Proj EULibHarn Paper Reading: Towards Efficient Large-Scale Interprocedural Program Static Analysis on Distributed Data-Parallel Computation
Abstract
和BigSpa一样https://www.cnblogs.com/xuesu/p/15760169.html
目的：large-scale 进程间分析
问题：计算代价和内存代价搞
本文：data-parallel, a join-process filter computation model
基于CFL可达性的进程间分析
面向云
效果：在百万行代码上能够进行准确进程间分析
I. Intro
II. Background
A. CFL Reachability
B. Two Examples
- Alias Analysis
- Dataflow Analysis
Alias Analysis
Dataflow Analysis
C. Worklist Algo
III. A distributed solution
IV. Offline batch static program analysis system
V. Online incremental static program
VI. System Design and usage
VII. EVALUATION
Q1: What is the performance of BigSpa? (§VI-A)
Q2: How does BigSpa compare to the state-of-the-art single-machine analysis system? (§VI-B)
Q3: How does BigSpa compare to other state-of-the-art distributed systems for analysis workloads? (§VI-C)
