1 / 40

A Probabilistic Pointer Analysis for Speculative Optimizations

Electrical and Computer Engineering University of Toronto. A Probabilistic Pointer Analysis for Speculative Optimizations. Jeff DaSilva Greg Steffan. ASPLOS XII. foo(int *a) { … while(…) { x = *a; … } } . Pointers Impede Optimization. Loop Invariant Code Motion.

keaton
Download Presentation

A Probabilistic Pointer Analysis for Speculative Optimizations

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Electrical and Computer Engineering University of Toronto A Probabilistic Pointer Analysis for Speculative Optimizations Jeff DaSilva Greg Steffan ASPLOS XII

  2. foo(int *a) { … while(…) { x = *a; … } } Pointers Impede Optimization Loop Invariant Code Motion Parallelize Can pointer analysis help? ASPLOS XII

  3. optimize Definitely Pointer Analysis Definitely Not Maybe *a = ~ ~ = *b How can we optimize cases? Maybe Pointer Analysis • Do pointers a and b point to the same location? • Repeat for every pair of pointers at every program point *a = ~ ~ = *b ASPLOS XII

  4. int *a, x; … while(…) { x = *a; … } int *a, x, tmp; … tmp = *a; while(…) { x = tmp; … } <verify, recover?> a is probably loop invariant Lets Speculate • Implement a potentially unsafe optimization • Verify and Recover if necessary ASPLOS XII

  5. Data Speculative Optimizations • EPIC Instruction sets • Support for speculative load/store instructions (eg. Itanium) • Speculative compiler optimizations • Dead store elimination, redundancy elimination, copy propagation, strength reduction, register promotion • Thread-level speculation (TLS) • Hardware and compiler support for speculative parallel threads • Transactional programming • Hardware and software support for speculative parallel transactions Heavy reliance on detailed profile feedback ASPLOS XII

  6. Maybe Recovery penalty (if unsuccessful) Overhead for verify Probability of success Expected speedup (if successful) Speculate? Ideally would be a probability Maybe Can We Quantify ? Maybe • Estimate the potential benefit for speculating: ASPLOS XII

  7. optimize optimize p = 1.0 Definitely Pointer Analysis PPA p = 0.0 Definitely Not 0.0 < p < 1.0 Maybe *a = ~ ~ = *b Probabilistic Pointer Analysis (PPA) Conventional Pointer Analysis • Do pointers a and b point to the same location? • Do this for every pair of pointers at every program point *a = ~ ~ = *b • With what probability p, do pointers a and b point to the same location? • Repeat for every pair of pointers at every program point Potential bonus: PPA doesn’t have to be safe ASPLOS XII

  8. PPA Research Objectives • Accurate points-to probability information • at every static pointer dereference • Scalable analysis • Goal: entire SPEC integer benchmark suite • Understand scalability/accuracytradeoff • through flexible static memory model Improve our understanding of programs ASPLOS XII

  9. Related Work ASPLOS XII

  10. Algorithm Design Choices • Fixed • Bottom Up / Top Down Approach • Linear transfer functions (for scalability) • One-level context and flow sensitive • Flexible • Edge profiling (or static prediction) • Safe (or unsafe) • Field sensitive (or field insensitive) ASPLOS XII

  11. int x, y, z, *b = &x; void foo(int *a) { if(…) b = &y; if(…) a = &z; else(…) a = b; while(…) { x = *a; … } } = Definitely = pointer = = pointed at Maybe Traditional Points-To Graph b a x y z UND Results are inconclusive ASPLOS XII

  12. int x, y, z, *b = &x; void foo(int *a) { if(…) b = &y; if(…) a = &z; else a = b; while(…) { x = *a; … } } = p = 1.0 = pointer p = = pointed at 0.0<p< 1.0 Probabilistic Points-To Graph 0.1 taken(edge profile) b a 0.2 taken(edge profile) 0.72 0.9 0.1 0.2 0.08 x y z UND Results provide more information ASPLOS XII

  13. Linear One - Level Interprocedural Probabilistic Pointer Analysis LOLLIPOPOur PPA Algorithm ASPLOS XII

  14. Fundamentals • Location sets (Wilson & Lam) • One or more mem locations • Classified as pointer, pointer target, or both • Matrix-based approach • Nice properties, optimized implementations • Two key matrices: • Points-to matrix • Linear transformation matrix ASPLOS XII

  15. Points-To Matrix Location Sets M-1 M Area Of Interest 1 2 … Pointer Sets … N-1 N I Location Sets All matrix rows sum to 1.0 ASPLOS XII

  16. x y z UND a b a b 0.72 0.9 0.1 0.2 x 0.08 y x y z UND z UND Points-To Matrix Example 0.72 0.08 0.20 0.90 0.10 1.0 1.0 1.0 1.0 I ASPLOS XII

  17. I I Computing New Points-To Matrices Points-To Matrix In Any Instruction Points-To Matrix Out ASPLOS XII

  18. The Fundamental PPA Equation Transformation Matrix Points-To Matrix In Points-To Matrix Out = This can be applied to any instruction(incl. function calls) ASPLOS XII

  19. Transformation Matrix Location Sets Pointer Sets 1 2 3 … N-1 N 1 2 Area of Interest … Pointer Sets ø I Location Sets … N-1 N All matrix rows sum to 1.0 ASPLOS XII

  20. a b x y z UND Transformation Matrix Example a b x y z UND S1:a = &z; 1.0 1.0 TS1 1.0 1.0 1.0 1.0 = ASPLOS XII

  21. PTout TS1 PTin = x y z UND a 1.0 1.0 0.72 0.08 0.20 0.90 0.10 b a b 0.72 0.9 0.1 0.2 x 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.08 y x y z UND z UND Example - The PPA Equation S1:a = &z; PTout = ASPLOS XII

  22. PTout TS1 PTin = x y z UND a b b a 0.9 0.1 x y x y z UND z UND Example - The PPA Equation S1:a = &z; 1.0 0.90 0.10 PTout = 1.0 1.0 1.0 1.0 I ASPLOS XII

  23. Basic Block S1: Instr S2: Instr S3: Instr I I Combining Transformation Matrices PTout TS2 TS3 TS1 PTin PTin TS1 PTin = PTout TBB PTin = ASPLOS XII

  24. Control flow - if/else p + q = 1.0 X Y p p q q TX TY + = ASPLOS XII

  25. N <L,U> X Y i 1 TY = U-L+1 Control flow - loops N TX = Both operations can be implemented efficiently <L,U>  <min,max> ASPLOS XII

  26.   Safe vs. Unsafe Pointer Assignment Instructions Safe?   Shadow/invisible vars: non-linear to handle ASPLOS XII

  27. Steensgaard FICI PA for shadow vars p p Safe q p_1 p_1 q + = p_1 q p_1 and q are considered aliased Safety for Shadow Variables Unsafe q q = pointer = shadow ptr = pointed at ASPLOS XII

  28. LOLLIPOP Implementation .spd SUIF Infrastructure Static Memory Model TF-Matrix Collector Points-To Matrix Propagator ICFG SMM BU TD Edge Profile .spx • Implementation details: • Sparse matrices • Efficient matrix algorithms • Result memoization Results Stats MATLAB C Library ASPLOS XII

  29. Measuring LOLLIPOP’s Efficiency and Accuracy ASPLOS XII

  30. SPEC2000 Benchmark Data Experimental Framework: 3GHz P4 with 1GB of RAM Scales to all of SPECint ASPLOS XII

  31. Reduced Input Set SPEC2000 Benchmark Data Experimental Framework: 3GHz P4 with 1GB of RAM More efficient than points-to profiling ASPLOS XII

  32. Average Dereference Size more accurate Easy SPEC2000 Benchmarks A one-level Analysis is often adequate (i.e. safe~=unsafe) ASPLOS XII

  33. Average Dereference Size more accurate Challenging SPEC 95/2000 Benchmarks Many improbable points-to relations can be pruned away ASPLOS XII

  34. while(…) { x = *a; … } { (0.72, ), (0.08, ), (0.2, ) } x y z Σ (max probability value) = Avg Certainty (num of loads & stores) Metric: Average Max Certainty Max probability value = 0.72 ASPLOS XII

  35. Probabilistic Certainty more accurate SPEC2000 Average Certainty On average, LOLLIPOP can predict a single likely points-to relation ASPLOS XII

  36. while(…) { x = *a; … } { (0.72, ), (0.08, ), (0.2, ) } { (0.88, ), (0.02, ), (0.1, ) } x x y y z z Pd Ps Σ ║Ps – Pd ║ 1 = NAED √2 (num of loads & stores) Metric: NAED • Normalized Average Euclidean Distance LOLIPoP’s Static prediction Pts-to Profiler’s Dynamic freq vector ASPLOS XII

  37. more accurate NAED Norm Avg Euclidean Dist (NAED) ASPLOS XII

  38. Summary • A novel matrix-based PPA algorithm • Scales to the SPECint 95/2000 benchmarks • One level context and flow sensitive • Future Work • Applying LOLLIPOP to TLS/TM • Further optimize LOLLIPOP’s implementation ASPLOS XII

  39. ASPLOS XII

  40. References • Manuvir Das, Ben Liblit, Manuel Fahndrich, and Jakob Rehof. Estimating the Impact of Scalable Pointer Analysis on Optimization. SAS 2001, 260-278. • Peng-Sheng Chen, Ming-Yu Hung, Yuan-Shin Hwang, Roy Dz-Ching Ju, and Jenq Kuen Lee. Compiler support for speculative multithreading architecture with probabilistic points-to analysis. PPOPP 2003, 25-36. • Jin Lin, Tong Chen, Wei-Chung Hsu, Peng-Chung Yew, Roy Dz-Ching Ju, Tin-Fook Ngai and Sun Chan, A Compiler Framework for Speculative Analysis and Optimizations. PLDI 2003, 289-299. • R.D. Ju, J. Collard, and K. Oukbir. Probabilistic Memory Disambiguation and its Application to Data Speculation. SIGARCH Comput. Archit. News 27 1999, 27-30. • Manel Fernandez and Roger Espasa. Speculative Alias Analysis for Executable Code. PACT 2002, 221-231. ASPLOS XII

More Related