1 / 19

Guiding Ispike with Instrumentation and Hardware (PMU) Profiles CGO’04 Tutorial 3/21/04

Ispike is a post-link optimizer designed for the Itanium architecture on Linux, requiring no source code. It leverages memory-centric optimizations like code and data layout management to achieve an average speedup of 10% over compiler-optimized programs such as GCC -O3 on SPEC CINT 2000. The tool provides insights into program characteristics, automates drive optimizations, and evaluates their effectiveness. This tutorial discusses various profiling techniques, tools for analyzing program performance, and methods for deriving new profiles from performance monitoring units (PMUs).

caine
Télécharger la présentation

Guiding Ispike with Instrumentation and Hardware (PMU) Profiles CGO’04 Tutorial 3/21/04

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. Guiding Ispike with Instrumentation and Hardware (PMU) ProfilesCGO’04 Tutorial3/21/04 CK. Luk chi-keung.luk@intel.com Massachusetts Microprocessor Design Center Intel Corporation

  2. What is Ispike? • A post-link optimizer for Itanium/Linux • No source code required • Memory-centric optimizations: • Code layout + prefetching, data layout + prefetching • Significant speedups over compiler-optimized programs: • 10% average speedup over gcc –O3 on SPEC CINT 2000 • Profile usages: • Understanding program characteristics • Driving optimizations automatically • Evaluating the effectiveness of optimizations CGO’04 Tutorial

  3. Profiles used by Ispike CGO’04 Tutorial

  4. Profile Example: D-EAR (cache) Top 10 loads in the D-EAR profile of the MCF benchmark Total sampled miss latency latency buckets CGO’04 Tutorial

  5. Profile Analysis Tools • A set of tools written for visualizing and analyzing profiles, e.g.,: • Control flow graph (CFG) viewer • Code-layout viewer • Load-latency comparator CGO’04 Tutorial

  6. CFG Viewer C For evaluating the accuracy of profiles CGO’04 Tutorial

  7. Code-layout Viewer C For evaluating code-layout optimization CGO’04 Tutorial

  8. Load-latency Comparator • For evaluating data-layout optimization and data prefetching CGO’04 Tutorial

  9. Deriving New Profiles from PMUs • New profile types can be derived from PMUs • Two examples: • Consumer stall cycles • D-cache miss strides CGO’04 Tutorial

  10. Consumer Stall Cycles Basic block A PC-sample count N1 I1: ld8 r2 = [r3];; /* other instructions */ I2: add r2 = r2, 1;; I3: st8 [r3] = r2 Question: • How many cycles of stall experienced by I2? (Note: not necessarily the load latency of I1) Method: • PC-sample count is proportional to (stall cycles * frequency) N2 N3 CGO’04 Tutorial

  11. Example: Two strided loads in MCF arc* arcin; node* tail; … while (arcin) { tail = arcin->tail; … arcin = tail->mark; } -192B -192B -192B arcin tail -120B -120B -120B D-cache Miss Strides Problem: • Detect strides that are statically unknown CGO’04 Tutorial

  12. Skipping phases (1 sample per 1000 misses) Time Inspection phases (1 sample per miss) A1 A2 A3 A4 Time A4-A3=3*48=144 A2-A1=5*48=240 A3-A2=7*48=336 Use GCD to figure out strides from miss addresses: A1, A2, A3, A4 are four consecutive miss addresses of a load. The load has a stride of 48 bytes. GCD(A2-A1, A3-A2 )=GCD(240,336)=48 GCD(A3-A2, A4-A3 )=GCD(336,144)=48 D-EAR based Stride Profiling • Sample load misses with 2 phases: CGO’04 Tutorial

  13. Performance Evaluation • Instrumentation vs. PMU profiles: • Profiling overhead • Performance impact • Ispike optimizations: • Code layout, instruction prefetching, data layout, data prefetching, inlining, global-data optimization, scalar optimizations • Baseline compilers: • Intel Electron compiler (ecc), version 8.0 Beta, -O3 • GNU C compiler (gcc), version 3.2, -O3 • Benchmarks: • SPEC CINT2000 (profiled with “training”, measured with “reference”) • System: • 1GHz Itanium 2, 16KB L1I/16KB L1D, 256KB L2, 3MB L3, 16GB memory • Red Hat Enterprise Linux AS with 2.4.18 kernel CGO’04 Tutorial

  14. Performance Gains with PMU Profiles BTB (1 sample/10K branches), D-EAR cache (1 sample/100 load misses) D-EAR stride (1 sample /100 misses in skipping, 1 sample/miss in inspection) • Up to 40% gain • Geo. means: 8.5% over Ecc and 9.9% over Gcc Gcc3.2 –O3 baseline Ecc8.0 –O3 baseline CGO’04 Tutorial

  15. Cycle Breakdown (Ecc Baseline) • Help understand if individual optimizations are doing a good job CGO’04 Tutorial

  16. PMU Profiling Overhead • Overhead reduced from 58% to 23% when lowering the BTB sampling rate by 10x. • Overhead reduced to 3% when lowering the D-EAR sampling rate by 10x. CGO’04 Tutorial

  17. Instrumentation Profiling Overhead Why is the overhead so large? • Training runs are too short to amortize the dynamic compilation cost • Techniques like ephemeral instrumentation yet to be applied CGO’04 Tutorial

  18. PMU vs. Instrumentation (Perf. Gains) profiling overhead >60x • PMU profiles can be as good as instrumentation profiles • Could be even better in some cases (e.g., mcf) • However, possible performance drops when samples are too sparse • E.g., gap and parser when Stride = <1/1000, 1/1> 59% 24% 3% CGO’04 Tutorial

  19. Reference “Ispike: A Post-link Optimizer for the Intel Itanium Architecture”, by Luk et. al. In Proceedings of CGO’04. http://www.cgo.org/papers/01_82_luk_ck.pdf CGO’04 Tutorial

More Related