Skip to content

Analysis of memory reuse and sharing profiles of parallel programs

Notifications You must be signed in to change notification settings

Riyuzakii/Memreuse-Analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

31 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Memory Reuse and Sharing Profile Analysis

Discrete analysis of memory reuse and sharing profiles of a set of parallel programs. Memory accesses are obtained through PIN's IARG_MEMORYOP_EA macro, which in turn is broken down into individual x86 machine accesses according to specified rules.

The analysis tool is developed as a part of the Assignment-2 for the course Advanced Computer Architecture (CS622A),
Fall Semester - 2019, instructed by Prof. Mainak Chaudhury.

The results of the simulation are compiled here. And the project report can be found here.

Group Members (G16)

Name Email Roll
Aditya Rohan [email protected] 160053
Aniket Pandey [email protected] 160113

Setup and Build Instructions

Clone the repository in source/tools/CS622Assignment path of pin-3.11 tool.

$ git clone [email protected]:Riyuzakii/Memreuse-Analysis.git <Path-to-Pin>/source/tools/CS622Assignment

Compiling and running the tool is easy. Like, really easy!

The root folder of the project contains a Makefile which is not something we're allowed to touch! As a result, each of the following commands need to be executed from within the scripts directory.

 > cd scripts
 > make setup          # Compiles instrumentation, analysis and target programs

Part 1: Collecting machine access trace

 > make run PART=1 PROG=x                   # Collects access trace for progx.c

where x ∈ {1, 2, 3, 4}. This will create addrtrace.out inside the result directory. Note: All subsequent parts will use this trace and running the above command with PROG=y where y ≠ x will overwrite the trace.

Part 2, 3, 4: Analysis of the obtained trace

 > make run PART=x

where x ∈ {2, 3, 4}. This will use the trace generated above.

About

Analysis of memory reuse and sharing profiles of parallel programs

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published