DOCSLIB.ORG

Automatic Non-Functional Testing and Tuning of Configurable Generators Mohamed Boussaa

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Automatic Non-Functional Testing and Tuning of Configurable Generators Mohamed Boussaa Automatic non-functional testing and tuning of configurable generators Mohamed Boussaa To cite this version: Mohamed Boussaa. Automatic non-functional testing and tuning of configurable generators. Math- ematical Software [cs.MS]. Université Rennes 1, 2017. English. NNT : 2017REN1S011. tel- 01598821v2 HAL Id: tel-01598821 https://tel.archives-ouvertes.fr/tel-01598821v2 Submitted on 19 Oct 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. ANNÉE 2017 THÈSE / UNIVERSITÉ DE RENNES 1 sous le sceau de l’Université Bretagne Loire pour le grade de DOCTEUR DE L’UNIVERSITÉ DE RENNES 1 Mention : Informatique École doctorale Matisse présentée par Mohamed BOUSSAA préparée à l’unité de recherche INRIA INRIA Rennes Bretagne Atlantique ISTIC Thèse soutenue à Rennes le 6 Septembre 2017 Automatic devant le jury composé de : Non-functional Hélène WAESELYNCK Directrice de recherche, LAAS-CNRS Toulouse / Testing and Tuning Rapporteuse Philippe MERLE Chargé de recherche, INRIA Lille / Rapporteur of Configurable Erven ROHOU Directeur de recherche, INRIA Rennes / Examinateur Generators Franck FLEUREY Directeur de recherche, SINTEF Oslo / Examinateur Jean-Marie MOTTU Maître de conférences, Université de Nantes / Examinateur Gerson SUNYÉ Maître de conférences, Université de Nantes / Examinateur Benoit BAUDRY Directeur de recherche, INRIA Rennes / Directeur de thèse Olivier BARAIS Professeur, Université de Rennes 1 / Co-directeur de thèse Acknowledgements This thesis would not have been completed without the help of others. I would like to take this opportunity to express my gratitude towards them and acknowledge them. First of all, I would like to offer my deepest gratitude to my supervisors Olivier Barais and Benoit Baudry for their useful advices, and for guiding my steps over more than three years. Special thanks are also due to my \third" supervisor Gesron Suny´e that always knew how to motivate me and help me to overcome the difficult times in my thesis. I owe special thanks to my Master's supervisor Marouane Kessentini, with whom I discovered the software engineering research world, for encouraging me to continue in the research field as a PhD candidate. I would also like to thank the members of my jury. My examiners, H´el`eneWaeselynck and Philippe Merle for having accepted to serve on my examination board and for the time they have invested in reading and evaluating this thesis. Thanks are also due to Erven Rohou for agreeing to preside over my jury. I would also like to thank Jean-Marie Mottou and Franck Fleury for attending as members of the jury. I greatly appreciate and wish to thank all the (former and current) members of the DiverSE research team for providing a great and friendly work atmosphere, and for all the different discussions, seminars, and DiverSE coffees. Finally, I would like to thank my family. Even so far away from home, they are always present in my life during these three years. I first express my gratitude to my parents Najoua and Noureddine for their support, comprehension, confidence and, especially for their patience. You have always been behind me and believed in me. Then, thanks to my sisters Imene and Jihene for encouraging me to succeed in everything I do. By now, we can say that we have a Doctor in the Boussaa family ! i Contents R´esum´een Fran¸cais vii Contexte........................................ vii Motivations...................................... vii Probl´ematiques.................................... ix Contributions..................................... x 1 Introduction1 1.1 Context ..................................... 1 1.2 Motivation .................................... 2 1.3 Scope of the thesis ............................... 4 1.4 Challenges .................................... 5 1.5 Contributions .................................. 6 1.6 Overview of this thesis ............................. 8 1.7 Publications ................................... 9 I Background and State of the Art 11 2 Background 13 2.1 Diversity in software engineering ........................ 14 2.1.1 Hardware heterogeneity ........................ 14 iii iv CONTENTS 2.1.2 Software diversity............................ 16 2.1.3 Matching software diversity to heterogeneous hardware: the marriage 18 2.2 From classical software development to generative programming . 22 2.2.1 Overview of the generative software development process . 23 2.2.2 Automatic code generation in GP: a highly configurable process . 26 2.2.3 Stakeholders and their roles for testing and tuning generators . 28 2.3 Testing code generators ............................. 29 2.3.1 Testing workflow ............................ 30 2.3.2 Types of code generators ........................ 30 2.3.3 Why testing code generators is complex? . 33 2.4 Compilers auto-tuning ............................. 34 2.4.1 Code optimization ........................... 35 2.4.2 Why compilers auto-tuning is complex? . 36 2.5 Summary: challenges for testing and tuning configurable generators . 37 3 State of the art 39 3.1 Testing code generators ............................. 40 3.1.1 Functional testing of code generators . 40 3.1.2 Non-functional testing of code generators . 46 3.2 Compilers auto-tuning techniques ....................... 51 3.2.1 Iterative compilation .......................... 51 3.2.2 Implementation of the iterative compilation process . 51 3.2.3 Iterative compilation search techniques . 53 3.3 Lightweight system virtualization for automatic software testing . 61 3.3.1 Application in software testing ..................... 62 3.3.2 Application in runtime monitoring . 64 3.4 Summary & open challenges .......................... 65 v II Contributions 67 To the reader: summary of contributions 69 4 Automatic non-functional testing of code generator families 73 4.1 Context and motivations............................ 74 4.1.1 Code generator families ......................... 74 4.1.2 Issues when testing a code generator family . 76 4.2 The traditional process for non-functional testing of a code generator family 76 4.3 Approach overview ............................... 78 4.3.1 An infrastructure for non-functional testing using system containers 78 4.3.2 A metamorphic testing approach for automatic detection of code generator inconsistencies ........................ 80 4.4 Evaluation .................................... 89 4.4.1 Experimental setup ........................... 90 4.4.2 Experimental methodology and results . 93 4.4.3 Threats to validity . 102 4.5 Conclusion .................................... 103 5 NOTICE: An approach for auto-tuning compilers 105 5.1 Motivation .................................... 107 5.2 Evolutionary exploration of compiler optimizations . 110 5.2.1 Novelty search adaptation . 110 5.2.2 Novelty search for multi-objective optimization . 113 5.3 Evaluation .................................... 114 5.3.1 Research questions . 114 5.3.2 Experimental setup . 115 5.3.3 Experimental methodology and results . 118 5.3.4 Discussions ............................... 127 vi CONTENTS 5.3.5 Threats to validity ........................... 128 5.3.6 Tool support overview ......................... 128 5.4 Conclusion.................................... 131 6 A lightweight execution environment for automatic generators testing 133 6.1 Introduction ................................... 133 6.2 System containers as a lightweight execution environment . 134 6.3 Runtime Monitoring Engine . 136 6.3.1 Monitoring Container . 136 6.3.2 Back-end Database Container . 137 6.3.3 Front-end Visualization Container . 138 6.4 The generator case study ............................ 138 6.5 Conclusion .................................... 141 III Conclusion and Perspectives 143 7 Conclusion and perspectives 145 7.1 Summary of contributions . 145 7.2 Perspectives ................................... 147 References 151 List of Figures 171 List of Tables 173 R´esum´een Fran¸cais Contexte Les techniques ´emergentes de l'ing´enierie dirig´eepar les mod`eleset de la programma- tion g´en´erative ont permis la cr´eationde plusieurs g´en´erateurs(g´en´erateurs de code et compilateurs). Ceux-ci sont souvent utilis´esafin de faciliter le d´eveloppement logiciel et automatiser le processus de g´en´erationde code `apartir des sp´ecificationsabstraites. En effet, les utilisateurs sont devenus capables de synth´etiserrapidement des artefacts logiciels pour une large gamme de plateformes logicielles et mat´erielles. En outre, les g´en´erateurs modernes comme les compilateurs C, sont devenus hautement configurables, offrant de nombreuses options de configuration `al'utilisateur de mani`ere`apersonnaliser facilement le code g´en´er´epour la plateforme mat´eriellecible. Par cons´equent, la qualit´elogicielle est devenue fortement corr´el´eeaux param`etresde configuration ainsi qu'au g´en´erateurlui- m^eme.Dans ce contexte, il est devenu indispensable de v´erifierle bon comportement des g´en´erateurs.D'une part, il est important de tester les programmes g´en´er´esafin de d´etecter des anomalies de g´en´erateurset les corriger. D'autre
Load more

Recommended publications
  • Loop Pipelining with Resource and Timing Constraints
    UNIVERSITAT POLITÈCNICA DE CATALUNYA LOOP PIPELINING WITH RESOURCE AND TIMING CONSTRAINTS Autor: Fermín Sánchez October, 1995 1 1 1 1 1 1 1 1 2 1• SOFTWARE PIPELINING 1 1 1 2.1 INTRODUCTION 1 Software pipelining [ChaSl] comprises a family of techniques aimed at finding a pipelined schedule of the execution of loop iterations. The pipelined schedule represents a new loop body which may contain instructions belonging to different iterations. The sequential execution of the schedule takes less time than the sequential execution of the iterations of the loop as they were initially written. 1 In general, a pipelined loop schedule has the following characteristics1: • All the iterations (of the new loop) are executed in the same fashion. • The initiation interval (ÍÍ) between the issuing of two consecutive iterations is always the same. 1 Figure 2.1 shows an example of software pipelining a loop. The DDG representing the loop body to pipeline is presented in Figure 2.1(a). The loop must be executed ten times (with an iteration index i G [0,9]). Let us assume that all instructions in the loop are executed in a single cycle 1 and a new iteration may start every cycle (otherwise the dependence between instruction A from 1 We will assume here that the schedule contains a unique iteration of the loop. 1 15 1 1 I I 16 CHAPTER 2 I I time B,, Prol AO Q, BO A, I Q, B, A2 3<i<9 I Steady State D4 B7 I Epilogue Cy I D9 (a) (b) (c) I Figure 2.1 Software pipelining a loop (a) DDG representing a loop body (b) Parallel execution of the loop (c) New parallel loop body I iterations i and i+ 1 will not be honored).- With this assumption,-the loop can be executed in a I more parallel fashion than the sequential one, as shown in Figure 2.1(b).
    [Show full text]
  • Research of Register Pressure Aware Loop Unrolling Optimizations for Compiler
    MATEC Web of Conferences 228, 03008 (2018) https://doi.org/10.1051/matecconf/201822803008 CAS 2018 Research of Register Pressure Aware Loop Unrolling Optimizations for Compiler Xuehua Liu1,2, Liping Ding1,3 , Yanfeng Li1,2 , Guangxuan Chen1,4 , Jin Du5 1Laboratory of Parallel Software and Computational Science, Institute of Software, Chinese Academy of Sciences, Beijing, China 2University of Chinese Academy of Sciences, Beijing, China 3Digital Forensics Lab, Institute of Software Application Technology, Guangzhou & Chinese Academy of Sciences, Guangzhou, China 4Zhejiang Police College, Hangzhou, China 5Yunnan Police College, Kunming, China Abstract. Register pressure problem has been a known problem for compiler because of the mismatch between the infinite number of pseudo registers and the finite number of hard registers. Too heavy register pressure may results in register spilling and then leads to performance degradation. There are a lot of optimizations, especially loop optimizations suffer from register spilling in compiler. In order to fight register pressure and therefore improve the effectiveness of compiler, this research takes the register pressure into account to improve loop unrolling optimization during the transformation process. In addition, a register pressure aware transformation is able to reduce the performance overhead of some fine-grained randomization transformations which can be used to defend against ROP attacks. Experiments showed a peak improvement of about 3.6% and an average improvement of about 1% for SPEC CPU 2006 benchmarks and a peak improvement of about 3% and an average improvement of about 1% for the LINPACK benchmark. 1 Introduction fundamental transformations, because it is not only performed individually at least twice during the Register pressure is the number of hard registers needed compilation process, it is also an important part of to store values in the pseudo registers at given program optimizations like vectorization, module scheduling and point [1] during the compilation process.
    [Show full text]
  • CS153: Compilers Lecture 19: Optimization
    CS153: Compilers Lecture 19: Optimization Stephen Chong https://www.seas.harvard.edu/courses/cs153 Contains content from lecture notes by Steve Zdancewic and Greg Morrisett Announcements •HW5: Oat v.2 out •Due in 2 weeks •HW6 will be released next week •Implementing optimizations! (and more) Stephen Chong, Harvard University 2 Today •Optimizations •Safety •Constant folding •Algebraic simplification • Strength reduction •Constant propagation •Copy propagation •Dead code elimination •Inlining and specialization • Recursive function inlining •Tail call elimination •Common subexpression elimination Stephen Chong, Harvard University 3 Optimizations •The code generated by our OAT compiler so far is pretty inefficient. •Lots of redundant moves. •Lots of unnecessary arithmetic instructions. •Consider this OAT program: int foo(int w) { var x = 3 + 5; var y = x * w; var z = y - 0; return z * 4; } Stephen Chong, Harvard University 4 Unoptimized vs. Optimized Output .globl _foo _foo: •Hand optimized code: pushl %ebp movl %esp, %ebp _foo: subl $64, %esp shlq $5, %rdi __fresh2: movq %rdi, %rax leal -64(%ebp), %eax ret movl %eax, -48(%ebp) movl 8(%ebp), %eax •Function foo may be movl %eax, %ecx movl -48(%ebp), %eax inlined by the compiler, movl %ecx, (%eax) movl $3, %eax so it can be implemented movl %eax, -44(%ebp) movl $5, %eax by just one instruction! movl %eax, %ecx addl %ecx, -44(%ebp) leal -60(%ebp), %eax movl %eax, -40(%ebp) movl -44(%ebp), %eax Stephen Chong,movl Harvard %eax,University %ecx 5 Why do we need optimizations? •To help programmers… •They write modular, clean, high-level programs •Compiler generates efficient, high-performance assembly •Programmers don’t write optimal code •High-level languages make avoiding redundant computation inconvenient or impossible •e.g.
    [Show full text]
  • Fast-Path Loop Unrolling of Non-Counted Loops to Enable Subsequent Compiler Optimizations∗
    Fast-Path Loop Unrolling of Non-Counted Loops to Enable Subsequent Compiler Optimizations∗ David Leopoldseder Roland Schatz Lukas Stadler Johannes Kepler University Linz Oracle Labs Oracle Labs Austria Linz, Austria Linz, Austria [email protected] [email protected] [email protected] Manuel Rigger Thomas Würthinger Hanspeter Mössenböck Johannes Kepler University Linz Oracle Labs Johannes Kepler University Linz Austria Zurich, Switzerland Austria [email protected] [email protected] [email protected] ABSTRACT Non-Counted Loops to Enable Subsequent Compiler Optimizations. In 15th Java programs can contain non-counted loops, that is, loops for International Conference on Managed Languages & Runtimes (ManLang’18), September 12–14, 2018, Linz, Austria. ACM, New York, NY, USA, 13 pages. which the iteration count can neither be determined at compile https://doi.org/10.1145/3237009.3237013 time nor at run time. State-of-the-art compilers do not aggressively optimize them, since unrolling non-counted loops often involves 1 INTRODUCTION duplicating also a loop’s exit condition, which thus only improves run-time performance if subsequent compiler optimizations can Generating fast machine code for loops depends on a selective appli- optimize the unrolled code. cation of different loop optimizations on the main optimizable parts This paper presents an unrolling approach for non-counted loops of a loop: the loop’s exit condition(s), the back edges and the loop that uses simulation at run time to determine whether unrolling body. All these parts must be optimized to generate optimal code such loops enables subsequent compiler optimizations. Simulat- for a loop.
    [Show full text]
  • A Hierarchical Region-Based Static Single Assignment Form
    A Hierarchical Region-Based Static Single Assignment Form Jisheng Zhao and Vivek Sarkar Department of Computer Science, Rice University fjisheng.zhao,[email protected] Abstract. Modern compilation systems face the challenge of incrementally re- analyzing a program’s intermediate representation each time a code transforma- tion is performed. Current approaches typically either re-analyze the entire pro- gram after an individual transformation or limit the analysis information that is available after a transformation. To address both efficiency and precision goals in an optimizing compiler, we introduce a hierarchical static single-assignment form called Region Static Single-Assignment (Region-SSA) form. Static single assign- ment (SSA) form is an efficient intermediate representation that is well suited for solving many data flow analysis and optimization problems. By partitioning the program into hierarchical regions, Region-SSA form maintains a local SSA form for each region. Region-SSA supports a demand-driven re-computation of SSA form after a transformation is performed, since only the updated region’s SSA form needs to be reconstructed along with a potential propagation of exposed defs and uses. In this paper, we introduce the Region-SSA data structure, and present algo- rithms for construction and incremental reconstruction of Region-SSA form. The Region-SSA data structure includes a tree based region hierarchy, a region based control flow graph, and region-based SSA forms. We have implemented in Region- SSA form in the Habanero-Java (HJ) research compiler. Our experimental re- sults show significant improvements in compile-time compared to traditional ap- proaches that recompute the entire procedure’s SSA form exhaustively after trans- formation.
    [Show full text]
  • The Effect of C Ode Expanding Optimizations on Instruction Cache Design
    May 1991 UILU-ENG-91-2227 CRHC-91-17 Center for Reliable and High-Performance Computing " /ff ,./IJ 5; F.__. =, _. I J THE EFFECT OF C ODE EXPANDING OPTIMIZATIONS ON INSTRUCTION CACHE DESIGN William Y. Chen Pohua P. Chang Thomas M. Conte Wen-mei W. Hwu Coordinated Science Laboratory College of Engineering UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN Approved for Public Release. Distribution Unlimited. UNCLASSIFIED ;ECURIrY CLASSIFICATION OF THiS PAGE REPORT DOCUMENTATION PAGE ii la. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS Unclassified None 2a. SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION / AVAILABILITY OF REPORT Approved for public release; 2b. DECLASSIFICATION I DOWNGRADING SCHEDULE distribution unlimited 4, PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S) UILU-ENG-91-2227 CRHC-91-17 _.NAME OFPERFORMING ORGANIZATION 7a. NAME OF MONITORING ORGANIZATION Coordinated Science Lab (If ap_lic,_ble) NCR, NSF, AI_, NASA University of Illinois 6b. OFFICEN/A SYMBOL 6cAOORESS(O_,S_,andZ;PCode) 7b. AOORESS(O_,$tate, andZIPCodc) Dayton, OH 45420 ii01 W. Springfield Avenue Washington DC 20550 Langley VA 20200 Urbana, IL 61801 NAME OF FUNDING/SPONSORING 18b.OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER NO0014-9 l-J- 1283 NASA NAG 1-613 ORGANIZATION 7a I (If aDplitable) 8c. ADDRESS (City, State, and ZlP Code) 10. SOURCE OF FUNDING NUMBERS WORK UNIT ELEMENT NO. NO. ACCESSION NO. 7b I 11. TITLE (Include Securf(y C/aclrification) The Effect of Code Expanding Optimizations on Instruction Cache Design I 12. PERSONAL AUTHOR(S) Chen, William, Pohua Chang, Thomas Conte and Wen-Mei Hwu 13a. TYPE OF REPORT i13b, TIME COVERED 114.
    [Show full text]
  • A Tiling Perspective for Register Optimization Fabrice Rastello, Sadayappan Ponnuswany, Duco Van Amstel
    A Tiling Perspective for Register Optimization Fabrice Rastello, Sadayappan Ponnuswany, Duco van Amstel To cite this version: Fabrice Rastello, Sadayappan Ponnuswany, Duco van Amstel. A Tiling Perspective for Register Op- timization. [Research Report] RR-8541, Inria. 2014, pp.24. hal-00998915 HAL Id: hal-00998915 https://hal.inria.fr/hal-00998915 Submitted on 3 Jun 2014 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. A Tiling Perspective for Register Optimization Łukasz Domagała, Fabrice Rastello, Sadayappan Ponnuswany, Duco van Amstel RESEARCH REPORT N° 8541 May 2014 Project-Teams GCG ISSN 0249-6399 ISRN INRIA/RR--8541--FR+ENG A Tiling Perspective for Register Optimization Lukasz Domaga la∗, Fabrice Rastello†, Sadayappan Ponnuswany‡, Duco van Amstel§ Project-Teams GCG Research Report n° 8541 — May 2014 — 21 pages Abstract: Register allocation is a much studied problem. A particularly important context for optimizing register allocation is within loops, since a significant fraction of the execution time of programs is often inside loop code. A variety of algorithms have been proposed in the past for register allocation, but the complexity of the problem has resulted in a decoupling of several important aspects, including loop unrolling, register promotion, and instruction reordering.
    [Show full text]
  • Graduate Computer Architecture Chapter
    Graduate Computer Architecture Chapter 4 – Explore Instruction Level Parallelism with Software Approaches 1 Basic Pipeline Scheduling Find sequences of unrelated instructions Compiler’s ability to schedule – Amount of ILP available in the program – Latencies of the functional units Latency assumptions for the examples – Standard MIPS integer pipeline – No structural hazards (fully pipelined or duplicated units – Latencies of FP operations: Instruction producing result Instruction using result Latency FP ALU op FP ALU op 3 FP ALU op SD 2 LD FP ALU op 1 LD SD 0 2 Basic Scheduling for (i = 1000; i > 0; i=i-1) Sequential MIPS Assembly Code Loop: LD F0, 0(R1) x[i] = x[i] + s; ADDD F4, F0, F2 SD 0(R1), F4 SUBI R1, R1, #8 BNEZ R1, Loop Pipelined execution: Scheduled pipelined execution: Loop: LD F0, 0(R1) 1 Loop: LD F0, 0(R1) 1 stall 2 SUBI R1, R1, #8 2 ADDD F4, F0, F2 3 ADDD F4, F0, F2 3 stall 4 stall 4 stall 5 BNEZ R1, Loop 5 SD 0(R1), F4 6 SD 8(R1), F4 6 SUBI R1, R1, #8 7 stall 8 BNEZ R1, Loop 9 stall 10 3 Static Branch Prediction: Using Compiler Technology How to statically predict branches? –Examination of program behavior Always predict taken (on average, 67% are taken) –Mis-prediction rate varies large (9%—59%) Predict backward branches taken, forward branches un-taken (mis-prediction rate < 30% -- 40%) –Profile-based predictor: use profile information collected from earlier runs 4 Mis-prediction Rate for a Profile-based Predictor FP is better than integer 5 Prediction-taken VS.
    [Show full text]
  • The Effect of Code Expanding Optimizations on Instruction Cache Design
    May 1991 UILU-EN G-91-2227 CRHC-91-17 Center for Reliable and High-Performance Computing THE EFFECT OF CODE EXPANDING OPTIMIZATIONS ON INSTRUCTION CACHE DESIGN William Y. Chen Pohua P. Chang Thomas M. Conte Wen-mei W. Hwu Coordinated Science Laboratory College of Engineering UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN Approved for Public Release. Distribution Unlimited. UNCLASSIFIED____________ SECURITY ¿LASSIPlOvriON OF t h is PAGE REPORT DOCUMENTATION PAGE 1a. REPORT SECURITY CLASSIFICATION 1b. RESTRICTIVE MARKINGS Unclassified None 2a. SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/AVAILABILITY OF REPORT 2b. DECLASSIFICATION/DOWNGRADING SCHEDULE Approved for public release; distribution unlimited 4. PERFORMING ORGANIZATION REPORT NUMBER(S) 5. MONITORING ORGANIZATION REPORT NUMBER(S) UILU-ENG-91-2227 CRHC-91-17 6a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION Coordinated Science Lab (If applicable) NCR, NSF, AMD, NASA University of Illinois N/A 6c ADDRESS (G'ty, Staff, and ZIP Code) 7b. ADDRESS (Oty, Staff, and ZIP Code) 1101 W. Springfield Avenue Dayton, OH 45420 Urbana, IL 61801 Washington DC 20550 Langley VA 20200 8a. NAME OF FUNDING/SPONSORING 8b. OFFICE SYMBOL ORGANIZATION 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER 7a (If applicable) N00014-91-J-1283 NASA NAG 1-613 8c. ADDRESS (City, State, and ZIP Cod*) 10. SOURCE OF FUNDING NUMBERS PROGRAM PROJECT t a sk WORK UNIT 7b ELEMENT NO. NO. NO. ACCESSION NO. The Effect of Code Expanding Optimizations on Instruction Cache Design 12. PERSONAL AUTHOR(S) Chen, William, Pohua Chang, Thomas Conte and Wen-Mei Hwu 13a. TYPE OF REPORT 13b. TIME COVERED 14. OATE OF REPORT (Year, Month, Day) Jl5.
    [Show full text]
  • LLVM Performance Optimization for Z Systems
    IBM Linux Technology Center LLVM performance optimization for z Systems Dr. Ulrich Weigand Senior Technical Staff Member GNU/Linux Compilers & Toolchain Date: Mar 27, 2017 © 2017 IBM Corporation IBM Linux Technology Center Agenda • LLVM on z Systems Performance History • Instruction-Set Architecture (ISA) Optimization • Processor Micro-Architecture Optimization • Outlook 2 © 2017 IBM Corporation IBM Linux Technology Center Performance History 3 © 2017 IBM Corporation LLVM on z Systems – performance history LLVM Performance vs. GCC pre-7.0 (Mar 2017) 4 -25% -20% -15% -10% -5% 0% IBM Linux Technology Center IBM Technology Linux 3.3 3.4 3.5.0 3.6.0 LLVM Version LLVM 3.7.0 © 2017 © IBM2017 Corporation 3.8.0 3.9.0 4.0.0 IBM Linux Technology Center 0% -5% Back-end changes: overview -10% -15% • LLVM 3.3 – Initial release -20% – No focus on performance; z10 only -25% • LLVM 3.4 – ISA exploitation 3.3 3.4 3.5.0 3.6.0 3.7.0 3.8.0 3.9.0 4.0.0 – Significantly improved z10 code generation; initial z196 & zEC12 support • LLVM 3.5 – Further ISA exploitation – More z196 & zEC12 instructions exploited • LLVM 3.6 – No performance-related changes • LLVM 3.7 – z13 vector ISA support • LLVM 3.8 – Only minor performance-related changes – Small improvements to floating-point code generation • LLVM 3.9 – Misc. code-gen changes / start of micro-arch tuning – Avoid false FPR dependencies, conditional sibcall/return, FP test data class • LLVM 4.0 – Focus on micro-architecture tuning – Post-RA scheduler, tune loop unrolling / strength reduction, tune load-on-condition
    [Show full text]
  • That Increase Instruction-Level Parallelism 1 Introduction
    Balanced Scheduling with Compiler Optimizations that Increase Instruction-Level Parallelism Jack L, LO and Susan J. Eggers Department of Computer Science and Engineering University of Washington { -j 10, eggers} @cs . washi-rlgton. edu Abstract and becomes uncertain: not only will the processor see both cache hits and misses, but each level in the memory hierarchy will also Traditional list schedulers order instructions based on an optimistic introduce a new set of Iatencies. Instead of handling all loads iden- estimate of the load latency imposed by the hardware and therefore tically, a non-blocking processor’s code scheduler could arrange cannot respond to variations in memory latency caused by cache instructions behind loads more intelligently to produce more effi- hits and misses on non-blocking architectures. In contrast, bal- cient code schedules. Unfortunately, a traditional instruction anced scheduling schedules instructions based on an estimate of scheduler fails to exploit thk opportunity because of its optimistic and architecture-based estimates; its resulting schedules may be the amount of instruction-level parallelism in the program, By scheduling independent instructions behind loads based on what therefore intolerant of variations in load latency. the program can provide, rather than what the implementation stip- Balanced scheduling [KE93] is an algorithm that can generate ulates in the best case (i.e., a cache hit), balanced scheduling can schedules that adapt more readily to the uncertainties in memory hide variations in memory latencies more effectively. latency. Rather than being determined by a predefine, architecttrr- ally-based value, load latency estimates are based on the number Since its success depends on the amount of instruction-level paral- of independent instructions that are available to hide the latency of lelism in the code, balanced scheduling should perform even better when more parallelism is available.
    [Show full text]
  • Code Scheduling
    Register Allocation (wrapup) & Code Scheduling CS2210 Lecture 22 CS2210 Compiler Design 2004/5 Constructing and Representing the Interference Graph ■ Construction alternatives: ■ as side effect of live variables analysis (when variables are units of allocation) ■ compute du-chains & webs (or ssa form), do live variables analysis, compute interference graph ■ Representation ■ adjacency matrix: A[min(i,j), max(i,j)] = true iff (symbolic) register i adjacent to j (interferes) CS2210 Compiler Design 2004/5 Adjacency List ■ Used in actual allocation ■ A[i] lists nodes adjacent to node i ■ Components ■ color ■ disp: displacement (in stack for spilling) ■ spcost: spill cost (initialized 0 for symbolic, infinity for real registers) ■ nints: number of interferences ■ adjnds: list of real and symbolic registers currently interfere with i ■ rmvadj: list of real and symbolic registers that interfered withi bu have been removed CS2210 Compiler Design 2004/5 1 Register Coalescing ■ Goal: avoid unnecessary register to register copies by coalescing register ■ ensure that values are in proper argument registers before procedure calls ■ remove unnecessary copies introduced by code generation from SSA form ■ enforce source / target register constraints of certain instructions (important for CISC) ■ Approach: ■ search for copies si := sj where si and sj do not interfere (may be real or symbolic register copies) CS2210 Compiler Design 2004/5 Computing Spill Costs ■ Have to spill values to memory when not enough registers can be found (can’t find k- coloring)
    [Show full text]