statistics spec.org class 3rd party library classes Apply filter • SPECjbb* 2005 source cannot be modified • 3rd party library classes hard to influence (view level filter) A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 19 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 Sort by Base Time Candidates for investigation A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 20 © 2008 Intel Corporation; made available under the EPL v1.0 Cumulative Time Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 First candidate class Hot methods • Next() and nextInt() are rather small. • Typically disappear in JIT inlining • JIT behavior invisible to TPTP • Class instrumentation impacts behavior • Message: be careful with getters/setters/etc • TPTP can be set to automatically filter them • We need to see them so we can double check JIT • No opportunity for tuning... Move along. A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 21 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 TreeMap iteration is expensive next() at the top Look into details A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 22 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 next() is hot next() invokes get() pretty much each time Problem! • Single generalized iterator impl. (e.g., KeyIterator, ValueIterator, EntryIterator) Also… • Ignores common cases for which • Creates isexcess “getter” objects. optimization possible • Inlining complicated due to interface call 5% boost from implementing three specialized iterators for common use cases A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP http://issues.apache.org/jira/browse/HARMONY-5232 23 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 Next hot classes: java.math.* Choose hot method for details A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 24 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Performance study: SPECjbb* 2005 – java.math.* (~10%) Short Quiz: Each BigDecimal.add() invokes BigInteger.add() java.math.* took ~10%? How did we get more then 30%? Do we really need toiscreate separate BigInteger Only 1 object (BigDecimal) created for each BigDecimalBigInteger, value? • Instead ofobject 3isobjects (BigDecimal, int[]) A BigDecimal a BigInteger plus a scaling factor Let’s incorporate small values (<19 digits) • Less GC• stress! directly into BigDecimal. Result: More then a 30% boost!!! A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP http://issues.apache.org/jira/browse/HARMONY-551 25 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project What to do next to analyze Harmony with TPTP? So now that we’ve tried TPTP, what is next? Use Eclipse TPTP profiler often to improve Harmony Work closely with Eclipse* TPTP developers So they better understand how we use TPTP Use heap profiler Harmony’s memory footprint is fairly large Target smaller heaps (e.g., for J2SE on smaller devices) Already touch tested it Use thread profiler Particularly excited about optimizing multithreaded Java for multicore platforms Already touch tested it A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 26 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Summarizing wish lists for TPTP and Harmony Features we would like to see in TPTP in future Better handling of huge datasets (multi-GB) Simpler and richer batch (scripted) collection Command line data reduction utility (offline analysis) Interface for filtering “easily” inlinable getter/setter methods at collection time Features we would like to see in Harmony in future Fuller JVMTI implementation e.g., RedefineClasses - needed for TPTP dynamic attach/detach Better memory footprint on large Eclipse* scenarios Further tuned Eclipse* workbench startup time Java* J2SE6 support A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 27 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Conclusion Our perspective Apache* Harmony making great progress Eclipse* TPTP allows great Java* code analysis Both projects have opportunities for further enhancement Look forward to future collaboration between Harmony and TPTP What we would like from you Go try Harmony (w/ Eclipse* platform) Tell us about your experience (and feel free to join) Go try Eclipse* TPTP profiler Tell us about your experience (and feel free to join) A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 28 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project TPTP Resources (profiling and beyond) Learn and Try http://www.eclipse.org/tptp/home/documents/conferences/eclipseCon2007 http://www.eclipse.org/tptp/home/documents/conferences/eclipseCon2007 /%283669%29%20Profiling%20Java%20applications%20using%20Eclips e%20TPTP%20v1_1.htm Webs and Wikis http://eclipse.org/TPTP http://wiki.eclipse.org/TPTP Downloads and Updates http://www.eclipse.org/tptp/home/downloads http://www.eclipse.org/tptp/home/downloads/updateManager.php News and Mail http://www.eclipse.org/tptp/home/project_info/general/mailnews.php Use and Participate http://wiki.eclipse.org/TPTP_User_Experiences_Profiling http://www.eclipse.org/tptp/home/project_info/general A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 29 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Harmony Resources Learn and Try http://harmony.apache.org/hdk.html http://harmony.apache.org/subcomponents/drlvm/index.html Webs and Wikis http://harmony.apache.org/ http://wiki.apache.org/harmony/ Downloads and Updates http://harmony.apache.org/download.cgi http://sourceforge.net/projects/eclipse-harmony/ Use and Participate http://harmony.apache.org/quickhelp_contributors.html A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 30 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Intel, the Intel logo, Itanium and VTune are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. * Other names and brands may be claimed as the property of others. A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 31 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Questions? A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 32 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project BACKUP A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 33 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Eclipse* TPTP -- The New Java* Profiler A set of libraries that attach to a JVM for recording Java application's behavior An extensible framework, consisting of a core runtime component (Martini), an agent managed by the Agent Controller (JPIAgent), and a set of data collection libraries built on top of the Martini runtime. Can be used to identify performance details such as classes or methods responsible for execution bottlenecks, analyze application heap to find memory leaks and visualize threading behavior. Output in the form of XML fragments (XML4Profiling) Can be launched from the Eclipse IDE or as a standalone program using Java command-line options Applications under test can reside in Eclipse workspace, binaries on file system, or hosted in a J2EE application server. A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 34 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Apache* Harmony Status: Overall API completeness: ~99% Remaining 1% is in Swing as well as spread over all code Non conformant – Sun* is silent on TCK issue. Compatibility is proven by a number of free and commercial apps running on Harmony VM completeness: ~100% (according to the spec) Interpreter, 2 JITs, DPGO Optimized JIT is based on StarJIT – initially tuned to show best performance on Intel® Itanium® processors – has a good large enough set of HLOs Parallel, generational GC with innovative algorithms. Two commercial-quality VMs are working with Harmony classlibs: IBM* J9* BEA* JRockit* Performance is good on server benchmarks, client still needs some attention. Large set of test suites: Unit, VTS (conformance), reliability, stress, performance, etc. A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 35 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project How we collected a snapshot w/o filters 1. Configure profiling to save to a file Destination tab in workbench 2. Configure aggregated stats with 3600 second refresh Avoid auto refresh… Only manual 3. Start the run and immediately pause collection 4. 5. 6. 7. 8. 9. Pause button… Could not find a way to start paused… Wait until steady state reached Resume collection Wait a few minutes Pause collection Wait for data to come across Terminate Do not save 10. Restart Eclipse Ensures maximum memory available for file import 11. Import from where you saved file in (1). A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 36 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Heap profile: a first try A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 37 © 2008 Intel Corporation; made available under the EPL v1.0 Eclipse* TPTP Project Apache* Harmony Project Thread profile: a first try A Harmonious Combo: Experiences Profiling Apache* Harmony with Eclipse* TPTP 38 © 2008 Intel Corporation; made available under the EPL v1.0